Analisis de Soluciones Tecnologicas para El Tratamiento de La Osteoartritis

ESCUELA POLITCNICA SUPERIOR DE MONDRAGON
UNIBERTSITATEA
MONDRAGON UNIBERTSITATEKO GOI ESKOLA
POLITEKNIKOA
Trabajo presentado para la obtencin del ttulo de

Titulua eskuratzeko lana
GRADO EN INGENIERA MECNICA

MEKANIKAKO INGENIARITZA GRADUA
Ttulo del Trabajo Lanaren izenburua
ANLISIS DE SOLUCIONES TECNOLGICAS PARA EL

TRATAMIENTO DE LA OSTEOARTRITIS
Autor Egilea DIEGO MAURO POJMAEVICH
Curso Ikasturtea 2015/2016

Ttulo del Trabajo Lanaren izenburua
ANLISIS DE SOLUCIONES TECNOLGICAS PARA EL

TRATAMIENTO DE LA OSTEOARTRITIS
Nombre y apellidos del autor

Egilearen izen-abizenak
POJMAEVICH , DIEGO MAURO
Nombre y apellidos del/los director/es del trabajo

Zuzendariaren/zuzendarien izen-abizenak
GARAI, GAIZKA
Lugar donde se realiza el trabajo

Lana egin deneko lekua
Mondragon Goi Eskola Politeknikoa Jose M Arizmendiarrieta, S.
Coop.
Curso acadmico
Ikasturtea
2015/2016
El autor/la autora del Trabajo Fin de Grado, autoriza a la Escuela Politcnica Superior de Mondragon
Unibertsitatea, con carcter gratuito y con fines exclusivamente de investigacin y docencia, los derechos de
reproduccin y comunicacin pblica de este documento siempre que: se cite el autor/la autora original, el uso
que se haga de la obra no sea comercial y no se cree una obra derivada a partir del original.
Gradu Bukaerako Lanaren egileak, baimena ematen dio Mondragon Unibertsitateko Goi Eskola Politeknikoari Gradu
Bukaerako Lanari jendeaurrean zabalkundea emateko eta erreproduzitzeko; soilik ikerketan eta hezkuntzan erabiltzeko
eta doakoa izateko baldintzarekin. Baimendutako erabilera honetan, egilea nor den azaldu beharko da beti, eragotzita
egongo da erabilera komertziala baita lan originaletatik lan berriak eratortzea ere.
Summary
Summary
Resumen
La osteoartritis es una enfermedad producida por el desgaste del

cartlago, tejido que hace de amortiguador al proteger los extremos de los huesos y que
favorece el movimiento de la articulacin. Es la enfermedad reumtica ms frecuente,
especialmente entre personas de edad avanzada y afecta en mayor medida a mujeres
que a hombres. La osteoartritis de rodilla ha despertado un especial inters en los
ltimos aos, ya que se ha convertido en una de las principales causas de discapacidad
motriz tanto en hombres como en mujeres en todo el mundo.
Este proyecto tuvo como objetivo obtener un modelo 3D de la rodilla de un

paciente afectado por osteoartritis, a partir de imgenes mdicas especficas, para luego
imprimir un prototipo de la misma. Para esto, utilizando una base de datos que contiene
imgenes mdicas provistas por OAI, se seleccion el estudio que contena las
imgenes ms ntidas y ptimas para su procesamiento. Se trabaj con ese estudio y
se logr segmentar correctamente la parte sea de la rodilla utilizando un software
llamado 3D Slicer. Luego de la segmentacin, se lograron generar los archivos
necesarios para la impresin del prototipo deseado.
Luego se concluy en que es posible segmentar una rodilla con osteoartritis

utilizando imgenes mdicas de pacientes reales para luego imprimir un prototipo 3D
que pueda ser utilizado en futuros anlisis.
III
Anlisis de soluciones tecnolgicas para el tratamiento de la osteoartritis
Summary
Summary
Osteoarthritis is a disease caused by loss of cartilage tissue, which acts as a

cushion to protect the ends of bones and helps the movement of the joint. It is the most
common rheumatic disease, especially among older people and affects more women
than men. Knee osteoarthritis has produced particular interest in latest years because it
has become one of the major causes of physical disability in both men and women
worldwide.
This project aimed to obtain a 3D model of the knee of a patient affected by

osteoarthritis, based on specific medical images and then print a prototype of the knee.
For this, using a database containing medical images provided by OAI, the study
containing the sharpest and best for processing images was selected. Then, working
with the study, the bone part of the knee was properly segmented using a software called
3D Slicer. After segmentation, files needed for printing the desired prototype were
generated.
After all, it was concluded that it is possible to segment a knee with osteoarthritis
using actual patient medical images and then print a 3D prototype that can be used in
future analyzes.
IV
Contents
Contents
Contents
Summary ...................................................................................................................... I
Resumen ................................................................................................................. III
Summary .................................................................................................................IV
Contents ......................................................................................................................V
Chapter 1: Introduction .............................................................................................. 1

1.1. Introduction ..................................................................................................... 3
1.2. Goals and project phases ............................................................................... 5
Chapter 2: Knee anatomy and kinematics ................................................................ 7

2.1. Anatomy of the knee joint ............................................................................... 9
2.1.1 Articular surfaces ......................................................................................... 9
2.1.1.1 Lower extremity of the femur ................................................................. 9
2.1.1.2 Upper extremity of the tibia .................................................................. 11
2.1.1.3 Menisci ................................................................................................ 12
2.1.1.4. Patella ................................................................................................ 13
2.1.2. Articular ligaments ..................................................................................... 14
2.1.3. Structures around the joint ........................................................................ 15
2.3.1. Muscles ................................................................................................. 15
2.1.3.2. Arteries ............................................................................................... 16
2.1.3.3. Nerves ................................................................................................ 16
2.1.3.4. Cartilage ............................................................................................. 16
2.2. Kinematics of the knee ................................................................................. 17
2.2.1. Osteokinematics of the knee ..................................................................... 17
2.2.1.1. Flexion/Extension ............................................................................... 17
2.2.1.2. Axial rotation....................................................................................... 18
2.2.2. Arthrokinematics of the knee ..................................................................... 19
2.2.2.1. Rolling and gliding .............................................................................. 19
Chapter 3: Osteoarthritis .......................................................................................... 21

3.1. Physiopathology of osteoarthritis .................................................................. 23
3.1.1. Symptoms ................................................................................................. 24
3.1.2. Causes ...................................................................................................... 24
3.2. Diagnosis and treatment ............................................................................... 25
3.2.1. Physical Activity ........................................................................................ 26

Contents
3.2.2. Weight Management ................................................................................. 27

3.2.3. Pain and Anti-inflammatory Medications .................................................... 27
3.2.4. Assistive Devices ...................................................................................... 27
3.2.5. Surgery ..................................................................................................... 27
3.3. Types of surgeries ........................................................................................ 28
3.3.1. Total knee arthroplasty or total knee replacement ..................................... 28
3.3.2. Unicompartmental knee arthroplasty or partial knee replacement ............. 29
3.3.3. Patellofemoral joint arthroplasty ................................................................ 30
3.3.4. Complex replacement or revision .............................................................. 30
3.4. Surgical approaches ..................................................................................... 31
3.4.1. Traditional Surgery .................................................................................... 31
3.4.2. Minimally Invasive Surgery ........................................................................ 32
3.4.3. Computer-Assisted Surgery (CAS) ............................................................ 33
3.4.4. Risks ......................................................................................................... 33
3.4.4.1. Deep vein thrombosis ......................................................................... 34
3.4.4.2. Loss of Motion .................................................................................... 34
3.4.4.3. Instability ............................................................................................ 34
3.4.4.4. Infection.............................................................................................. 34
3.4.5. Recovery Time .......................................................................................... 35
Chapter 4: Knee prostheses ....................................................................................... 37

4.1. Knee replacement implants .......................................................................... 39
4.1.1. Implant Components ................................................................................. 39
4.1.2. History of total knee arthroplasty ............................................................... 39
4.1.2.1. Beginning ........................................................................................... 40
4.1.2.2. The hinged prosthesis ........................................................................ 40
4.1.2.3. The bicompartmental prosthesis ......................................................... 41
4.1.2.4. The tricompartmental prosthesis......................................................... 41
4.1.2.5. Unicompartmental knee arthroplasty .................................................. 42
4.1.2.6. Patellar resurfacing............................................................................. 44
4.1.3. Types of prostheses designs ..................................................................... 44
4.1.3.1. Posterior cruciate ligament retention vs. substitution .......................... 44
4.1.3.2. Mobile vs. fixed bearing prostheses .................................................... 46
4.1.3.3. Non-cemented vs. cemented prostheses ............................................ 47
4.1.3.4. Constrained condylar prostheses ....................................................... 47
4.1.3.5. High flexion type prostheses ............................................................... 48
4.1.3.6. Ultracongruent polyethylene bearings................................................. 49
4.1.4. Materials ................................................................................................... 50
4.1.4.1. Material Criteria .................................................................................. 50

Contents
4.1.4.2. Most used materials ........................................................................... 50

4.1.4.3. Metals................................................................................................. 50
4.1.4.4. Ceramics ............................................................................................ 54
4.1.4.5. Polymeric Biomaterials ....................................................................... 59
Chapter 5: Manual segmentation of a knee............................................................. 61

5.1. Manual segmentation of a knee ....................................................................... 63
5.1.1. Introduction to 3D Slicer ............................................................................ 64
5.1.2. Loading DICOM files ................................................................................. 66
5.1.3. Bone segmentation ................................................................................... 68
5.1.4. Building 3D model ..................................................................................... 71
5.1.5. Saving model ............................................................................................ 72
Chapter 6: Conclusion.............................................................................................. 73
6.1. Conclusin .................................................................................................... 75
6.2. Conclusion .................................................................................................... 76
6.3. Trabajo futuro ............................................................................................... 77
6.4. Future work................................................................................................... 78
Chapter 7: References .............................................................................................. 79

7.1 References ................................................................................................... 81
Appendix ................................................................................................................... 83
APPENDIX A: DICOM Image Release Notes .......................................................... 85
APPENDIX B: 3D Slicer Editor Effects .................................................................. 113

Contents
Contents tables
Table 1. Maximum flexion angle for a normal knee joint in different stages ................. 17
Table 2. Flexion angle required by a normal knee joint for different daily activities ...... 18
Table 3. Rage of axial rotation depending on the flexion angle of a normal knee joint. 18
Table 4. Comparison of recovery times for PKR and TKR .......................................... 35
Table 5. Advantages and disadvantages of unicompartmental knee replacement ...... 43
Table 6. Mechanical Property Requirements of Co-Cr Alloys ...................................... 52
Table 7. Mechanical Properties of Ti and its Alloys (ASTM F136) ............................... 53
Table 8. Desired Properties of Implantable Bioceramics ............................................. 56
Table 9. Physical Property Requirements of Alumina and Zirconia ............................. 56
Table 10. Physical Properties of Calcium Phosphate .................................................. 58
Table 11. Requirements for Biomedical Polymers ....................................................... 59
Contents figures
Fig. 1. Comparison between normal and damaged cartilage......................................... 3
Fig. 2. Prevalence of osteoarthritis of the knee, by age group, sex, and region. ............ 4
Fig. 3. Knee anatomy .................................................................................................... 9
Fig. 4. Lower extremity of the femur ............................................................................ 10
Fig. 5. Upper extremity of the tibia .............................................................................. 11
Fig. 6. Medial and lateral meniscus ............................................................................. 12
Fig. 7. Patella. Anterior facet (A); Posterior facet (B). .................................................. 13
Fig. 8. Posterior ligaments .......................................................................................... 14
Fig. 9. Anterior ligaments ............................................................................................ 14
Fig. 10. Intracapsular ligaments .................................................................................. 15
Fig. 11. Posterior view of knee muscles ...................................................................... 15
Fig. 12. Popliteus muscle ............................................................................................ 15
Fig. 13. Knee arteries.................................................................................................. 16
Fig. 14. Knee nerves ................................................................................................... 16
Fig. 15. Normal flexion/extensin angle ...................................................................... 17
Fig. 16. How joint subluxation or dislocation would result were a convex surface to roll
on a fixed concave surface without gliding at the same time. ...................................... 20
Fig. 17. How joint surfaces would gap in some areas and impinge in others were a
concave surface to roll on a fixed convex surface without gliding at the same time. .... 20
Fig. 18. Comparison between a healthy knee and a knee with osteoarthritis. ............. 25
Fig. 20. Good candidate for total knee replacement: ................................................... 26
Fig. 19. Good candidate for partial knee replacement: ................................................ 26
Fig. 21. An artificial knee joint in place ........................................................................ 28
Fig. 22. A unicompartmental knee replacement .......................................................... 29
Fig. 23. A kneecap replacement ................................................................................. 30

Contents
Fig. 24. Complex replacement .................................................................................... 31

Fig. 25. Traditional approach ...................................................................................... 31
Fig. 26. Comparison between traditional and MIS ....................................................... 32
Fig. 27. Computed-assited surgery ............................................................................. 33
Fig. 28. Typical Knee implant ...................................................................................... 39
Fig. 29. Arthritis limited to the lateral compartment. .................................................... 43
Fig. 31. Posterior-substitution prosthesis .................................................................... 45
Fig. 30. Posterior-retention prosthesis ........................................................................ 45
Fig. 32. Mobile bearing knee prosthesis ...................................................................... 46
Fig. 33. Constrained condylar prosthesis .................................................................... 48
Fig. 34. High flexion prosthesis ................................................................................... 48
Fig. 35. Ultracongruent polyethylene bearing .............................................................. 49
Fig. 36. Home screen of 3D Slicer .............................................................................. 64
Fig. 38. Changing visualizing of imaging data. ............................................................ 65
Fig. 37. Drop down menu to select modules. .............................................................. 65
Fig. 39. Toolbar........................................................................................................... 66
Fig. 40. Load DICOM Data button ............................................................................... 66
Fig. 41. DICOM Browser ............................................................................................. 67
Fig. 42. Loading DICOM data ..................................................................................... 67
Fig. 43. DICOM data loaded ....................................................................................... 68
Fig. 45. Threshold Effect ............................................................................................. 69
Fig. 44. Editor module ................................................................................................. 69
Fig. 46. Selected threshold ......................................................................................... 70
Fig. 47. After apply PaintEffect.................................................................................... 70
Fig. 48. Modelmaker effect.......................................................................................... 71
Fig. 49. 3D model ....................................................................................................... 71
Fig. 50. Save model to 3D printing compatible format ................................................. 72

Contents

Chapter 1: Introduction
Introduction
1.1. Introduction
Osteoarthritis also known as degenerative arthritis, degenerative joint disease,
or osteoarthrosis, is a type of joint disease that results from breakdown of joint cartilage
(Fig.1) and underlying bone.
According to the World Health Organization (WHO),
Osteoarthritis is a degenerative joint disease, which mainly affects the articular

cartilage. It is associated with ageing and will most likely affect the joints that have been
continually stressed throughout the years including the knees, hips, fingers, and lower
spine region
As it is known, cartilage is the slippery

tissue that covers the ends of bones in a joint
and let them move smoothly. Healthy cartilage
allows bones to glide over each other and also
helps absorb shock of movement. In
osteoarthritis, the top layer of cartilage breaks
down and wears away. This allows bones under
the cartilage to rub together when you move.
The rubbing causes pain, swelling, and loss of
motion of the joint. Over time, the joint may lose
its normal shape. Also, bone spurs may grow on Fig. 1. Comparison between normal and damaged
cartilage
the edges of the joint. Bits of bone or cartilage
can break off and float inside the joint space,
which causes more pain and damage.
The most commonly involved joints are those near the ends of the fingers, at the
base of the thumb, neck, lower back, knees, and hips. Joints on one side of the body are
often more affected than those on the other. Usually the problems come on over years.
It can affect work and normal daily activities. Unlike other types of arthritis, only the joints
are typically affected.
Few data are available on the incidence of osteoarthritis because of the problems
of defining it and how to determine its onset. Estimates from Australia indicate that the
incidence of osteoarthritis is higher among women than men among all age groups. For
women, the highest incidence is among those aged 6574 years, reaching approximately
13.5 per 1000 population per year; for men, the highest incidence occurs among those
aged >75 years (approximately 9 per 1000 population per year).The prevalence of
osteoarthritis increases with age, because the condition is not reversible. Men are
affected more often than women among those aged <45 years, whereas women are
affected more frequently among those aged >55 years.
3
Introduction
Worldwide estimates are that 9.6% of men and 18.0% of women aged >60 years
have symptomatic osteoarthritis, and approximately 250 million people are affected by
osteoarthritis of the knee (3.6% of the population). Furthermore radiographic studies of
US and European populations aged >45 years show higher rates for osteoarthritis of the
knee: 14.1% for men and 22.8% for women. In the other hand, hip osteoarthritis is less
common, with a radiographic prevalence of 1.9% among men and 2.3% among women
aged >45 years.
By the year 2004, osteoarthritis globally causes moderate to severe disability in

43.4 million people. For this reason, knee osteoarthritis is likely to become the fourth
most important global cause of disability in women and eighth most important in men.
Talking about the money, it is the second-most costly condition billed to Medicare and
private insurance. Fig. 2 shows the prevalence of osteoarthritis of the knee by age group,
sex and region.
Fig. 2. Prevalence of osteoarthritis of the knee, by age group, sex, and region, 2000. WHO. A regions = developed
countries in North America, Western Europe, Japan, Australia, and New Zeland. AF = countries in sub-Saharan Africa.
AM BD = developing countries in the Americas. EM = countries in the Eastern Mediterranean and North African
regions. EM BC = developing countries in Europe. SEA = countries in South-east Asia. WP B = countries in the
Western Pacific region.
In general, osteoarthritis is more prevalent in Europe and the USA than in other
parts of the world. African-American women are more prone than white women to
osteoarthritis of the knee but not of the hip. Osteoarthritis of the hip occurs more often in
European whites than in Jamaican blacks, African blacks, or Chinese.
4
Introduction
1.2. Goals and project phases

The project objective is to generate and print a 3D model of a patients knee with
osteoarthritis. For this purpose the following tasks were carried out:
1. Analysis of knee anatomy and kinematics.
2. DICOM image selection.
3. Segmentation and generation of the 3D model.
4. Print 3D model.
5
Introduction
6
Chapter 2: Knee anatomy and kinematics
.
Knee anatomy and kinematics
2.1. Anatomy of the knee joint

The knee joint is the largest joint in the human body and also one of the most
important ones. It plays an essential role in movement related to carrying the body weight
in horizontal (running and walking) and vertical (jumps) directions.
Functionally it is classified as a ginglymus or hinge-joint and according to its

structure is classified as a synovial joint.
It consists of three articulations in one:
Two condyloid joints, one between

each condyle of the femur and the
corresponding meniscus and condyle
of the tibia.
One partly arthrodial, between the

patella and the femur.
Fig. 3. Knee anatomy
Also, the knee is composed of three functional compartments (Fig. 3):
Patellofemoral compartment, which consists of the patella, or "kneecap", and

the patellar groove on the front of the femur through which it slides.
Medial compartment (inner), which links the femur with the tibia medially.
Lateral compartment (outer), which links the femur with the tibia laterally.
2.1.1 Articular surfaces

2.1.1.1 Lower extremity of the femur
It has a cuboid form (Fig. 4). Its transverse diameter is greater than its antero-
posterior and it consists of two oblong eminences known as the condyles (lateral and
medial).
Condyles
Anteriorly, the condyles are slightly prominent and are separated by a smooth
shallow articular depression called the patellar surface. Posteriorly, they project
considerably and a deep notch, the intercondylar fossa of femur, is present between
them.
9
Fig. 4. Lower extremity of the femur
The lateral condyle is the more prominent and is the broader both in its antero-
posterior and transverse diameters.
The medial condyle is the longer and, when the femur is held with its body
perpendicular, projects to a lower level.
The condyles are not quite parallel with one another; the long axis of the lateral
is almost directly antero-posterior, but the medial runs backward and medialward.
Epicondyles
Each condyle is surmounted by an elevation, the epicondyle.
The medial epicondyle is a large convex eminence to which the tibial collateral
ligament of the knee-joint is attached. At its upper part is the adductor tubercle,
already referred to, and behind it is a rough impression which gives origin to the
medial head of the Gastrocnemius.
The lateral epicondyle, smaller and less prominent than the medial, gives
attachment to the fibular collateral ligament of the knee-joint. Directly below it is
a small depression from which a smooth well-marked groove curves obliquely
upward and backward to the posterior extremity of the condyle.
Articular surface
The articular surface of the lower end of the femur occupies the anterior, inferior,
and posterior surfaces of the condyles.
Its front part is named the patellar surface and articulates with the patella; it
presents a median groove which extends downward to the intercondyloid fossa and two
convexities, the lateral of which is broader, more prominent, and extends farther upward
than the medial.
The lower and posterior parts of the articular surface constitute the tibial
surfaces for articulation with the corresponding condyles of the tibia and menisci. These
surfaces are separated from one another by the intercondyloid fossa and from the
patellar surface by faint grooves which extend obliquely across the condyles.
10
2.1.1.2 Upper extremity of the tibia
It is large, and expanded into two eminences known as the condyles, a lateral
and a medial one (Fig. 5).
Facets
The superior articular surface presents two smooth articular facets.
The medial facet, oval in shape, is slightly concave from side to side, and from
before backward.
The lateral facet, nearly circular, is concave from side to side, but slightly convex
from before backward, especially at its posterior part, where it is prolonged on to
the posterior surface for a short distance.
The central portions of these facets articulate with the condyles of the femur,
while their peripheral portions support the menisci of the knee joint, which here intervene
between the two bones.
Fig. 5. Upper extremity of the tibia
Intercondyloid eminence
Between the articular facets, but nearer the posterior than the anterior aspect of
the bone, is the intercondyloid eminence (spine of tibia), surmounted on either side by a
prominent tubercle, on to the sides of which the articular facets are prolonged; in front of
and behind the intercondyloid eminence are rough depressions for the attachment of the
anterior and posterior cruciate ligaments and the menisci.
11
2.1.1.3 Menisci
The menisci or meniscus are two crescent-shaped fibrocartilaginous structures

(Fig. 6), which serve to deepen the surfaces of the head of the tibia for articulation with
the condyles of the femur. Also the menisci serve to protect the ends of the bones from
rubbing on each other, and play a role in shock absorption, and may be cracked, or torn,
when the knee is forcefully rotated and/or bent.
The peripheral border of each meniscus is thick, convex, and attached to the
inside of the capsule of the joint; the opposite border is thin, concave, and free. The
upper surfaces of the menisci are concave, and in contact with the condyles of the femur;
their lower surfaces are flat, and rest upon the head of the tibia; both surfaces are
smooth, and invested by synovial membrane.
Each meniscus covers approximately the peripheral two-thirds of the

corresponding articular surface of the tibia.
Fig. 6. Medial and lateral meniscus
The medial meniscus is nearly semicircular in form, a little elongated from before
backward, and broader behind than in front; its anterior end, thin and pointed, is
attached to the anterior intercondyloid fossa of the tibia, in front of the anterior
cruciate ligament; its posterior end is fixed to the posterior intercondyloid fossa
of the tibia, between the attachments of the lateral meniscus and the posterior
cruciate ligament.
The lateral meniscus is nearly circular and covers a larger portion of the articular
surface than the medial one. It is grooved laterally for the tendon of the Popliteus,
which separates it from the fibular collateral ligament. Its anterior end is attached
in front of the intercondyloid eminence of the tibia, lateral to, and behind, the
anterior cruciate ligament, with which it blends; the posterior end is attached
behind the intercondyloid eminence of the tibia and in front of the posterior end
of the medial meniscus.
12
2.1.1.4. Patella
The patella also known as the kneecap, is a thick, circular-triangular bone which
articulates with the femur and covers and protects the anterior articular surface of the
knee joint.
The patella is the largest sesamoid bone in the body. It has an apex facing
downwards, pointed in shape, and gives attachment to the patellar ligament. The front
and back surfaces are joined by a thin margin and towards centre by a thicker margin.
The tendon of the quadriceps femoris muscle attaches to the base of the patella,
with the vastus intermedius muscle attaching to the base itself, and the vastus lateralis
and vastus medialis are attached to outer lateral and medial borders of patella
respectively.
The front surface (Fig. 7A) is divided into three parts:
The upper third is coarse, flattened, and rough, and serves for the attachment of
the tendon of the quadriceps.
The middle third has numerous vascular canaliculi.
The lower third culminates in the apex which serves as the origin of the patellar
ligament.
The posterior surface (Fig. 7B) is divided into two parts.
The upper three-quarters articulates with the femur and is subdivided into a
medial and a lateral facet by a vertical edge which varies in shape.
The lower part has vascular canaliculi filled and is filled by fatty tissue, the
infrapatellar fat pad.
In the adult the articular surface is about 12 cm2 and covered by cartilage, which
can reach a maximal thickness of 6 mm in the center at about 30 years of age.
Fig. 7. Patella. Anterior facet (A); Posterior facet (B).
13
2.1.2. Articular ligaments

The bones involved are connected together by ligaments. This ligaments are
commonly grouped into two types, intracapsular and extracapsular ones and classified
as follows:
Intracapsular ligaments (Fig.10)

o Cruciate ligaments
Anterior Cruciate Ligament (ACL)
Posterior Cruciate Ligament (PCL)
o Transverse ligament
o Meniscofemoral ligaments
Anterior meniscofemoral ligament or

ligament of Humphrey
Posterior meniscofemoral ligament or

ligament of Wrisberg
Fig. 8. Posterior ligaments
Extracapsular ligaments
o Anterior ligaments (Fig. 9)
Patellar ligament
o Posterior ligaments (Fig. 8)
Oblique Popliteal ligament
Arcuate Popliteal ligament
o Medial ligaments
Medial Collateral ligament
o Lateral ligaments
Lateral Collateral ligament
Fig. 9. Anterior ligaments
14
Fig. 10. Intracapsular ligaments
2.1.3. Structures around the joint
2.3.1. Muscles
In front, and at the sides
Quadriceps femoris
Laterally
Biceps femoris (Fig. 11)
Popliteus (Fig. 12)
Medially (Fig. 11)
Sartorius Fig. 11. Posterior view of knee muscles
Gracilis
Semitendinosus
Semimembranosus
Behind
Popliteus (Fig. 12)
Plantaris (Fig. 11)
Gastrocnemius (Fig. 11)
Fig. 12. Popliteus muscle
15
2.1.3.2. Arteries
The knee arteries (Fig. 13) supplying

the joint are:
The descending genicular
The genicular branches of the

popliteal
The recurrent branches of the

anterior tibial
The descending branch from the

lateral femoral circumflex of the
profunda femoris.
Fig. 13. Knee arteries
2.1.3.3. Nerves
The knee nerves (Fig. 14) are:
The tibial
The common peroneal.
The femoral
The sural
Fig. 14. Knee nerves

2.1.3.4. Cartilage
Cartilage is a thin, elastic tissue that protects the bone and makes certain that
the joint surfaces can slide easily over each other. Cartilage ensures supple knee
movement.
There are two types of joint cartilage in the knees:
Fibrous cartilage, which has tensile strength and can resist pressure (menisci).
Hyaline cartilage, which covers the surface along which the joints move.
Cartilage will wear over the years and has a very limited capacity for self-
restoration. The newly formed tissue will generally consist of a large part of fibrous
cartilage of lesser quality than the original hyaline cartilage. As a result, new cracks and
tears will form in the cartilage over time.
16
2.2. Kinematics of the knee

Kinematics is the branch of classical mechanics which describes the motion of
points, bodies and systems of bodies without consideration of the causes of motion. It is
divided into osteokinematics and arthrokinematics.
Osteokinematic studies the bone motion without consideration of the articular

surfaces, and describes the axis and planes of movement.
Arthrokinematic studies the intrinsic motion of the joint, especially the

relationship between articular planes during motion.
The main movements of the knee joint are flexion/extension and less the
internal and external rotation. Due to articular incongruences and variations of
ligament elasticity, the knee undergoes anterior or posterior gliding movements of the
tibia or femur.
2.2.1. Osteokinematics of the knee
2.2.1.1. Flexion/Extension
Flexion and extension

movement take place in the sagittal
plane with a horizontal axis across
femoral condyles. This axis present a
slight obliquity, lower at the medial face
of the articulation, which cause the tibia
to move laterally during extension and
Fig. 15. Normal flexion/extensin angle
medially during maximum flexion.
The knee present biarticular muscles which generate a variation of the range of
motion of the knee. If the hip is in certain positions, some muscles exert forces over the
knee joint and produce a limitation in its normal range of motion (Fig. 15 and Table 1).
Table 1. Maximum flexion angle for a normal knee joint in different stages.
Flexion angle
(max)
Normal 130 to 140
Hyperextension 120
of the hip
Maximum 160
flexion of the hip
17
Also, for different daily live activities exist a required range of motion that your
knee joint has to be capable to achieve in order to do the activity (Table 2).
Table 2. Flexion angle required by a normal knee joint for different daily activities
Activity Flexion
angle required
Knee extension 0
Walking 60
Climbing stairs 80
Descending stairs 90
Sitting down 80-110
Sedentary position 90
Functional activity 5-10
2.2.1.2. Axial rotation
The axial rotation takes place in the longitudinal plane and could be internal
(medial) or external (lateral). This knee motion is produced by the movement of the tibia
as a consequence of articular incongruences and the ligamentous laxity of the
articulation. The range of rotation depends on the flexion/extension angle of the knee
(Table 3).
Table 3. Rage of axial rotation depending on the flexion angle of a normal knee joint
Flexion angle Rotation
0 No rotation
60 to 70 Starts to rotate
90 External rotation 40
Internal rotation 30
18
2.2.2. Arthrokinematics of the knee

Articular surfaces move with respect to one another by simultaneously rolling,
gliding, and spinning.
Roll: the rolling of one joint surface on another. During this type of movement
new points on each surface come into contact throughout the motion.
Glide: linear movement of a joint surface parallel to the plane of the adjacent
joint surface. In other words, one point on one joint surface (remains the same)
contacts new points on the other joint surface.
Spin: the rotation of the movable joint surface on the fixed adjacent surface.
Essentially, the same point on each surface remains in contact with each other.
2.2.2.1. Rolling and gliding
If the moving joint surface rolls on its partner without simultaneously gliding, the
surfaces would separate (gap or subluxate) in some places and impinge in others.
In order to preserve joint integrity, roll and glide must occur simultaneously, and
must occur in opposite directions. For example, the convex joint surface must roll
posteriorly (curved arrow) and glide anteriorly (straight arrow) (Fig. 16). Also to preserve
joint integrity, roll and glide must occur simultaneously, and in the same direction. For
example, the concave joint surface must roll posteriorly (curved arrow) and glide
posteriorly (straight arrow) (Fig. 17).
We can predict and understand the relationship between bony shapes at a joint
surface and the surfaces' movements by applying the rules of concavity and convexity.
Each joint or articulation involves two bony surfaces, one that is convex and one
that is concave.
1) When the concave surface is fixed and the convex surface moves on it, the
convex surface rolls and glides in opposite directions.
2) When the convex surface is fixed and the concave surface moves on it, the
concave surface rolls and glides in the same direction.
19
Fig. 16. How joint subluxation or dislocation would result were a convex surface to roll on a fixed concave surface
without gliding at the same time.
Fig. 17. How joint surfaces would gap in some areas and impinge in others were a concave surface to roll on a fixed
convex surface without gliding at the same time.
20
Chapter 3: Osteoarthritis
Osteoarthritis
3.1. Physiopathology of osteoarthritis

While osteoarthritis is a degenerative joint disease that may cause gross cartilage
loss and morphological damage to other joint tissues, more subtle biochemical changes
occur in the earliest stages of osteoarthritis progression:
The water content of healthy cartilage is finely balanced by compressive force

driving water out & swelling pressure drawing water in.
Collagen fibers exert the compressive force, whereas the GibbsDonnan

effect & cartilage proteoglycans create osmotic pressure which tends to draw
water in.
However, during onset of osteoarthritis, the collagen matrix becomes more

disorganized and there is a decrease in proteoglycan content within cartilage. The
breakdown of collagen fibers results in a net increase in water content. This increase
occurs because whilst there is an overall loss of proteoglycans (and thus a decreased
osmotic pull), it is outweighed by a loss of collagen. Without the protective effects of the
proteoglycans, the collagen fibers of the cartilage can become susceptible to
degradation and thus exacerbate the degeneration. Inflammation of the synovium (joint
cavity lining) and the surrounding joint capsule can also occur, though often mild
(compared to what occurs in rheumatoid arthritis). This can happen as breakdown
products from the cartilage are released into the synovial space, and the cells lining the
joint attempt to remove them.
Other structures within the joint can also be affected:
Ligaments within the joint become thickened and fibrotic and the
Menisci can become damaged and wear away.
Menisci can be completely absent by the time a person undergoes a joint

replacement. New bone outgrowths, called "spurs" or osteophytes, can form on the
margins of the joints, possibly in an attempt to improve the congruence of the articular
cartilage surfaces in the absence of the menisci. The subchondral bone volume
increases and becomes less mineralized (hypomineralization). All these changes can
cause problems functioning. The pain in an osteoarthritic joint has been related to
thickened synovium and subchondral bone lesions.
23
Osteoarthritis
3.1.1. Symptoms
Initially, symptoms may occur only following exercise, but over time may become
constant. The most common symptoms are:
Joint pain and stiffness.
Joint swelling.
Decreased range of motion.
Weakness or numbness of the arms and legs (when the back is affected).
3.1.2. Causes
Causes of osteoarthritis may include previous joint injury, abnormal joint or limb
development, and inherited factors. Osteoarthritis is believed to be caused by
mechanical stress on the joint and low grade inflammatory processes.
Some of the factors that contribute to the development of osteoarthritis are:
Genes: Various genetic traits can make a person more likely to develop
osteoarthritis. One possibility is a rare defect in the bodys production of collagen,
the protein that makes up cartilage. This abnormality can cause osteoarthritis to
occur as early as age 20. Other inherited traits may result in slight defects in the
way the bones fit together so that cartilage wears away faster than usual.
Weight: Being overweight puts additional pressure on hips and knees. Many
years of carrying extra pounds can cause the cartilage that cushions joints to
break down faster. Research has shown there is a link between being overweight
and having an increased risk of osteoarthritis.
Injury and overuse: Repetitive movements or injuries to joints (such as a

fracture, surgery or ligament tears) can lead to osteoarthritis. Some athletes, for
example, repeatedly damage joints, tendons and ligaments, which can speed
cartilage breakdown. Certain careers that require standing for long periods of
time, repetitive bending, heavy lifting or other movements can also make
cartilage wear away more quickly.
Others: Several other factors may contribute to osteoarthritis. These factors

include bone and joint disorders like rheumatoid arthritis, certain metabolic
disorders such as hemochromatosis, which causes the body to absorb too much
iron, or acromegaly, which causes the body to make too much growth hormone.
24
Osteoarthritis
3.2. Diagnosis and treatment

Diagnosis is based on the orthopedic examination, which consists of:
Medical history: Your orthopedic surgeon will gather information about your
general health and ask you about the extent of your knee pain with specifically
concerning of the location of your pain and your ability to function. If your pain is
located almost entirely on either the inside portion or outside portion of your
knee, then you may be a candidate for a partial knee replacement. If you have
pain throughout your entire knee or pain in the front of your knee (under your
kneecap) you may be better qualified for a total knee replacement.
Physical examination: Your doctor will closely examine your knee trying to
determine the exact location of your pain. Your doctor will test your knee for
range of motion and ligament quality. If your knee is too stiff, or if the ligaments
in your knee feel weak or torn, then your doctor will probably not recommend
partial knee replacement.
Image scans: Your doctor will order several x-rays of your knee to see the
pattern of arthritis. In practice, an X-ray scan may confirm the diagnosis (Fig.
18), but some surgeons may also order a magnetic resonance imaging (MRI)
scan to better evaluate the cartilage. These images help to determine the extent
of damage and deformity in your knee. In the diagnostic process is expected to
see some typical changes on X-ray scans, for example: joint space narrowing,
subchondral sclerosis (increased bone formation around the joint),
subchondral cyst formation, and osteophytes.
Your orthopedic surgeon

will review the results of your
evaluation with you and discuss
whether total knee replacement is
the best method to relieve your
pain and improve your function.
In addition, your orthopedic

surgeon will explain the potential
risks and complications of total
knee replacement, including those
related to the surgery itself and
Fig. 18. Comparison between a healthy knee (left) and a knee those that can occur over time
with osteoarthritis (right).
after your surgery.
As of today, there is no definitive cure for osteoarthritis, but treatments are

available to manage symptoms. Long-term management of the disease will include
several factors:
25
Osteoarthritis
Managing symptoms, such as pain, stiffness and swelling

Improving joint mobility and flexibility
Maintaining a healthy weight
Getting enough of exercise
Depending on how severe is the clinical case of the patient, doctors would chose
the correct treatment. For example, the x-ray of fig. 19 shows a good candidate for a
PKR; it shows on the left a normal joint space for the lateral compartment, but the medial
compartment has severe arthritis with "bone-on-bone" degeneration, and the same knee
after partial knee replacement on the right image. On the other hand, the x-ray of fig. 20
shows a good candidate for a TKR; it shows on the right image a severe arthritis in both
medial and lateral compartments, and this patient after total knee replacement on the
right image.
Fig. 19. Good candidate for partial knee replacement: Fig. 20. Good candidate for total knee replacement:
The most common treatment includes exercise, efforts to decrease joint

stress, support groups, and pain medications. In some cases, if pain interferes with
normal life despite other treatments, joint replacement surgery may help. The most
common types of treatments are briefly described in the following subsections.
3.2.1. Physical Activity
One of the most beneficial ways to manage osteoarthritis is to get moving.

Studies show that simple activities like walking around the neighborhood or taking a fun,
easy exercise class can reduce pain and help maintain (or attain) a healthy weight.
Strengthening exercises build muscles around osteoarthritis-affected joints,

easing the burden on those joints and reducing pain. Range-of-motion exercise helps
maintain and improve joint flexibility and reduce stiffness. Aerobic exercise helps to
improve stamina and energy levels and also help to reduce excess weight. Slow, gentle
stretching of joints may improve flexibility, lessen stiffness and reduce pain. Exercises
such as yoga and tai chi are great ways to manage stiffness.
26
Osteoarthritis
3.2.2. Weight Management
Excess weight adds additional stress to weight-bearing joints, such as the hips,
knees, feet and back. Losing weight can help people with osteoarthritis reduce pain and
limit further joint damage. The basic rule for losing weight is to eat fewer calories and
increase physical activity.
3.2.3. Pain and Anti-inflammatory Medications
Medicines for osteoarthritis are available as pills, syrups, creams or lotions, or

they are injected into a joint. They include:
Analgesics. These are pain relievers and include acetaminophen, opioids

(narcotics) and an atypical opioid called tramadol.
Nonsteroidal anti-inflammatory drugs (NSAIDs). These are the most
commonly used drugs to ease inflammation and related pain. NSAIDs include
aspirin, ibuprofen, naproxen and celecoxib.
Corticosteroids. Corticosteroids are powerful anti-inflammatory medicines.
They are taken by mouth or injected directly into a joint.
Hyaluronic acid. Hyaluronic acid occurs naturally in joint fluid, acting as a shock
absorber and lubricant. However, the acid appears to break down in people with
osteoarthritis.
3.2.4. Assistive Devices
Assistive devices can help with function and mobility. These include items, such
as like scooters, canes, walkers, splints, shoe orthotics or helpful tools, such as jar
openers, long-handled shoe horns, steering wheel grips or exoskeletons.
3.2.5. Surgery
If problems are significant and more conservative management is ineffective, joint

replacement surgery or resurfacing may be recommended. Joint surgery can repair or
replace severely damaged joints, especially hips or knees. An artificial joint, however,
only lasts a limited amount of time.
As an example, knee replacement, or knee arthroplasty, is a surgical procedure

to replace the weight-bearing surfaces of the knee joint to relieve pain and disability and
is most commonly performed for osteoarthritis. Knee replacement surgery can be
performed as a partial or a total knee replacement. In general, the surgery consists of
replacing the diseased or damaged joint surfaces of the knee with metal and plastic
components shaped to allow continued motion of the knee.
27
Osteoarthritis
3.3. Types of surgeries

A knee replacement (also called knee arthroplasty) might be more accurately
termed a knee "resurfacing" because only the surface of the bones are actually replaced.
There are four basic steps to a knee replacement procedure:
Prepare the bone: The damaged cartilage surfaces at the ends of the femur
and tibia are removed along with a small amount of underlying bone.
Position the metal implants: The removed cartilage and bone is replaced with
metal components that recreate the surface of the joint. These metal parts may
be cemented or "press-fit" into the bone.
Resurface the patella: The undersurface of the patella (kneecap) is cut and
resurfaced with a plastic button. Some surgeons do not resurface the patella,
depending upon the case.
3.3.1. Total knee arthroplasty or total knee replacement

Most total knee replacement surgeries involve replacing the joint surfaces at:
1) Lower extremity of the femur.
2) Upper extremity of the tibia.
A TKR may also involve replacing the

under-surface of your kneecap (patella) with a
smooth plastic dome, but some surgeons
prefer to preserve the natural one if possible.
If the patient has had a previous

operation to remove the patella altogether
(patellectomy), this wont stop the patient
having a knee replacement, but it may affect the
type of replacement part (prosthesis) the
surgeon uses.
The new parts are normally cemented in

place (Fig. 21). If cement is not used then the
surface of the component facing the bone is
textured or coated to encourage bone to grow
onto it, forming a natural bond.
Another common technique is to use a

mobile plastic bearing which isnt firmly fixed to
Fig. 21. An artificial knee joint in place
the metal parts. This may help to reduce wear
on your new joint, though it isnt has been
shown to provide better long-term results.
28
Osteoarthritis
3.3.2. Unicompartmental knee arthroplasty or partial knee

replacement
If arthritis affects only one side of your knee usually the inner side it may be
possible to have a half-knee replacement (sometimes called unicompartmental or partial
replacement) (Fig.22).
This type of surgery involves less

interference with the knee than a TKR,
and usually rely on a quicker recovery
and better function. Partial knee
replacements can be carried out through
a smaller cut (incision) than a total knee
replacement, using techniques called
reduced invasive or minimally invasive
surgery. A smaller incision may further
reduce the recovery time.
Partial knee replacement isnt

suitable for everyone because you need
to have strong, healthy ligaments within
your knee.
Fig. 22. A unicompartmental knee replacement
Research shows people who have PKR are more likely to have the knee revised
than people who have a TKR about 1 person in 10 needs further surgery after 10 years.
Even though the operation involves less interference with the knee it is often a
more complex operation than total knee replacement.
PKR can be considered at any age:
For younger people, it offers the opportunity to preserve more bone, which is
helpful if you need revision surgery at a later stage.
For older people, it is a less stressful operation with less pain and less risk of
bleeding.
The outcome of the surgery, however, depends on the type of arthritis, rather
than your age.
29
Osteoarthritis
3.3.3. Patellofemoral joint arthroplasty
This type of surgery consist on replace just the under-surface of the kneecap and
its groove (the trochlea) if these are the only parts affected by arthritis (Fig. 23).
The surgery has a higher rate of failure than TKR, may be caused by the arthritis
progressing to other parts of your knee. The operation is only suitable for about 1 in 40
people with osteoarthritis. However, if the arthritis doesnt progress the outcome of
patellofemoral joint replacement can be good. Additionally its a less major operation and
has a faster recovery time.
Some surgeons advise a total knee replacement because of its more predictable
results, but others feel that its better to preserve the rest of the knee joint if it isnt affected
by arthritis.
Fig. 23. A kneecap replacement
3.3.4. Complex replacement or revision
Some people may need a more complex type of knee replacement. The usual
reasons for this are:
Major bone loss due to arthritis or fracture

Major deformity of the knee
Weakness of the main knee ligaments
If youre having a second or third joint replacement in the same knee.
30
Osteoarthritis
These knee replacements usually have a longer stem, which allows the
component to be more securely fixed into the bone cavity (Fig. 24). The components
may also interlock in the center of the knee to form a hinge to give greater stability. Extra
pieces of metal and/or plastic may be used to make up for any removed or badly
damaged bone.
Fig. 24. Complex replacement
3.4. Surgical approaches
There are different surgical approaches (as well as the approach to anesthesia,
whether general or regional) thats best suited to each patient. In order to ensure a
smooth procedure, an orthopedic surgeon will map out the patients knee anatomy in
advance so the surgeon may plan the surgical approach and anticipate special
instruments or devices.
The three most used surgical approaches today are:
Tradition surgery
Minimally Invasive Surgery (MIS)
Computer-Assisted Surgery (CAS)
3.4.1. Traditional Surgery
In the traditional approach, the surgeon

makes an 8 to 12 inch incision (Fig. 25) and
operates on the knee using standard surgical
technique. Generally, the incision is made along
the front and toward the middle (midline or
Fig. 25. Traditional approach
anteromedial) or along the front and to the side
(anterolateral) of the knee.
31
Osteoarthritis
The traditional surgical approach usually involves cutting into the quadriceps
tendon in order to turn the kneecap over and expose the arthritic joint.
This approach typically requires three to five recovery days in the hospital and
about 12 weeks of recovery time.
3.4.2. Minimally Invasive Surgery
A minimally invasive approach reduces

the incision to 3 to 4 inches (Fig. 26). The
kneecap is pushed to the side rather than being
turned over and this results in a smaller cut into
the quadriceps tendon and less trauma to the
quadriceps muscle. Because the surgeon cuts
less muscle, healing occurs faster and you are
Fig. 26. Comparison between traditional and MIS
likely to experience better range of motion after
recovery.
A minimally invasive procedure reduces tissue trauma, pain, and decreases

blood loss, which drive to a consequently faster recovery.
This procedure modifies the techniques used in traditional surgery while using
the same implants from traditional surgery. Manufacturers provide specialized
instruments that help to place the implant accurately but also allow for incisions to be
made as small as possible.
Since the only change between MIS and traditional surgery is in the surgical
technique, the long-term clinical outcomes are similar.
There are two main types of minimally invasive approaches:
Quadriceps-Sparing Approach
After making a minimal incision, the surgeon shifts the kneecap to the side
and cuts away the arthritic bone without cutting through the quadriceps tendon.
This method is less invasive than traditional surgery because it spares the
quadriceps muscle from as much trauma as possible.
This approach take longer to perform but may result in a faster rehab
process because there is little to no trauma to the underlying thigh muscle,
making it easier to walk sooner after the operation.
32
Osteoarthritis
Lateral Approach
The surgeon enters the knee joint laterally, or from the side of the knee.
The lateral approach is less invasive than traditional surgery because it spares
much of the quadriceps, making it easier for patients to return to walking faster.
This approach is rarely used. It is more common for patients whose knees
tend to bend outwards.
3.4.3. Computer-Assisted Surgery (CAS)
Increasingly, surgeons are also turning to computer-assisted methods for both

TKRs and PKRs involving both traditional and minimally invasive procedures.
A surgeon enters a patients

anatomical data into a computera
process called registrationand the
computer generates a 3-D model of the
knee. The software provides the surgeon
with a more precise, computer-
aided image of the knee. The computer
helps the surgeon (Fig. 27) align the knee
components more precisely in the bone
and increases the odds that the device
Fig. 27. Computed-assited surgery
will work effectively.
Approximately 10% of patients receiving a traditional TKR experience problems

due to improper alignment. A computer-based approach also allows a surgeon to operate
with a smaller incision and benefits the patient by reducing recovery time. A more precise
fit can also reduce wear and increase the longevity of the new joint.
3.4.4. Risks
Risks and complications in knee replacement are similar to those associated with
all joint replacements. The most serious complication are:
Infection of the joint (less than 1% of patients)

Deep vein thrombosis (up to 15% of patients)
Nerve injuries (12% of patients)
Persistent pain or stiffness (823% of patients)
Prosthesis failure (2% of patients at 5 years)
33
Osteoarthritis
There is increased risk in complications for obese people going through total
knee replacement. The morbidly obese should be advised to lose weight before surgery
and, if medically eligible, would probably benefit from bariatric surgery.
Fracturing or chipping of the polyethylene platform inserted onto of the tibial

component may be a concern. These fragments may become lodged in the knee and
create pain or may move into other parts of the body.
3.4.4.1. Deep vein thrombosis
According to the American Academy of Orthopedic Surgeons (AAOS), deep vein

thrombosis in the leg is "the most common complication of knee replacement surgery...
prevention... may include periodic elevation of patient's legs, lower leg exercises to
increase circulation, support stockings and medication to thin your blood."
3.4.4.2. Loss of Motion
The knee at times may not recover its normal range of motion (0135 degrees
usually) after total knee replacement. Much of this is dependent on pre-operative
function. Most patients can achieve 0110 degrees, but stiffness of the joint can occur.
In some situations, manipulation of the knee under anesthetic is used to reduce post-
operative stiffness. There are also many implants from manufacturers that are designed
to be "high-flex" knees, offering a greater range of motion.
3.4.4.3. Instability
In some patients, the kneecap dislocates to the outer side of the knee. This is
painful and usually needs to be treated by surgery to realign the kneecap. However this
is quite rare. Knee replacement implants can now last up to 20 years.
3.4.4.4. Infection
The current classification of AAOS divides prosthetic infections into four types.
Type 1 (positive intraoperative culture): Two positive intraoperative cultures

Type 2 (early postoperative infection): Infection occurring within first month after
surgery
Type 3 (acute hematogenous infection): Hematogenous seeding of site of
previously well-functioning prosthesis
Type 4 (late chronic infection): Chronic indolent clinical course; infection present
for more than a month
34
Osteoarthritis
While it is relatively rare, periprosthetic infection remains one of the most

challenging complications of joint arthroplasty. A detailed clinical history and physical
remain the most reliable tool to recognize a potential periprosthetic infection. In some
cases the classic signs of fever, chills, painful joint, and a draining sinus may be present,
and diagnostic studies are simply done to confirm the diagnosis.
3.4.5. Recovery Time
The recovery time for any operation varies. Here is a comparison of typical
recovery times for partial vs. total knee replacement (Table 4).
Table 4. Comparison of recovery times for total knee replacement and partial knee replacement
Total Knee Partial Knee

Replacement Replacement
Hospital Stay 3-4 days 1-2 days
In Patient Rehab Stay 1-2 weeks None
Outpatient Rehab 12 weeks 6 weeks
Time using walker or crutches 6 weeks 2 weeks
Need for pain meds 6 weeks 2 weeks
Loss of work 6-12 weeks 2-3 weeks
Complete Recovery 6-12 months 6-12 weeks
35
Osteoarthritis
36
Chapter 4: Knee prostheses
Knee prostheses
4.1. Knee replacement implants

There are several kinds of replacement knee joint as well as different surgical
methods. Taking into account the condition of your knee and your general health, doctors
and orthopedic surgeons should choose the best option for each patient. The brand and
design used depends on many factors, including your needs (based on your age, weight,
activity level, and health), doctor's experience and familiarity with the device, and the
cost and performance record of the implant.
4.1.1. Implant Components

Three bone surfaces may be replaced in a total knee replacement and compound
a typical knee implant (Fig. 28):
The lower ends of the femur: The metal

femoral component curves around the end
of the femur. It is grooved so the kneecap
can move up and down smoothly against
the bone as the knee bends and
straightens.
The top surface of the tibia: The tibial
component is typically a flat metal platform
with a cushion of strong, durable plastic,
called polyethylene. Some designs do not
have the metal portion and attach the
polyethylene directly to the bone. For
Fig. 28. Typical Knee implant
additional stability, the metal portion of the
component may have a stem that inserts
into the center of the tibia bone.
The back surface of the patella: The patellar component is a dome-shaped
piece of polyethylene that duplicates the shape of the patella.
4.1.2. History of total knee arthroplasty

Many types of prosthesis are used for total knee arthroplasty, and the evolution
of knee arthroplasty which has a history of almost 40 years, involves repetitious cycles
of failure and development.
During its early stage (1970-1974), instruments of the unicondylar,

duocondylar, or hinged types were used, but these were eventually abandoned due to
low success rates. A replacement for the total condylar type was successfully developed
and became the model for total knee arthroplasty. Recently, unicondylar arthroplasty has
39
Knee prostheses
produced good results in selected patients, and arthroplasty of the constrained or hinged
type have been proven useful for revision surgery or combined surgery, respectively. To
solve the problem of the fixed bearing joint, a mobile bearing joint has been developed,
and non-cemented fixed knee arthroplasty is receiving renewed attention. Much effort is
being expended on the developments of new materials, such as, ceramics and cross-
linked polyethylene, on new designs that maximize function and endurance, and on
minimally invasive surgery.
4.1.2.1. Beginning
During the late 19th and early 20th century, interposition arthroplasty was
attempted using soft tissues. In 1860, Verneuil proposed interposition arthroplasty,
involving the insertion of soft tissue to reconstruct the joint surface. Since then, pig
bladder, nylon, femoral sheath, anterior bursa of the knee, cellophane, and many other
materials have been used, but results have been disappointing.
The use of metallic interposition arthroplasty began in the late 1930s. Campbell
and Smith-Peterson proposed metal femoral mold arthroplasty, and McKeever and Mac
Intosh proposed hemiarthroplasty of the tibia, but all produced unsatisfactory results in
terms of minimizing pain, and high rate of failures of the interposition, and thus, these
procedures were not widely recognized.
Ferguson attempted resection arthroplasty for ankylosis or severe deformity

caused by tuberculosis or infection. This procedure involved resecting cartilage from the
knee joint and allowing knee joint movement along the subchondral surface. When too
little bone was removed, knees spontaneously fused, and when more bone was
removed, knee had good motion but poor stability. Accordingly, these operations were
attempted in only the most severe cases because their results were poor.
4.1.2.2. The hinged prosthesis
In the 1950s, Walldius developed a hinged prosthesis that replaced the joint
surfaces of the femur and tibia, as subsequently, modifications of the basic hinged
prosthesis design were made by many surgeons.
The hinged prosthesis allows the intramedullary stem to align with the artificial
knee joint by itself, and is technically easy to perform since all ligaments and soft tissues
can be removed due to the mechanical and structural stability of the prosthesis.
During the 1950s and 1960s, hinged total knee arthroplasty provided satisfactory
results for a longer period of time in more patients than any other knee arthroplasty
design used. However, this method could not be widely used since this type of simple
hinged prosthesis cannot replace the complex movements of the knee joint and
40
Knee prostheses
because of a high failure rate due to early loosening caused by overloading the
prosthesis and bone contact surface or by infection.
4.1.2.3. The bicompartmental prosthesis
In 1971, Gunston developed polycentric knee arthroplasty. Gunstons knee

arthroplasty retained the collateral and cruciate ligaments to help absorb stress, and
consisted of relatively flat tibial interposition of high-density polyethylene and a round
femoral prosthesis, which replaced the posterior portion of femoral condyles. These
components were fixed to bones with bone cement, and replaced the complex
movements of femoral roll-back. Polycentric knee arthroplasty was initially successful
due to improved mobility and movement range, but the fixation it provided was not
sufficient.
Freeman designed a femoral and tibial prosthesis to work in a roller-in-trough

manner by the strength of collateral ligament. The anterior and posterior cruciate
ligaments were usually removed, and the tibial prosthesis did not have an
intramedullary stem to minimize the risk of infection and to maximize knee joint function
for salvage procedures. However, the loosening of the tibial prosthesis became a
major drawback.
In the mid 1970s, duocondylar interposition was designed to resemble the

anatomic structure of the knee joint. The femoral prosthesis connected with two
unicondylar prostheses via an anterior bridge, and formed a joint, which was
considerable wider than previous polycentric knee arthroplasties, with two flat tibial
instruments. However, the design had the drawbacks of frequent destruction or
deformity of the tibial prosthesis.
4.1.2.4. The tricompartmental prosthesis
In the early 1970s, three types of condylar prostheses were developed:
1) Ranawart developed the duocondylar prosthesis
a. Preserved the anterior and posterior cruciate ligaments
b. Used bone cement for fixation to bone
2) Coventry developed the geometric prosthesis
3) Townley developed the anatomic prosthesis
However, the geometric and anatomic types were not produced continuously due
to early loosening of fixation. The duocondylar type was further developed to
produce the first total condylar prosthesis with a tibial stem by Walker in 1976 and
41
Knee prostheses
became the early model for todays posterior cruciate ligament substitution knee
arthroplasty.
The total condylar prosthesis is a design that:
Removes anterior and posterior cruciate ligaments
The femoral prosthesis Is made of chrome cobalt and has:
o A symmetric femur with a double curve
o A flat patellar trochlear groove
The tibial prosthesis Is completely made of polyethylene and has:
o Good conformity in the flexion and extension states
o Anterior and posterior lips in the tibial joint surface
o Eminence in the mid joint surface which provides anteroposterior and

mediolateral stability
o A stem which can endure asymmetric loading
The patellar prosthesis is of the half-ball type and is completely made of

polyethylene, with a fixation lug in the middle, which is fixed with bone cement.
The features of this early total condylar knee are used in most of todays
prostheses According to early clinical reports, the average mobile angle was 90-100
degrees. To solve this problem, Insall added a cam to the femoral prosthesis and a
post to the tibial prosthesis for posterior cruciate ligament substitution knee
arthroplasty to accelerate the posterior location of the femoral prosthesis when
flexed at about 70 degrees, thus enhancing flexion. These Insall-Burnstein and kinematic
interpositions became the foundation of modern knee arthroplasty.
Despite the developments of modern joint replacement designs, complications of

the femoro-patellar joint were frequent after knee arthroplasty in the 1980s and 1990s,
which led to the development of todays knee arthroplasty which increases contact
surface in the femoro-patellar joint and prevents lateral displacement of patellar
bone.
4.1.2.5. Unicompartmental knee arthroplasty
Although it has been used since its introduction in 1950s, the results of
unicompartmental knee arthroplasty remain controversial. In the early 1970s, several
authors reported unsatisfactory results for unicompartmental knee arthroplasties but
over the next decade, better surgical techniques and proper patient selection
improved results.
42
Knee prostheses
Unicompartmental knee arthroplasty can be used in cases with up to moderate

arthritis and when diseased is confined to one compartment (Fig. 29). Only the damaged
compartment is replaced with metal and plastic and the healthy cartilage and bone in the
rest of the knee is left alone.
Fig. 29. Arthritis limited to the lateral compartment.
4.1.2.5.1. Advantages vs. disadvantages

Multiple studies have shown that modern unicompartmental knee replacement
performs very well in the vast majority of patients who are appropriate candidates. The
recently reported long-term endurance of unicompartmental knee arthroplasty is about
85-95%, which is similar to total arthroplasty. Therefore, if patients are properly selected
and an adequate technique is used, it may be a good surgical option.
Table 5. Advantages and disadvantages of unicompartmental knee replacement
Advantages Disadvantages
Preserves cruciate ligaments Less predictable pain relief
Recovering almost the full range of Potential need for more surgery
motion of the normal knee joint Deep vein thrombosis (blood clots)
Easier revision surgery Infections after surgery
Smaller incision and smaller scars
Less post-operative pain
Quicker operation
Shorter recovery period
Less trauma of soft tissue
Reduce blood loss during surgery
Reduce costs
43
Knee prostheses
However, two of the most significant benefits are:
1) Patients report that their replaced knee feels more natural as compared to a
total knee arthroplasty.
2) Leave other options open to further advances. By not replacing the rest of the
knee with metal and plastic, if other options exist in years to come for arthritis in
these areas then a partial knee replacement does not burn that bridge
4.1.2.6. Patellar resurfacing
Patellar resurfacing was described as early as 1955. The first patellar resurfacing
materials were metallic components, but this design was limited because of problems
concerning metal to cartilage articulation. Subsequently, the polyethylene patellar
prosthesis was developed and satisfactory results were obtained. Present day knee
arthroplasty became total knee replacement when patellar component was added.
4.1.3. Types of prostheses designs

Nowadays, many types of prostheses are used for total knee arthroplasty.
However, controversy exists regarding which prostheses are the most appropriate for
individual surgeons and specific patients. Therefore, we compare the advantages and
disadvantages of each type of interposition knee arthroplasty.
4.1.3.1. Posterior cruciate ligament retention vs. substitution
All knee arthroplasties require anterior cruciate ligament removal, but retention of
the posterior cruciate ligament depends on the type of arthroplasty.
The preservation type, in which posterior cruciate ligament is preserved (Fig. 30)
Is better for performing functions, such as, climbing stairs
Simplify revision surgery due to less loss of bone
Is appropriate for a patient whose posterior cruciate ligament is healthy

enough to continue stabilizing the knee joint.
o Soft tissue balance is not easy to achieve for knee joints with
degenerative arthritis
Increases the risk of early failure due to polyethylene insert

overloading caused by posterior cruciate ligament unbalanced
tension.
44
Knee prostheses
The substitution type, in which posterior cruciate ligament is replaced (Fig. 31)
Even degenerative knee joints with relatively severe deformities can

achieve ligament balance
When flexed at 60-70 degrees, the post of the tibial polyethylene contacts
the cam of the femoral component and induces posterior placement of
femoral bone, which allows relatively satisfactory rollback and can
achieve sufficient knee joint flexion.
o Revision surgery is difficult due to bone loss at the intercondylar notch
o Fracture may occur intra-operatively or post-operatively in patients with

small femurs.
From the biomechanical perspective, neither posterior cruciate ligament

preservation nor substitution types can totally replace the biomechanics of the
normal knee joint. Furthermore, many clinical studies have concluded that there is no
significant difference between these two types of prosthesis.
Fig. 30. Posterior-retention prosthesis

Fig. 31. Posterior-substitution prosthesis
45
Knee prostheses
4.1.3.2. Mobile vs. fixed bearing prostheses
In the fixed-bearing prosthesis, the polyethylene of the tibial component is

attached firmly to the metal implant beneath. The femoral component then rolls on this
cushioned surface. In some cases, excessive activity and/or extra weight can cause a
fixed-bearing prosthesis to wear down more quickly. Worn components can loosen from
the bone and cause pain. Loosening is a major reason some artificial joints fail. If you
are younger, more active, and/or overweight, your doctor may recommend a rotating
platform/mobile-bearing knee replacement. These implants are designed for potentially
longer performance with less wear.
In the mobile-bearing prosthesis (Fig. 32), the polyethylene insert can rotate short
distances inside the metal tibial tray. This is designed to allow patients a few degrees of
greater rotation to the medial and lateral sides of their knee. Compared with fixed-bearing
designs, mobile-bearing knee implants require more support from soft tissues, such as
the ligaments surrounding the knee. If the soft tissues are not strong enough, mobile-
bearing knees are more likely to dislocate. They also may cost more than fixed-bearing
implants.
In Europe, this mobile bearing prosthesis has been used for decades with good
clinical results, but recent reports have found no significant differences between this
mobile bearing prosthesis and fixed bearing polyethylene. In addition, there are no
studies which show better durability, improvement in pain, or improvement of function
with a mobile bearing design.
Fig. 32. Mobile bearing knee prosthesis
46
Knee prostheses
4.1.3.3. Non-cemented vs. cemented prostheses
Two types of fixation are used to hold knee implants in place.
Cemented fixation uses a fast-curing bone cement

(polymethylmethacrylate).
Non-cemented fixation relies on new bone growing into the surface of the
implant for fixation.
Non-cemented implants are made of a material that attracts new bone growth.
Most are textured or coated so that the new bone actually grows into the surface of the
implant. However, because they depend on new bone growth for stability, non-cemented
implants require a longer healing time than cemented replacements. In addition, non-
cemented implants, unfortunately, showed higher failure rates than cemented knee
arthroplasties due to aseptic loosening and bone loss.
There is also hybrid fixation. In hybrid fixation for total knee replacement, the
femoral component is inserted without cement, and the tibial and patellar components
are inserted with cement.
Because non-cemented implants have not been used as long as cemented

implants, comparisons after long-term use are not possible. However, some studies
have shown that non-cemented fixation has success rates comparable to those of
cemented fixation. Nevertheless, non-cemented knee arthroplasty has not widely
adopted, but recent material developments have resulted in materials that enhance bone
ingrowth which has led to the use of non-cemented knee arthroplasty in young patients.
4.1.3.4. Constrained condylar prostheses
Revision total knee arthroplasty is often associated with poorer outcomes due to
bone loss and ligament damage, which can result in ligamentous laxity and imbalance.
A constrained condylar knee design (Fig. 33) was developed to resist coronal moments
in the plane caused by soft tissue deficiency. Constrained condylar knee designs have
the advantage of allowing changes in the center of rotation during flexion, and thereby,
theoretically impart less tangential anterior-posterior stress across the prosthetic
interface.
Excessive constraint is a problem and this causes failure by loosening the

prosthesis. Thus, in difficult knee arthroplasty cases, usage may be determined during
surgery by taking into consideration the need for constraint. For example, in severe
valgus knee joints, the polyethylene insert may be a good candidate, but posterior
cruciate substitution tibial bearing is recommended over the constrained type.
47
Knee prostheses
Fig. 33. Constrained condylar prosthesis
4.1.3.5. High flexion type prostheses
Generally, postoperative knee motion range for total knee arthroplasty is less
than 120 degrees. Recently, to obtain motion ranges similar to those of the normal knee
joint, high flexion femoral prostheses (Fig. 34) with a thickened posterior portion of
femoral prosthesis and a wider contact surface with the bearing are being used to
reduce contact pressure and wear. To prevent collision between the patellar ligament
and bearing at high degrees of flexion, a high flexion bearing with an oblique cutting
of the anterior bearing has been developed.
Furthermore, many authors have reported that high flexion knee arthroplasty can
result in smaller contact loadings and wider ranges of motion than previous knee
arthroplasties. However, in a clinical study that used both knee implants, high flexion
knee arthroplasty did not show a significant increase in knee joint flexion range.
This issue needs to be proven by long-term follow up over 10 to 15 years.
Fig. 34. High flexion prosthesis
48
Knee prostheses
4.1.3.6. Ultracongruent polyethylene bearings
The most important thing to remember when performing posterior cruciate

ligament preserving knee arthroplasty is to balance the posterior cruciate ligament and
prevent instability by ligament disruption when flexed. For these reasons, deep-dished
polyethylene insert (also called ultracongruent insert) was developed (Fig. 35). This
bearing insert has moderate conformity in coronal and sagittal planes, which can prevent
edge loading caused from paradoxical anterior translation due to elevation of the anterior
lip of the prosthesis, prevent elevation in flexion, and prevent posterior subluxation.
Ultracongruent bearings can:
Reduce cam-and-post wear or fracture that may occur after posterior

cruciate ligament substitution knee arthroplasties
Prevent bone loss at the intercondylar cutting site
This bearing represents a new concept in that it can also maintain the posterior
cruciate ligament and provide moderate conformity in total knee arthroplasty. Further
long-term clinical follow-up is required along with comparative clinical trials of posterior
cruciate ligament preservation, substitution, and sacrificing techniques.
Fig. 35. Ultracongruent polyethylene bearing
49
Knee prostheses
4.1.4. Materials
4.1.4.1. Material Criteria
The construction materials used must meet several criteria:
They must be biocompatible; that is, they can be placed in the body without
creating a rejection response.
They must fulfil mechanical and structural requirements; for example, they must
be strong enough to take weight bearing loads, flexible enough to bear stress
without breaking, and able to move smoothly against each other as required.
They must be durable and able to retain their strength and shape for a long time.
4.1.4.2. Most used materials
Implants are made of metal alloys, ceramic material, or strong plastic parts, and
can be joined to your bone by acrylic cement.
Metal parts are made of titanium or cobalt-chromium (CoCr) based alloys.
Plastic parts are made of ultra-high molecular weight polyethylene (UHMWP).
Ceramic parts are made of alumina (Al2O3), zirconia (ZrO2), silica (SiO2),
hydroxyapatite (Ca10(PO4)6(OH)2), titanium nitride (TiN), silicon nitride
(Si3N4)
4.1.4.3. Metals
Metals are used as biomaterials due to their excellent electrical and thermal
conductivity and mechanical properties. Since some electrons are independent in
metals, they can quickly transfer an electric charge and thermal energy. The mobile free
electrons act as the binding force to hold the positive metal ions together. This attraction
is strong, as evidenced by the closely packed atomic arrangement resulting in high
specific gravity and high melting points of most metals. Since the metallic bond is
essentially non-directional, the position of the metal ions can be altered without
destroying the crystal structure resulting in a plastically deformable solid.
Some metals are used as passive substitutes for hard tissue replacement such
as total hip and knee joints, for fracture healing aids as bone plates and screws, spinal
fixation devices, and dental implants because of their excellent mechanical properties
and corrosion resistance. Some metallic alloys are used for more active roles in devices
such as vascular stents, catheter guide wires, orthodontic arch wires, and cochlea
implants.
50
Knee prostheses
The biocompatibility of the metallic implant is of considerable concern because

these implants can corrode in an in vivo environment. The consequences of corrosion
are the disintegration of the implant material per se, which will weaken the implant, and
the harmful effect of corrosion products on the surrounding tissues and organs.
4.1.4.3.1. Co-Cr Alloys
There are basically two types of cobaltchromium alloys:
1) CoCrMo alloy (castable)

2) CoNiCrMo alloy (wrought by hot forging)
The castable CoCrMo alloy has been used for many decades in dentistry and,
relatively recently, in making artificial joints. The wrought CoNiCrMo alloy is relatively
new, now used for making the stems of prostheses for heavily loaded joints such as the
knee and hip.
The American Society for Testing Materials (ASTM) lists four types of CoCr alloys
which are recommended for surgical implant applications:
Cast CoCrMo alloy (F75)

Wrought CoCrWNi alloy (F90)
Wrought CoNiCrMo alloy (F562)
Wrought CoNiCrMoWFe alloy (F563)
At the present time only two of the four alloys are used extensively in implant
fabrications, the castable CoCrMo and the wrought CoNiCrMo alloy.
The two basic elements of the CoCr alloys form a solid solution of up to 65% Co.
The molybdenum is added to produce finer grains which results in higher strengths
after casting or forging. The chromium enhances corrosion resistance as well as solid
solution strengthening of the alloy. The CoNiCrMo alloy originally called MP35N
(Standard Pressed Steel Co.) contains approximately 35% Co and Ni each. The alloy is
highly corrosion resistant to seawater (containing chloride ions) under stress.
There is a considerable difficulty of cold working on CoNiCrMo alloy, especially

when making large devices such as hip joint stems. Only hot-forging can be used to
fabricate a large implant with the alloy. The abrasive wear properties of the wrought
CoNiCrMo alloy are similar to the cast CoCrMo alloy (about 0.14 mm/yr in joint simulation
tests with ultra-high molecular weight polyethylene acetabular cup); however, the former
is not recommended for the bearing surfaces of joint prosthesis because of its poor
frictional properties with itself or other materials. The superior fatigue and ultimate tensile
strength of the wrought CoNiCrMo alloy make it suitable for the applications which
51
Knee prostheses
require long service life without fracture or stress fatigue. Such is the case for the stems
of the hip joint prostheses. This advantage is better appreciated when the implant has to
be replaced, since it is quite difficult to remove the failed piece of implant embedded
deep in the femoral medullary canal.
The mechanical properties required for CoCr alloys are given in Table 6. As with
the other alloys, the increased strength is accompanied by decreased ductility. Both the
cast and wrought alloys have excellent corrosion resistance.
Table 6. Mechanical Property Requirements of Co-Cr Alloys
The metallic products released from the prosthesis because of wear, corrosion,
and fretting may damage organs and local tissues. In vitro studies have indicated that
particulate Co is toxic to human osteoblast like cell lines and inhibits synthesis of type-I
collagen, osteocalcin and alkaline phosphatase in the culture medium. However,
particulate Cr and CoCr alloy are well tolerated by cell lines with no significant toxicity.
The toxicity of metal extracts in vitro have indicated that Co and Ni extracts at 50%
concentration appear to be highly toxic since all viability parameters were altered after
24 h. However, Cr extract seems to be less toxic than Ni and Co.
The modulus of elasticity for the CoCr alloys does not change with the changes
in their ultimate tensile strength. The values range from 220 to 234 GPa which are higher
than other materials such as stainless steels. This may have some implications of
different load transfer modes to the bone in artificial joint replacements, although the
effect of the increased modulus on the fixation and longevity of implants is not clear.
52
Knee prostheses
4.1.4.3.2. Ti Alloys
Attempts to use titanium for implant fabrication dates to the late 1930s. Titaniums
lightness (4.5 g/cm3) and good mechanical and chemical properties are salient
features for implant application. There are four grades of unalloyed commercially pure
titanium for surgical implant applications. The impurity contents separate them; oxygen,
iron, and nitrogen should be controlled carefully. Oxygen in particular has a great
influence on the ductility and strength.
Table 7. Mechanical Properties of Ti and its Alloys (ASTM F136)
One titanium alloy (Ti6Al4V) is widely used to manufacture implants and its
chemical requirements are given in Table 1.7. The main alloying elements of the alloy
are aluminum (5.56.5%) and vanadium (3.54.5%). The Ti6Al4V alloy has
approximately the same fatigue strength (550 MPa) of CoCr alloy after rotary bending
fatigue tests. Titanium is an allotropic material, which exists as a hexagonal close packed
structure (hcp, -Ti) up to 882 C and body-centered cubic structure (bcc, -Ti) above that
temperature. Titanium alloys can be strengthened and mechanical properties varied by
controlled composition and thermomechanical processing techniques. The addition of
alloying elements to titanium enables it to have a wide range of properties:
Aluminum tends to stabilize the -phase, that is increase the transformation

temperature from - to -phase
Vanadium stabilizes the -phase by lowering the temperature of the
transformation from to .
The -alloy has a single-phase microstructure which promotes good weldability.

The stabilizing effect of the high aluminum content of these groups of alloys makes
excellent strength characteristics and oxidation resistance at high temperature.
The addition of controlled amounts of -stabilizers causes the higher strength -

phase to persist below the transformation temperature which results in the two-phase
system. The precipitates of -phase will appear by heat treatment in the solid solution
temperature and subsequent quenching, followed by aging at a somewhat lower
temperature. The aging cycle causes the coherent precipitation of some fine particles
from the metastable , imparting structure may produce local strain field capable of
53
Knee prostheses
absorbing deformation energy. Cracks are stopped or deterred at the -particles, so that
the hardness is higher than for the solid solution. The higher percentage of -stabilizing
elements (13%V in Ti13V11Cr3Al alloy) results in a microstructure that is substantially
which can be strengthened by heat-treatment.
Another Ti alloy (Ti13Nb13Zr) with 13%Nb and 13%Zr showed martensite

structure after water quenched and aged, which showed high corrosion resistant with
low modulus (E = 79 MPa). Formation of plates of martensite induces considerable
elastic distortion in the parent crystal structure and increases strength.
The mechanical properties of the commercially pure titanium and its alloys are
given in Table 1.7. The modulus of elasticity of these materials is about 110 GPa except
13Nb13Zr alloy. From Table 7 one can see that the higher impurity content of the cp-Ti
leads to higher strength and reduced ductility. When compared by the specific strength
(strength per density) the titanium alloys exceed any other implant materials. Titanium,
nevertheless, has poor shear strength making it less desirable or bone screws, plates,
and similar applications. It also tends to gall or seize when in sliding contact with itself or
another metal. Titanium derives its resistance to corrosion by the formation of a solid
oxide layer to a depth of 10 nm.
Under in vivo conditions the oxide (TiO2) is the only stable reaction product.
However, micro motion at the cement-prosthesis and cement-bone are inevitable and
consequently, titanium oxide and titanium alloy particles are released in cemented joint
prosthesis. Sometimes this wear debris accumulates as periprosthetic fluid collections
and triggers giant cell response around the implants. This cystic collection continued to
enlarge and aspiration revealed dark heavily stained fluid containing titanium wear
particles and histiocytic cells. Histological examination of the stained soft tissue showed
fibrin necrotic debris and collagenous, fibrous tissue containing a histiocytic and foreign
body giant cell infiltrate. The metallosis, black staining of the periprosthetic tissues, has
been implicated in knee implant.
4.1.4.4. Ceramics
Ceramics are refractory, polycrystal line compounds, usually inorganic, including

silicates, metallic oxides, carbides and various refractory hydrides, sulfides, and
selenides. Oxides such as Al2O3, MgO, SiO2, and ZrO2 contain metallic and nonmetallic
elements and ionic salts, such as NaCl, CsCl, and ZnS. Exceptions to the preceding
include covalently bonded ceramics such as diamond and carbonaceous structures like
graphite and pyrolized carbons.
54
Knee prostheses
Ceramics in the form of pottery have been used by humans for thousands of
years. Until recently, their use was somewhat limited because of their inherent
brittleness, susceptibility to notches or micro-cracks, low tensile strength, and low
impact strength. However, within the last 100 years, innovative techniques for
manufacturing ceramics have led to their use as high tech materials. In recent years,
humans have realized that ceramics and their composites can also be used to augment
or replace various parts of the body, particularly bone. Thus, the ceramics used for the
latter purposes are classified as bioceramics. Their relative inertness to the body fluids,
high compressive strength, and aesthetically pleasing appearance led to the use of
ceramics in dentistry as dental crowns. Some carbons have found use as implants
especially for blood interfacing applications such as heart valves. Due to their high
specific strength as fibers and their biocompatibility, ceramics are also being used as
reinforcing components of composite implant materials and for tensile loading
applications such as artificial tendon and ligaments.
Unlike metals and polymers, ceramics are difficult to shear plastically due to the
(ionic) nature of the bonding and minimum number of slip systems. These characteristics
make the ceramics non-ductile and are responsible for almost zero creep at room
temperature. Consequently, ceramics are very susceptible to notches or micro
cracks because instead of undergoing plastic deformation (or yield) they will fracture
elastically on initiation of a crack. At the crack tip the stress could be many times higher
than the stress in the material away from the tip, resulting in a stress concentration which
weakens the material considerably. The latter makes it difficult to predict the tensile
strength of the material (ceramic).
This is also the reason ceramics have low tensile strength compared to
compressive strength. If a ceramic is flawless, it is very strong even when subjected to
tension. Flawless glass fibers have twice the tensile strengths of high strength steel (7
GPa).
Ceramics are generally hard; in fact, the measurement of hardness is calibrated

against ceramic materials. Diamond is the hardest, with a hardness index of 10 on Mohs
scale, and talc (Mg3Si3O10COH) is the softest ceramic (Mohs hardness 1), while
ceramics such as alumina (Al2O3; hardness 9), quartz (SiO2; hardness 8), and apatite
(Ca5P3O12F; hardness 5) are in the middle range. Other characteristics of ceramic
materials due to the chemical bonding within ceramics are:
High melting temperatures

Low conductivity of electricity and heat
55
Knee prostheses
In order to be classified as a bioceramic, the ceramic material must meet or

exceed the properties listed in Table 8.
Table 8. Desired Properties of Implantable Bioceramics
4.1.4.4.1. Alumina (Al2O3)
The ASTM specifies that alumina for implant use should contain 99.5% pure
alumina and less than 0.1% combined SiO2 and alkali oxides (mostly Na2O) (F603-78).
Alpha alumina has a rhombohedral crystal structure. The single crystal form of
alumina has been used successfully to make implants. Single crystal alumina can be
made by feeding fine alumina powders onto the surface of a seed crystal which is slowly
withdrawn from an electric arc or oxy-hydrogen flame as the fused powder builds up.
Single crystals of alumina up to 10 cm in diameter have been grown by this method.
The strength of polycrystalline alumina depends on its grain size and porosity.
Generally, the smaller the grains, the lower the porosity and the higher the
strength. The ASTM standards (F603-78) requires a flexural strength greater than 400
MPa and elastic modulus of 380 GPa (Table 9).
Table 9. Physical Property Requirements of Alumina and Zirconia
56
Knee prostheses
Single crystal alumina has been used in orthopedics and dental surgery for
almost 20 years. Alumina is usually a quite hard material, its hardness varies from 20 to
30 GPa. This high hardness permits its use as an abrasive and as bearings for watch
movements. Both polycrystalline and single crystal alumina have been used clinically.
The high hardness is accompanied by low friction and wear and inertness to the in
vivo environment. These properties make alumina an ideal material for use in joint
replacements.
4.1.4.4.2. Zirconia (ZrO2)
Pure zirconia can be obtained from chemical conversion of zircon (ZrSiO4), which
is an abundant mineral deposit. Zirconia has a high melting temperature (Tm = 2953 K)
and chemical stability. It undergoes a large volume change during phase changes at high
temperature in pure form; therefore, a dopant oxide such as Y2O3 is used to stabilize
the high temperature (cubic) phase. We have used 6 mole% Y2O3 as dopant to make
zirconia for implantation in bone. Zirconia produced in this manner is referred to as
partially stabilized zirconia. However, the physical properties of zirconia are somewhat
inferior to that of alumina (Table 2.5).
4.1.4.4.3. Calcium phophate
Calcium phosphate has been used in the form of artificial bone. This material has
been synthesized and used for manufacturing various forms of implants, as well as for
solid or porous coatings on other implants.
Calcium phosphate can be crystallized into salts such as hydroxyapatite and -

whitlockite depending on the Ca: P ratio, presence of water, impurities, and temperature.
In a wet environment and at lower temperatures (<900C), it is more likely that hydroxyl-
or hydroxyapatite will form, while in a dry atmosphere and at a higher temperature, -
whitlockite will be formed. Both forms are very tissue compatible and are used as bone
substitutes in a granular form or a solid block. The apatite form of calcium phosphate is
considered to be closely related to the mineral phase of bone and teeth.
The mineral part of bone and teeth is made of a crystalline form of calcium
phosphate similar to hydroxyapatite [Ca10(PO4)6(OH)2]. The apatite family of mineral
[A10(BO4)6X2] crystallizes into hexagonal rhombic prisms and has unit cell dimensions
a = 9.432 and c = 6.881 .
The ideal Ca: P ratio of hydroxyapatite is 10: 6 and the calculated density is 3.219
g/cm3. Substitution of OH with fluoride gives the apatite greater chemical stability due to
the closer coordination of fluoride (symmetric shape) as compared to the hydroxyl
57
Knee prostheses
(asymmetric, two atoms) by the nearest calcium. This is why fluoridation of drinking water
helps in resisting caries of the teeth.
Table 10. Physical Properties of Calcium Phosphate
The mechanical properties of synthetic calcium phosphates vary considerably

(Table 10). The wide variations in properties of polycrystalline calcium phosphates are
due to the variations in the structure and manufacturing processes. Depending on the
final firing conditions, the calcium phosphate can be calcium hydroxyapatite or -
whitlockite. In many instances, both types of structures exist in the same final product.
Polycrystalline hydroxyapatite has a high elastic modulus (40 to 117 GPa). Hard
tissue such as bone, dentin, and dental enamel are natural composites which contain
hydroxyapatite (or a similar mineral), as well as protein, other organic materials, and
water. Enamel is the stiffest hard tissue, with an elastic modulus of 74 GPa, and contains
the most mineral. Dentin (E = 21 GPa) and compact bone (E = 12 to 18 GPa) contain
comparatively less mineral. The Poissons ratio for the mineral or synthetic
hydroxyapatite is about 0.27 which is close to that of bone (0.3).
Among the most important properties of hydroxyapatite as a biomaterial is its

excellent biocompatibility.
Hydroxyapatite appears to form a direct chemical bond with hard tissues. On

implantation of hydroxyapatite particles or porous blocks in bone, new lamellar
cancellous bone forms within 4 to 8 weeks.
58
Knee prostheses
4.1.4.5. Polymeric Biomaterials
Synthetic polymeric materials have been widely used in medical disposable

supply, prosthetic materials, dental materials, implants, dressings, extracorporeal
devices, encapsulants, polymeric drug delivery systems, tissue engineered products,
and orthodoses as that of metal and ceramics substituents.
The main advantages of the polymeric biomaterials compared to metal or ceramic

materials are:
Ease of manufacturability to produce various shapes

Ease of secondary processability
Reasonable cost
Availability with desired mechanical and physical properties
The required properties of polymeric biomaterials are similar to other

biomaterials, that is, biocompatibility, sterilizability, adequate mechanical and physical
properties, and manufacturability as given in Table 11.
Table 11. Requirements for Biomedical Polymers
Although hundreds of polymers are easily synthesized and could be used as

biomaterials only ten to twenty polymers are mainly used in medical device fabrications
from disposable to long-term implants.
4.1.4.5.1. Polyethylene (PE)
Polyethylene is available commercially in five major grades:
1) High density (HDPE)

2) Low density (LDPE)
3) Linear low density (LLDPE)
4) Very low density (VLDPE)
5) Ultra-High molecular weight (UHMWPE)
59
Knee prostheses
HDPE is polymerized in a low temperature (6080C), and at a low pressure (10

kg/cm2) using metal catalysts. A highly crystalline, linear polymer with a density ranging
from 0.94 to 0.965 g/cm3 is obtained. LDPE is derived from a high temperature (150
300C) and pressures (10003000 kg/cm2) using free radical initiators. A highly
branched polymer with lower crystallinity and densities ranging from 0.915 to 0.935
g/cm3 is obtained. LLDPE (density: 0.910.94 g/cm3) and VLDPE (density: 0.88 0.89
g/cm3), which are linear polymers, are polymerized under low pressures and
temperatures using metal catalysts with co-monomers such as 1-butene, 1-hexene, or
1-octene to obtain the desired physical properties and density ranges.
HDPE is used in pharmaceutical bottles, nonwoven fabrics, and caps. LDPE is

found in flexible container applications, nonwoven-disposable and laminated (or
coextruded with paper) foil, and polymers for packaging. LLDPE is frequently employed
in pouches and bags due to its excellent puncture resistance and VLDPE is used in
extruded tubes. UHMWPE (MW >2106 g/mol) has been used for orthopedic implant
fabrications, especially for load-bearing applications such as an acetabular cup of total
hip and the tibial plateau and patellar surfaces of knee joints.
4.1.4.5.2. Polymethylmetacrylate (PMMA)
Commercial PMMA is an amorphous (Tg: 105C and density: 1.15 to 1.195

g/cm3) material with good resistance to dilute alkalis and other inorganic solutions.
PMMA is best known for its exceptional light transparency (92% transmission), high
refractive index (1.49), good weathering properties, and as one of the most
biocompatible polymers.
PMMA can be easily machined with conventional tools, molded, surface coated,
and plasma etched with glow or corona discharge. PMMA is used broadly in medical
applications such as a blood pump and reservoir, an IV system, membranes for blood
dialyzer, and in in vitro diagnostics. It is also found in contact lenses and implantable
ocular lenses due to excellent optical properties, dentures, and maxillofacial prostheses
due to good physical and coloring properties, and bone cement for joint prostheses
fixation (ASTM standard F451).
60
Chapter 5: Manual segmentation of a knee
Manual segmentation of a knee
5.1. Manual segmentation of a knee
This tutorial describes the whole process, used in this project, for preparing data
to 3D printing. Based on DICOM files provided by the OsteoArthritis Initiative (OAI), we
will see how we can generate the 3D model of a real patients knee and then obtain an
STL file which is necessary to run the 3D printing.
The objective of this tutorial is to guide step by step through the process of manual
segmentation of knee structures from MRI data. To do this we will use 3D Slicer, a free
and open source software package for image analysis and scientific visualization.
The tutorial is divided into five key steps as follows:
Introduction to 3D Slicer
Loading DICOM files
Bone segmentation
Building 3D model
Saving model STL file generation
Before to start using 3D Slicer its necessary to select the DICOM file that we are
going to use over this process. All the information about DICOM files used in this project
are available in Appendix A.
The selection process consisted on: use the MRI T2* studies (SAG_3D_DESS)
because they has the smallest slice thickness (0.7 mm) which implies a better image
resolution; and from those files select the enrollment visit. At that point the amount of
studies were 427 (212 left knees and 215 right knees) from 215 different patients. Then
8 studies were selected from that group based on its image quality.
Finally one study was selected from the last group based on its detailed image
quality and its feasibility of been segmented. Therefore the study selected was the right
knee from patient ID: 9034812.
63
5.1.1. Introduction to 3D Slicer

First of all, you have to install 3D Slicer on your computer. To do that, you can
download the software installer from http://download.slicer.org/ and follow the
instructions.
Once youve installed 3D Slicer on your computer, you will see a welcome screen
like this (Fig. 36).
Fig. 36. Home screen of 3D Slicer
Slicer has a modular design, so depending on what you are doing or what tools
you need youll open up different modules within Slicer. Depending on which module you
are using, the left side of your user interface will change.
Using the drop down menu at the top left of the interface you can access to
different modules (Fig. 37).
The menu is divided in three sections, so you can find the module you need in
three different ways. First you can select it from All Modules which is a list of all the
modules available. Beneath that you have a list of the most common used modules, and
finally at the bottom of the menu you have modules organized by categories.
On the other hand, the right part of the interface is where you are visualizing your
imaging data. In the conventional view, youll see the 3D aspect of your data at the top,
and three slice view at the bottom. You can change the view using the layout icon on the
toolbar (Fig. 38).
64
Fig. 37. Drop down menu to select modules.
Fig. 38. Changing visualizing of imaging data.
65
At the top side of the user interface is placed the toolbar, where we can find
different features (Fig. 39).
Fig. 39. Toolbar
5.1.2. Loading DICOM files

Before start the segmentation process you need to load the DICOM data into
Slicer. In order to do that you click on Load DICOM Data button (Fig. 5) and a DICOM
Browser window pops up (Fig. 40).
Fig. 40. Load DICOM Data button
In this window you can import the DICOM data needed into the browser and then
select which one you want to load into Slicer. In order to do that, you click on Import
(Fig. 41), then a file browser pops up and you select the folder where the DICOM data is
saved and click Import.
After that you will see the data imported into the DICOM Browser (Fig. 42). In the
top row you will see the patients information, in the mid row the studies and beneath that
the series.
66
Fig. 41. DICOM Browser
Finally, to load the DICOM data you need to select the studies you want to load
and then click on Load (Fig. 42).
Fig. 42. Loading DICOM data
When the data is loaded you will see the medical images in the sagittal, coronal
and axial view at the right side of the interface (Fig. 43).
67
Fig. 43. DICOM data loaded
Once youve loaded the data you can scroll through each slice view using the
scroll pad. Also by left clicking and dragging you can change the brightness (up and
down) and the contrast (left and right), and by right clicking and dragging you can zoom
in (down) and zoom out (up). Also you can pan the data by holding the shift key and left
clicking and dragging.
5.1.3. Bone segmentation

Select the editor module (Fig. 44). You will see two main tags: Create and
Select Label Maps and Edit Selected Label Map.
In order to segment the bone part you need to create a label map. A label map is
a 3D scalar volume where each of the voxels of that volume is a number and that number
indicates the type of tissue at that location. All this label volume are associated with a
color that matches the number with a color and textures.
To create a label map you need to be sure that the original volume is selected as
Master Volume and the new label map generated is selected as Merge Volume. There
are a great range of tools and effects in the Editor module you can use to manipulate the
label map (see Appendix B).
There is not a unique way to segment the bone part of the knee, so the following
steps are only one possibility of many of them.
68
Fig. 44. Editor module
First of all we are going to use the Threshold Effect. Before you click on the tool
icon you need to select the appropriate label value, so you need to click on the color
swatch and select, from the full list of available labels, the color which correspond to
bone (number 2). Now you can click on the threshold effect icon. Then you need to
define two values, the lower and upper bound of the threshold. In this case the bone part
comes up with a lower intensity because of T2 MRI shows bone in darker colors. For this
reason the recommended threshold is in between 20 and 90 (Fig. 45).
Fig. 45. Threshold Effect
69
Once set the threshold click on apply button and the effect will be done (Fig. 46).
Fig. 46. Selected threshold
The next step consist on use PaintEffect. The objective is to paint only the knee
part which correspond to bone, so you have to paint with label 2 all the bone part, and
with label 0 (background) the rest. Repeat this step for each slice through the volume
and for each view (sagittal, coronal and axial).This step takes a lot of time and has to be
repeated many times to obtain a great result (Fig. 47).
Fig. 47. After apply PaintEffect
Even though the only effects used in this project were Threshold Effect and
Paint Effect, there are other useful effect you can use. For example, Change Island
is a great effect which allows you to separate two different anatomical structures if they
are clearly separated in two islands (a group of connected pixels with the same label).
70
5.1.4. Building 3D model

To build the 3D model you need to use the effect Modelmaker in Editor Module
(Fig. 48). Click on the corresponding icon, select the bone color from the color swatch,
choose a name for the model and click on apply.
Fig. 48. Modelmaker effect
After that, you will see the 3D model generated in the right side (Fig. 49).
Fig. 49. 3D model
71
5.1.5. Saving model

Once you have created the 3D model, is necessary to save it in a file compatible
with 3D printers (STL file). In order to do that, click on Save button at the top left of the
interface and a new window will pop up. In this window three columns appear, from left
to right you will see first the file name of all the different pieces of information that are
part of your slicer interface at the moment, then at the next column are the file format in
which you can see that information, and at the final column are the directory where you
can find those files. Find the model generated by its name, select it, then select STL
(.stl) in file format column, choose a directory where you want to save it and finally click
on save (Fig. 50).
Fig. 50. Save model to 3D printing compatible format
72
Chapter 6: Conclusion
Conclusion
6.1. Conclusin
En conclusin, se logr generar e imprimir un modelo 3D de una rodilla con

osteoartritis en base a imgenes mdicas, mediante el uso del software 3D Slicer. Se
logr la segmentacin de la parte sea de la rodilla, pero no la de sus elementos blandos
como cartlago, ligamentos o msculos, ya que esto excede los propsitos de este
proyecto. Se lograron utilizar imgenes mdicas especialmente tomadas para el
diagnstico de osteoartritis como son las RMN del tipo T2*, a pesar de no ser las ms
adecuadas para la visualizacin del hueso. Por otro lado, no fue posible utilizar
herramientas de segmentacin automtica, debido principalmente al tipo de imgenes
antes mencionadas y a la calidad de las mismas.
Por lo tanto, queda demostrado que es posible segmentar la parte sea de una
rodilla afectada por osteoartritis, basndose en imgenes mdicas reales y especficas
para el diagnstico de dicha enfermedad, y que una vez segmentada puede imprimirse
un prototipo 3D de la misma para futuros anlisis y estudios.
75
Conclusion
6.2. Conclusion
In conclusion, it was possible to generate and print a 3D model of knee with

osteoarthritis based on medical images and using 3D Slicer software. The bone part of
the knee was segmented, but not the soft tissue such as cartilage, ligaments or muscles,
because it was out of the project objective. Medical images taken especially for diagnosis
of osteoarthritis such as MRI T2* were successfully used, even though they are not
appropriate for bone visualization. On the other hand, automatic segmentation tools
could not be used, mainly due to the quality and type of the abovementioned images.
Therefore, it is demonstrated that it is possible to segment the bone part of a knee

affected by osteoarthritis, based on real and specific medical images, and after that a 3D
prototype can be printed and used for future analysis and studies.
76
Conclusion
6.3. Trabajo futuro
Las propuestas sobre trabajo futuro podran ser entre otras:
Utilizacin de otros softwares de segmentacin, tanto para lograr mejores

resultados como as tambin reducir el tiempo de esta tarea. En la misma lnea, podra
pensarse incluso en el desarrollo de algoritmos para lograr una segmentacin
automtica.
A modo de complementar lo ya hecho, podra trabajarse sobre la segmentacin

de las partes blandas (cartlago y ligamentos) y de ser posible lograr imprimirlos con los
materiales adecuados y as poder tener un modelo an ms real de la rodilla. A esto se
le podra agregar el lograr un prototipo funcional de la misma, donde pueda apreciarse
la correcta cinemtica de la rodilla.
Por otro lado, se podra investigar ms en lo que confiere al diseo de implantes

y lograr disear un prototipo de implante personalizado para este paciente. Esto ltimo
seria de particular relevancia e inters, ms all del campo de la biomedicina, ya que
adems de ayudar en el diagnstico y tratamiento del paciente, mejorando su calidad
de vida y reinsertndolo ms rpidamente en sus tareas diarias, tambin a nivel
socioeconmico significa una reduccin en tiempos de hospitalizacin y
consecuentemente una reduccin en costos de servicio mdico.
77
Conclusion
6.4. Future work
Proposals for future work could be:
To use other segmentation software to achieve better results as well as reduce

the time of this task. Also, to develop algorithms to achieve automatic segmentation.
As a complement, it would be useful to work on the segmentation of soft tissue

(cartilage and ligaments) and if possible to achieve, to print those parts with suitable
materials so we can have an even more realistic model of the knee. Other interesting
topic would be to achieve a functional prototype, which shows the correct kinematics of
the knee.
On the other hand, it would be interesting to do a further investigation about the

implant design process, in order to achieve a customized implant prototype for this
patient. This would be of particular relevance and interest, not only into the biomedicine
field, helping on the diagnosis and treatment of patients, improving their life quality and
reinserting them faster in their daily tasks, but also into the socioeconomic field, reducing
hospitalization times and consequently reducing medical services costs.
78
Chapter 7: References
References
7.1 References
Wikipedia, (2015). Osteoarthritis. [online] Available at:
https://en.wikipedia.org/wiki/Osteoarthritis
Arthritis.org, (2015). Osteoarthritis Causes. [online] Available at:
http://www.arthritis.org/about-arthritis/types/osteoarthritis/causes.php
Woolf, A. and Pfleger, B. (2003). Burden of major musculoskeletal conditions. Bulletin of the
World Health Organization, [online] 81(9), pp.646-656. Available at:
http://www.scielosp.org/scielo.php?pid=S0042-
96862003000900007&script=sci_arttext#FIG01
Who.int, (2015). WHO | Chronic rheumatic conditions. [online] Available at:
http://www.who.int/chp/topics/rheumatic/en/
Gray, H. and Clemente, C. (1985). Anatomy of the human body. Philadelphia: Lea & Febiger.
Wikipedia, (2015). Lower extremity of femur. [online] Available at:
https://en.wikipedia.org/wiki/Lower_extremity_of_femur
Wikipedia, (2015). Upper extremity of tibia. [online] Available at:
https://en.wikipedia.org/wiki/Upper_extremity_of_tibia
Wikipedia, (2015). Knee. [online] Available at: https://en.wikipedia.org/wiki/Knee
Kaltenborn, F.M. (1989). Manual Mobilization of the Extremity Joints (4th ed.). Minneapolis:
OPTP
Eun-Kyoo Song, Jong-Keun Seon, Jae-Young Moon and Yim Ji- Hyoun (2013). The
Evolution of Modern Total Knee Prostheses, Arthroplasty - Update, Prof. Plamen Kinov (Ed.),
ISBN: 978-953-51-0995-2, InTech, DOI: 10.5772/54343. Available from:
http://www.intechopen.com/books/arthroplasty-update/the-evolution-of-modern-total-knee-
prostheses
PANESSO, Mara Claudia[et al.] (2008). Biomecnica clnica de la rodilla.- Facultad de
Rehabilitacin y Desarrollo Humano. Bogot: Editorial Universidad de Rosario, 2009. ISBN:
1794-1318. Available from: http://www.urosario.edu.co/urosario_files/4f/4f59d9d9-1c91-4115-
9206-7b2b96342c14.pdf
Sanjuan Cerver R., Jimnez Honrado P. J., Gil Monz E. R., Snchez Rodrguez R. J.
Fenollosa Gmez J. Biomechanics of the knee Patologa del Aparato Locomotor, 2005; 3 (3):
189-200
Moon.ouhsc.edu, (2015). Arthrokinematics and osteokinematics. [online] Available at:
http://moon.ouhsc.edu/dthompso/namics/arthkin.htm
Arthritis.org, (2015). Osteoarthritis Treatment. [online] Available at:
http://www.arthritis.org/about-arthritis/types/osteoarthritis/treatment.php
MedicineNet, (2015). Osteoarthritis Causes, Symptoms, Diagnosis and Treatment
Information on MedicineNet.com. [online] Available at:
http://www.medicinenet.com/script/main/art.asp?articlekey=90187
Www5.aaos.org, (2015). AAOS Informed Patient Tutorial. [online] Available at:
http://www5.aaos.org/icm/default.cfm?screen=icm005_s02_p1
Arthritisresearchuk.org, (2015). Complex or revision knee replacement | Arthritis Research
UK. [online] Available at: http://www.arthritisresearchuk.org/arthritis-information/surgery/knee-
replacement/different-types/complex-knee-replacement.aspx
81
References
Arthritisresearchuk.org, (2015). Total knee replacement | Arthritis Research UK. [online]

Available at: http://www.arthritisresearchuk.org/arthritis-information/surgery/knee-
replacement/different-types/total-knee-replacement.aspx
Arthritisresearchuk.org, (2015). Unicompartmental (partial) knee replacement | Arthritis
Research UK. [online] Available at: http://www.arthritisresearchuk.org/arthritis-
information/surgery/knee-replacement/different-types/partial-knee-replacement.aspx
Arthritisresearchuk.org, (2015). What are the different types of knee replacement surgery? |
Arthritis Research UK. [online] Available at: http://www.arthritisresearchuk.org/arthritis-
information/surgery/knee-replacement/different-types.aspx
Arthritisresearchuk.org, (2015). Kneecap replacement (patellofemoral arthroplasty) | Arthritis
Research UK. [online] Available at: http://www.arthritisresearchuk.org/arthritis-
information/surgery/knee-replacement/different-types/kneecap-replacement.aspx
BoneSmart, (2015). Knee Implants. [online] Available at: http://bonesmart.org/knees/knee-
implants/
Healthline, (2015). Guide to Knee Replacement Implants and Their Manufacturers. [online]
Available at: http://www.healthline.com/health/total-knee-replacement-surgery/implant-
manufacturers#3
Healthline, (2015). Knee Replacement: Your Surgical Options. [online] Available at:
http://www.healthline.com/health/total-knee-replacement-surgery/surgical-options#1
Kneereplacement.com, (2015). Types of Knee Replacements. [online] Available at:
https://www.kneereplacement.com/DePuy_technology/knee_options/types
Kneereplacement.com, (2015). Knee Replacement Basics. [online] Available at:
https://www.kneereplacement.com/DePuy_technology/knee_options/basics
My.clevelandclinic.org, (2015). Partial Knee Replacement Surgery & Recovery Information |
Cleveland Clinic. [online] Available at: https://my.clevelandclinic.org/services/orthopaedics-
rheumatology/treatments-procedures/hic-partial-knee-replacement
Orthoinfo.aaos.org, (2015). Knee Replacement Implants -OrthoInfo - AAOS. [online]
Available at: http://orthoinfo.aaos.org/topic.cfm?topic=a00221
Orthoinfo.aaos.org, (2015). Total Knee Replacement-OrthoInfo - AAOS. [online] Available at:
http://orthoinfo.aaos.org/topic.cfm?topic=a00389
Orthoinfo.aaos.org, (2015). Unicompartmental Knee Replacement-OrthoInfo - AAOS. [online]
Available at: http://orthoinfo.aaos.org/topic.cfm?topic=A00585
Thomas, P. (2014). Clinical and diagnostic challenges of metal implant allergy using the
example of orthopaedic surgical implants. Allergo J Int, 23(6), pp.179-185.
Wikipedia, (2015). Knee replacement. [online] Available at:
https://en.wikipedia.org/wiki/Knee_replacement
Wikipedia, (2015). Unicompartmental knee arthroplasty. [online] Available at:
https://en.wikipedia.org/wiki/Unicompartmental_knee_arthroplasty
Wong, J. and Bronzino, J. (2007). Biomaterials. Boca Raton: CRC Press.
Wiki.slicer.org, (2016). Documentation/Nightly/Modules/Editor - SlicerWiki. Available at:
http://wiki.slicer.org/slicerWiki/index.php/Documentation/Nightly/Modules/Editor.
82
Appendix
Appendix
APPENDIX A: DICOM Image Release Notes

TABLE OF CONTENTS
1. PARTICIPANTS FOR WHOM IMAGES ARE AVAILABLE

1.1 Participant Subcohorts & Demographics
1.2 Participant Clinical, Measurement and Biomarker Data
2. IMAGES AVAILABLE
2.1 Image Identifiers (Barcodes)
2.2 Radiographs
2.2.1 Fixed Flexion Knee Radiography and the Synaflexer Frame
2.2.2 Full-Limb X-rays
2.3 Knee MRI
2.4 Thigh MRI
2.5 Rules for Release of Acquired Images
2.5.1 Radiographs
2.5.2 Knee MRI
2.6 Future Updates to this Release
3. DISK CONTENTS
3.1 About the Files in Each Participants Folder
3.2 File and Folder Structure
4. DICOM INFORMATION IN IMAGES

4.1 Radiograph DICOM Tags
4.1.1 Note about pixel spacings in radiographs acquired and stored digitally
4.1.2 Note about pixel spacings in radiographs acquired on film and then digitized
4.1.3 Note about pixel spacings in radiographs acquired digitally, then printed on film
4.2 MRI DICOM Tags
5. DICOM VIEWING SOFTWARE AND RESOURCES

5.1 Freely Available DICOM Image Viewing Software
5.1.1 ImageJ
5.1.2 Other DICOM viewers
5.2 Viewing Software Installation
5.2.1 DicomWorks
5.2.2 ImageJ with DICOM Im/Ex plugin
5.2.3 Osiris and OsiriX
5.3 Viewing the Images
5.3.1 DicomWorks
5.3.2 ImageJ with DICOM Im/Ex Plugin
5.3.3 Osiris 4.19 for Windows.
6. ACKNOWLEDGEMENTS
7. REFERENCES
APPENDIX A: Details of Participants in Group A
APPENDIX B: Details of Participants in Group B
APPENDIX C: Details of Participants in Group C
APPENDIX D: Details of Participants in Group D
APPENDIX E: Details of Participants in Entire Cohort
APPENDIX F: Details of Participants in Group F
APPENDIX G: Details of Participants in Group G
APPENDIX X: MRI Pulse Sequence Parameters
85
Appendix
The data on this hard disk comprise x-ray and MRI images from the Osteoarthritis Initiative (OAI), and related files
that describe the data, along with information and software that might be useful for viewing the images. Information
about the protocols and study design for OAI can be found at the OAI Online website: http://www.oai.ucsf.edu/
Any feedback regarding this image release should be directed to OAIOnlineFeedback@psg.ucsf.edu , and
technical questions should be addressed to OAIImageHelp@psg.ucsf.edu
Updates since previous version (October, 2011):

72-month visit images are now available for the whole OAI cohort, and these include x-rays and knee MRIs.
Details of which images have been acquired with known image acquisition or protocol deviations are now
provided in the 72-month imaging meta-data datasets (see Section 1.2).
The rules for releasing MR images (section 2.5.2) were modified slightly to allow for a small number of
Sagittal DESS images where there are duplicate acquisitions (one where the scanner gradients were out of
specification ranges, and hence any quantitative measurements may be inaccurate). These images can be
identified using the imaging meta-data datasets and related documentation.
Due to software upgrades on the OAI MRI scanners, parametric maps of calculated T2 values become
available for many of the MR images acquired at the 72-month visit (see Sections 2.3 and 2.3.1 for details).
1. PARTICIPANTS FOR WHOM IMAGES ARE AVAILABLE
1.1 Participant Subcohorts & Demographics

Images are available for all 4796 OAI Study participants, of whom 1390 participants are in the Progression
subcohort, 3284 participants are in the incidence subcohort and 122 participants are non-exposed controls. For
details of subcohort definitions see OAI Online at http://www.oai.ucsf.edu/. The following two tables show the age,
sex and race distributions at enrollment for the whole OAI Study cohort of 4796 participants:
Distribution of Race by Sex
White or Caucasian Non-White Unknown Total

Male 1666 323 3 1992
Female 2124 678 2 2804
Total 3790 1001 5 4796
Distribution of Age by Sex
Age (years)
45 to 49 50 to 59 60 to 69 70 to 79 Total
Male 259 739 495 499 1992
Female 290 923 968 623 2804
Total 549 1662 1463 1122 4796
If you requested one of the smaller image releases containing a subset of the whole cohort, Appendices A thru G give
similar demographic data for each special subset of participants. If you requested a list of OAI IDs which you generated
yourself, you will have to determine the demographics of those participants from the OAI Clinical datasets available
online (see next section 1.2).
1.2 Participant Clinical, Measurement and Biomarker Data

Questionnaire and clinic measurement data are available in SAS datasets at the OAI Online website
http://www.oai.ucsf.edu/ . To obtain the most up-to-date clinical data on these participants, you will need to retrieve
the latest version of the clinical datasets. The ENROLLEES SAS datasets contains variables that let you
86
Appendix
easily determine whether a participant has knee MRI and/or knee x-rays available at specific visits. For a specific
visit, the sum of two variables (or value or one variable) is used. The following table explains their use:
VISIT Variables to use to check for No Knee Knee MRI Knee x-rays Knee MRI &
Knee MRI and X-rays Images Knee x-rays
Baseline V00IMAGESC(a)+V00IMAGESE 0 1 2 3
12-month V01IMAGESC(a)+V01IMAGESE 0 1 2 3
18-month V02IMAGESD 0 1 n/a(b) n/a (b)
24-month V03IMAGESC(a) +V03IMAGESE 0 1 2 3
30-month V04IMAGESG 0 1 n/a(b) n/a(b)
36-month V05IMAGESC +V05IMAGESE 0 1 2 3

48-month V06IMAGESC + V06IMAGESE 0 1 2 3
(a) Non-zero value for variables V00/V01/V03IMAGESC means that images for a participant were released in an early
image release of the specific visit prior to the release of images for the whole cohort
(b) MRI images only are acquired in subsets of participants at 18-month and 30-month interim visits.
Variables associated with contacts prior to enrollment are identified by a prefix starting with P (e.g., P01 variables are
from the screening visit where the baseline fixed flexion knee radiographs are acquired). Variables associated with
enrollment visit are prefixed V00, variables at 12-month follow-up are prefixed V01, variables at 18-month interim visit
are prefixed V02, variables at 24-month follow-up are prefixed V03, variables at 30-month follow-up are prefixed V04,
variables at 36-month follow-up are prefixed V05, variables at 48-month follow-up visit are prefixed V06, and
variables at 72-month follow-up visit are prefixed V08.
To determine exactly which MRI pulse sequences are available in which knee(s) for participants at specific visits,
and to determine the availability of all types of x-ray images meta-data can be downloaded from OAI Online
(http://www.oai.ucsf.edu)
For each visit, meta-data describing the type, acquisition, and quality of the images in this data release can be found
in a pair of SAS datasets, XRAYxx and MRIxx, where xx is 00 for baseline, 01 for the 12-month, 02 for the 18-month
visit, 03 for the 24-month visit, 04 for the 30-month visit, 05 for the 36-month visit, 06 for the 48-month visit, and 08 for
the 72-month visit. For example, the dataset XRAY00 contains the x-ray meta-data from baseline and XRAY01
contains the x-ray meta-data from the 12-month follow-up visit. Please note the x-ray and MRI datasets have multiple
records per participant. For example, the dataset XRAY00 contains one record per baseline radiograph expected or
released for each of the participants in this image release. The dataset MRI00 contains one record per knee per
baseline standard OAI MRI pulse sequence (DICOM series), excluding MRI localizer sequences, for each of the
participants (section 2 of this document outlines the types of x-ray and MRI images that can be available for each
knee). Further information can be found in the zipped documentation that accompanies each SAS dataset.
The imaging meta-data datasets MRIxx and XRAYxx will contain flag variables (VxxXQCFLAG for x-ray
VxxMQCFLAG for MRI) to indicate images that have been acquired with various image acquisition problems, and
the variables VxxXQCCMNT for x-ray, and VxxMQCCMNT for MRI will be a short text description of the problem.
The MR flag and comment variables (V08MQCFLAG and V08MQCCMNT respectively) are available at the 72-
month visit. For visits prior to 72-month, these variables currently do not exist, but will be added in future data
releases.
Further details of SAS datasets can be found in the document Overview of Structure, Use, and Conventions for OAI
data at: http://www.oai.ucsf.edu/datarelease/DataClinical.asp and descriptions of the clinical and imaging data
version conventions can be found in the document OAI Data Release Version Conventions at:
http://www.oai.ucsf.edu/datarelease/docs/DataReleaseVersionConventions.pdf
87
Appendix
DICOM Image Release Notes February 21, 2013

2. IMAGES AVAILABLE
Depending on your request, the hard drive supplied may contain various combinations (or subsets) of the
following:
Image Release 0.E.1 - if you requested baseline visit (screening & enrollment visit) images Image
Release 1.E.1 - if you requested 12-month follow-up visit images
Image Release 2.D.1 - if you requested 18-month interim visit MRI images
Image Release 3.E.1 - if you requested 24-month follow-up visit images
Image Release 4.G.1 - if you requested 30-month interim visit MRI images
The visit identifier (the first digit of the image release identifier) matches the visit designator number described for
clinical data variables in the previous section (eg: 0 = baseline, 1 = 12-month visit, 2 = 18-month visit, 3 = 24-month
visit, 4 = 30-month visit, 5 = 36-month visit, 6= 48-month visit, 8= 72-month visit).
The letter designation represents the specific participants for whom images are provided:
E represents the Entire OAI cohort of 4796 participants. Other
letter designations represent special subsets of the entire cohort:
C means that images for that specific visit are not yet available for the entire cohort, but only for a
subset of 2686 participants (Group C) recruited early in the study (see Appendix C for their details). D
represents the 287 participants with MRI scans obtained at an 18-month interim visit (Group D). (See
Appendix D for their details.)
G represents the 483 participants with MRI scans obtained at a 30-month interim visit (Group G). (See
Appendix G for their details.)
Appendices A and B provide information about further subgroups of participants who were included in early
image releases.
If you provided a specific list of participants for whom you required images, and you find that there are some
participants from your list without images on the hard drive, then this means that the particular participant(s) did not
have images acquired at the relevant visit. We encourage you to analyze the image meta-data XRAYxx and MRIxx
SAS datasets to determine which participants had images of different types acquired.
Please note that in the future, these image releases may be updated with additional images, updated images, or
deleted images. Later releases will be identified by a higher number in the final digit of the Image Release Version. A
list of updates will be kept on the OAI Online website in the Changes to previously released images link at
http://www.oai.ucsf.edu/datarelease/DataImaging.asp
Typical examples of images available can be found online at:

http://www.oai.ucsf.edu/datarelease/ExampleImages.asp
Manuals describing all image acquisition procedures and parameters can be found online at:
http://www.oai.ucsf.edu/datarelease/operationsmanuals.asp
2.1 Image Identifiers (Barcodes)

Every x-ray image has a unique 12 digit barcode associated with it and the first 5 digits are always 01660. These
barcodes are used to identify x-ray images in the xray meta-data datasets (variables V00XRBARCD,
V01XRBARCD, V03XRBARCD, V05XRBARCD, V06XRBARCD, and V08XRBARCD for baseline, 12-month, 24-
month, 36-month, 48-month, and 72-month images in the XRAY00, XRAY01, XRAY03, XRAY05, XRAY06, and
XRAY08 SAS datasets, respectively).
Each MRI sequence has a unique 12 digit barcode associated with it and the first 5 digits are always 01661. These
barcodes are used to identify MRI sequences (DICOM series) in the MRI meta-data datasets (variables
V00MRBARCD, V01MRBARCD, V02MRBARCD, V03MRBARCD, V04MRBARCD, V05MRBARCD, V06MRBARCD,
and V08MRBARCD for baseline, 12-month, 18-month, 24-month images, 30-month, 36-month,
88
Appendix

48-month, and 72-month images in the MRI00, MRI01, MRI02, MRI03, MRI04, MRI05, MRI06, and MRI08 SAS
datasets, respectively).
Please note that there are long (12 digit) and short (8 digit) versions of the barcodes. The 8-digit versions are used
for folder names on the hard drive, and can be converted to the 12-digit version by prefixing them with 0166. If you
import barcode data into spreadsheet or database programs, we recommend that you ensure that any leading
zeros in the data are not lost.
2.2 Radiographs (Participants Eligible for Imaging, Visit When Image Acquired)
Baseline Images:
Bilateral PA fixed flexion knees (all participants, Screening visit) AP
Pelvis (all participants, Enrollment visit)
PA Hand (all participants, Enrollment visit; unilateral dominant hand in 3 of 5 clinical centers, bilateral hand
acquired at 2 of 5 clinical centers)
Lateral knee radiographs (both knees, non-exposed controls only, Enrollment visit)
Follow-up Images:
Bilateral PA fixed flexion knees (all participants, annual follow-up visits)
AP Pelvis (48-month visit, and also 12-month visit if not acquired at baseline)
PA Hand (48-month visit, and also 12-month visit, if not acquired at baseline; unilateral dominant hand in 3
of 5 clinical centers, bilateral hand acquired at 2 of 5 clinical centers)
Lateral knee radiographs (both knees, non-exposed controls only, annual follow-up visits) AP
full limb (progression and incidence subcohort, at one of the annual follow-up visits)
2.2.1 Fixed Flexion Knee Radiography and the Synaflexer Frame

This device is designed to standardize positioning and allow a check of the geometry used during fixed flexion knee
radiography. The phantom consists of two parallel rows of metal beads of known separations encased in a plexiglass
frame of known dimensions. On the radiographs, these metal beads appear in two vertical lines:
http://www.oai.ucsf.edu/datarelease/ExampleParticipantImages/JPEG/Img2-0-0-0.html
In an optimally acquired OAI fixed flexion radiograph, these lines will be almost exactly parallel, and the spacing
between the beads for the line on the left of the image will be slightly greater than for the line on the right side of the
image. Within each line of metal beads, the beads will have almost exactly equal spacings. The row of beads with the
largest spacing is offset 1 (44.5mm) posteriorly from the point on which the patella presses against the frame
during the exposure, and is hence in approximately the same coronal plane as the center of the tibio-femoral joint.
The other row of beads is offset posteriorly from the point at which the patella presses against the frame. The
actual vertical spacing of the beads is 1 (25.4mm), so by measuring the inter-bead spacing on the row of beads on
the left side of the image, the conversion from image pixels to true distance at the joint line can be performed.
2.2.2 Full-Limb X-rays

Radiographs of the entire lower extremities are taken together in an upright weight-bearing position, and the images
include complete visualization of the femoral head, the knee and talus of the foot. The objective of this examination
is to measure knee alignment. Both extremities are usually on one image, but for participants with extreme
malalignment, or for large participants, there may be more than one image per participant, and each limb may be on
a separate image.
2.3 Knee MRI (All Participants, Enrollment visit and annual follow-up visits)
The sequences typically acquired for both left and right knees are:
Localizer (3-Plane) (MP_LOCATOR)
Coronal Intermediate Weighted Turbo Spin Echo (COR_IW_TSE)
Sagittal 3D DESS with Water Excitation (SAG_3D_DESS)
Sagittal Intermediate Weighted Turbo Spin Echo with Fat Suppression (SAG_IW_TSE_FS)
The right knee typically also has two additional sequences acquired:
Coronal T1-weighted 3D Flash with Water Excitation (COR_T1W_3D_FLASH_WE)
89
Appendix

Sagittal T2 Mapping Sequence with a small field of view (SAG_T2_MAP)
If the right knee is not being imaged, then these two additional sequences may be acquired for the left knee.
Multi-planar reformattings (MPR) were also performed from the SAG_3D_DESS images into:
Coronal Plane MPR (COR_MPR_SAG_3D_DESS)
Axial Plane MPR (AXIAL_MPR_SAG_3D_DESS)
For participants with a parametric T2 map calculated using the Siemens syngo MapIt software, there will be a set
of images labeled SAG_T2_CALC to match the SAG_T2_MAP images from which they were calculated. See
Section 2.3.1 for details about participants for whom these parametric images are available.
The parameters used during the acquisition of each pulse sequence are given in Appendix X at the end of this
document. The abbreviations in parentheses after each pulse sequence above (e.g.: COR_IW_TSE) are those used
to describe the sequences in the DICOM image headers, and can usually be used in viewing software to identify
images for loading or viewing.
Examples of assessments that can be performed using the OAI MR examination are:
1. Coronal IW TSE (COR_IW_TSE) enables assessment of the medial collateral ligament (MCL), lateral
collateral ligament (LCL), osteophytes and cysts (medial - lateral central femur and central tibia), sclerosis
(central femur and tibia), and the meniscal body.
2. Sagittal 3D DESS with water excitation (SAG_3D_DESS) enables quantification of cartilage volume over the
entire knee (patellofemoral and femorotibial joints). Another primary use of the 3D DESS acquisition is to
identify osteophytes in both the original sagittal (superior-inferior patella, anterior-posterior femur and tibia) as
well as in the coronal (medial / lateral femur and tibia) and axial (medial-lateral patella) MPR. Secondarily, it
also potentially provides assessment of subarticular marrow edema and cysts both in the original sagittal
plane as well as in the coronal (central femur and tibia) and axial (patella) MPR. This latter marrow
assessment does not have proven sensitivity and specificity, but is presumed to be less sensitive than a fat
suppressed IW or T2W.
3. Sagittal IW TSE with fat suppression (SAG_IW_TSE_FS) enables identification of subarticular marrow
edema and cysts as well as quantitation of the joint effusion. The large (20cm) imaging field of view (FOV)
covers the suprapatellar bursae as well as dissecting popliteal cysts. Additional assessments enabled by
this acquisition include cartilage quality (signal hetereogeneity / T2 lesion), marrow edema and cysts,
anterior cruciate ligament (ACL), posterior cruciate ligament (PCL), as well as osteophytes (superior
inferior patella, anterior posterior femur and tibia).
4. Coronal T1W 3D FLASH with water excitation (COR_T1W_3D_FLASH_WE) enables quantitation of cartilage
volume over the central load-bearing compartment of the knee (femorotibial joint). Another primary use of the
3D FL acquisition is to identify medial / lateral osteophytes on the femur and tibia in the original coronal
plane. Secondarily, it also potentially provides assessment of subarticular marrow edema and cysts in the
coronal plane (central femur and tibia). This latter marrow assessment does not have proven sensitivity and
specificity, but is presumed to be less sensitive than a fat suppressed IW or T2W.
5. Sagittal T2 map (SAG_T2_MAP) is a 7 echo sequence (every 10msec) acquired using a 12cm imaging
FOV. The resulting image contrasts include PD, IW, and T2W. These images enable assessment of
subchondral bone (PD, T2W) for sclerosis, cysts and edema, the meniscal horns (PD), and for cartilage
morphology and quality (PD, IW and T2W).
2.3.1 Images of Calculated T2 Values

During the imaging period of the 72-month visit, the software of the OAI was upgraded to include Siemens syngo
MapIt software (https://healthcare.siemens.com/magnetic-resonance-imaging/options-and-upgrades/clinical-
applications/syngo-mapit) which allows calculating of parametric maps of T2 values from the multiecho
SAG_T2_MAP images. For each site, there is a specific date on which this occurred:
Site A: SAG_T2_MAP scans from 12/21/2011 will have a matching set of parametric T2 images Site
B: SAG_T2_MAP scans from 6/3/2010 will have a matching set of parametric T2 images
90
Appendix

Site C: SAG_T2_MAP scans from 7/12/2010 will have a matching set of parametric T2 images Site
D: SAG_T2_MAP scans from 5/19/2010 will have a matching set of parametric T2 images Site E:
SAG_T2_MAP scans from 12/20/2011 will have a matching set of parametric T2 images
2.4 Thigh MRI (All Participants, Baseline or 12-month visit, with follow-up at 24-month visit or 36-
month visit, and 48-month visit)
Thigh MRI scans were obtained on the majority of participants at baseline (enrollment visit), and if not obtained at
that visit, an additional attempt was made to acquire one at the 12-month follow-up visit. Images were then also
acquired at the 24-month annual visit, and if not obtained at that visit, an additional attempt was made to acquire one
at the 36-month follow-up visit. Thigh MRIs were also acquired at the 48-month follow-up visit. The images available
comprise axial T1-weighted images with 15 contiguous slices 3.0mm thick with the most distal slice being 10cm
superior to the right medial femoral epiphysis. These images are labeled AX_T1_THIGH. There are also localizer
sequences (labeled MP_LOCATOR_THIGH) and in the majority of participants, also an additional image (labeled
PRESCRIPTION_THIGH) which shows the levels of the axial slices overlaid on the localizer sequence.
For further details of the thigh acquisition protocol, see the MRI Operations manual available on OAI online at:
http://www.oai.ucsf.edu/datarelease/operationsmanuals.asp
2.5 Rules for Release of Acquired Images

Rules for which images are released (for example, when more than one instance of a radiograph or MRI series are
acquired at a given visit - baseline, 12-month follow-up visit, etc.) are described below:
2.5.1 Radiographs
a. Images with pending radiograph-related queries will not be included in a release.
b. Only one image (acquisition) of a radiograph of each type from a given visit will be released.
b.1 Usually, this will be an "accepted" instance (i.e., either an image that is of acceptable quality based on
central QA review or one that was not selected for central QA review and was accepted by default without
QA review). In cases where there is more than one "accepted" instance of a type from a given visit, the
central imaging QA center will identify the instance of the best overall quality for release.
b.2 Unacceptable quality (based on central QA review) images will not be released, except when there is
no "accepted" instance available, in which case an unacceptable quality image of that type will be released.
When no "accepted" instance is available and more than one unacceptable quality image exists, the central
imaging QA center will identify the best overall quality unacceptable image for release.
c. An exception to the one released instance rule occurs when more than one acquisition was needed to cover
all of the required anatomy, such as occurs for the bilateral screening knee in participants with extreme
varus deformity. In this case, each of two complementary images has its own unique barcode, and both will
be released.
2.5.2 Knee MRI
a. Images with pending MRI-related queries will not be included in a release.
b. All unique and "accepted" instances (i.e., either a series of acceptable quality based on central QA review
or one that was not selected for central QA review and was accepted by default without QA review) of
an MRI series will be included in a release.
b.1 Only one copy of each unique instance of an MRI series will be released (i.e., exact duplicates of the
same instance of a series will be deleted).
91
Appendix
b.2 Unacceptable quality (based on central QA review) instances of a series will not be released, except
when there is no "accepted" instance of that series. In that case, all unacceptable quality instances of that
series will be released. In such cases, careful examination of the available images should be undertaken to
determine which one should be analyzed, keeping in mind that sometimes both may be sub-optimal and
need to be analyzed together. The central imaging QA center will attempt to exclude partial or aborted series
from a release when a complete series is available. This rule can be broken where there is a Sagittal DESS
acquired with the scanner gradients out of specification on one date, but a repeated Sagittal DESS of the
same knee acquired with the scanner gradients in specification on a later date (since the earlier scan is more
appropriate for visual inspection and the later for quantitative measurements). In such cases, the earlier scan
may have an unacceptable rating and the later scan an acceptable rating.
2.6 Future Updates

At some point in the future, additional images are likely to become available for baseline and follow-up visits of
participants in this image release. This will lead to updated releases that will contain those additional images, along
with the images currently in this release. There may also be some images that will require deletion or need to be
updated.
Please check the imaging page at OAI Online http://www.oai.ucsf.edu/datarelease/DataImaging.asp regularly for
possible updates.
3. DISK CONTENTS
The top level folder of this hard disk contains 3 files and a set of subfolders. The first 4 items in the list below will
always be present, and then various combinations of other items will be on the hard drive, depending on which visit(s)
and group of participants you requested images for:
DicomImageReleaseNotes.pdf:
This is the file that you are looking at now, in PDF format
OAIDataUserAgreementImages.pdf:
Before using these images a copy of this document should be reviewed, printed, signed, and sent to the OAI
Coordinating Center at the University of California at San Francisco.
Contents.csv
This text file lists the folders on the hard drive, along with ID of the participant for whom they belong, the date
that the images were acquired, the unique barcode identifying the images, along with the values of the DICOM
Study and Series descriptions. This data is extracted from the image DICOM header tags.
VIEWERS:
A folder containing software for viewing the images. All of the software may be used at no cost, but a
donation is asked for one of the programs. Details about the software are given later in this document.
0.[X].[n]:
Folder of baseline visit images containing image data stored in DICOM format, containing files in the
following manner:
Image folders (numeric names of 7 digits) contain DICOM image files for a particular participant. Images
for each participant are contained in the folder named for their OAI Release Participant ID #.
1.[X].[n]:
Folder of 12 month visit images containing image data stored in DICOM format, containing files in the
following manner:
92
Appendix

3.[X].[n]:
following manner:
5.[X].[n]:
following manner:
6.[X].[n]:
following manner:
8.[X].[n]:
following manner:
A value of C for letter [X] indicates that the images are from participants are in the initial 2686 participants in
Group C of the OAI cohort, and a value of E represents the remaining 2110 participants from the cohort.
The value of [n] represents the version number of the release, where .1 represents the first release, and values of
.2, .3, indicate later releases. Other letters represent the subsets described earlier (Groups A,B,D,G).
If you requested 18-month interim visit MRIs, then they will be in the following folder:
2.D.[n]:
Folder of 18-month images containing image data stored in DICOM format, containing files in the
following manner:
If you requested 30-month interim visit MRIs, then they will be in the following folder:
4.G.[n]:
Folder of 30-month images containing image data stored in DICOM format, containing files in the
following manner:
3.1 About the Files in Each Participants Folder

Each individual participants folder contains subfolders for each date on which a participant had images (format of
folder name is yyymmdd).
Each date subfolder contains folders identifying the unique barcode for the images contained within it. To get the
12 digit barcode, add 0166 to the start of the 8 digit folder name. Folders containing x-rays have names starting
with 0 and folders containing MRI have names starting with 1.
Earlier OAI image releases contained a small file in each participants folder which listed the images available, but
that data is now contained in the file contents.csv in the top level folder of the hard drive. The data in that file includes
the folder name, the participants ID, Study Date, Barcode Folder, Study Description and Series Description.
93
Appendix

3.2 File and Folder Structure
The folder name is simply the numeric ID of the participant in the study. Each participant's folder contains several
sub-folders named for the dates on which the participant underwent imaging. For example in folder 0.C.2/9007827,
participant 9007827 has sub-folders called 20040819 and 20041006 because they were imaged on August 19, 2004
and October 6, 2004 as part of the baseline visit schedule. There is one more level of sub-folders below this level:
one sub-folder for each image series acquired on that date. These sub-folders have unique 8-digit identifiers that are
assigned to the image series in the central OAI imaging database maintained at Synarc, Inc. If the 8-digit identifier
begins with 0 then the folder contains x-ray images, and if it starts with 1, then the folder contains MR images. (See
section 2.1 and 3.1 for more information about these identifiers.)
The image files are then contained within these subfolders. There is only one radiographic image per series, so each
radiographic image is named "001" and is the only file in the folder. The Magnetic Resonance (MR) series generally
comprise many images ("slices"), and have names like "001", "002", up to the number of images in the MR series.
4. DICOM INFORMATION IN IMAGES
The images have correctly specified DICOM Unique Identifiers (UIDs) for the correct relationships between
Participant ID, Study, Series, and Image UIDs, and therefore it should be possible to import them into PACs
systems or other DICOM viewing software that requires conformance with DICOM. The data have been tested
using Conquest DICOM server software (http://www.xs4all.nl/~ingenium/dicom.html ) and also with eFilm
Workstation software (http://www.efilm.ca), but should also work properly with other DICOM software.
In addition, to correctly specify DICOM UIDS in the images, we have attempted to provide a uniform set of additional
DICOM tags in the image files on this disk. Here we describe the values you can expect to find in the image file
headers.
4.1 Radiograph DICOM Tags

The following table lists the format of the specified DICOM header tags in x-ray images. When (unchanged if
digital) is listed, then the additional information in the table entry is specifically for images acquired on film and then
subsequently digitized.
DICOM Header Tags for X-ray Data
DICOM Tag DICOM Tag Name Comments
Number
(0008,0020) StudyDate (unchanged if digital)
[exam date from TF] e.g. 20051001
(0008,0030) StudyTime (unchanged if digital)
000000
(0008,0050) AccessionNumber Synarc Barcode, e.g. 016601234567
(0008,0060) Modality (unchanged if digital DX or CR)
RG = acquired on film and then digitized
CR = computed radiography
DX = digital radiography
(0008,0090) ReferringPhysiciansName Empty
(0008,1010) StationName Empty
(0008,1030) StudyDescription Bilateral PA Fixed Flexion Knee:
OAI^XRAY^[visit name]^KNEE
AP Pelvis:
OAI^XRAY^[visit name]^PELVIS
Full Limb:
OAI^XRAY^[visit name]^FULL LIMB
Unilateral PA Fixed Flexion Knee:
OAI^XRAY^[visit name]^KNEE LEFT
OAI^XRAY^[visit name]^KNEE RIGHT
94
Appendix

Number
PA Bilateral Hand:
OAI^XRAY^[visit name]^HAND
PA Unilateral Hand:
OAI^XRAY^[visit name]^HAND LEFT
OAI^XRAY^[visit name]^HAND RIGHT
Lateral Knee:
OAI^XRAY^[visit name]^LAT KNEE LEFT
OAI^XRAY^[visit name]^LAT KNEE RIGHT
[visit name] is either:
SCREENING VISIT, ENROLLMENT VISIT,
12 MONTH VISIT, 24 MONTH VISIT , 36 MONTH VISIT,
48 MONTH VISIT, 72 MONTH VISIT
(0008,103E) SeriesDescription Bilateral PA Fixed Flexion Knee

AP Pelvis
Full Limb
PA Fixed Flexion Right Knee
PA Fixed Flexion Left Knee
PA Bilateral Hand
PA Right Hand
PA Left Hand
Lateral Right Knee
Lateral Left Knee
(0010,0010) PatientsName [ParticipantID prefixed with OAI] e.g. OAI9007827

(0010,0020) PatientID [ParticipantID, no prefix] e.g. 9007827
(0010,0030) PatientsBirthDate Empty
(0010,0040) PatientsSex O (DICOM for other)
(0010,1010) PatientsAge Empty

(0012,0050) ClinicalTrialTimePointID Screening P01, Enrollment V00
12 Month V01, 24 Month V03
36 Month V05, 48 Month V06, 72 Month V08
(0012,0051) ClinicalTrialTimePointDescription Screening Visit
Enrollment Visit (Baseline)
12 month Annual Visit
(0018,0015) BodyPartExamined KNEE, PELVIS, HAND, LEG
(0020,0010) StudyID (both film and digital)
Bilateral PA Fixed Flexion Knee 1
AP Pelvis 3
Full Limb 6
PA Fixed Flexion Right Knee 7
PA Fixed Flexion Left Knee 8
PA Bilateral Hand 2
PA Right Hand 10
PA Left Hand 11
Lateral Right Knee 7
Lateral Left Knee 8
(0020,0011) SeriesNumber (unchanged if digital)
1
All private DICOM tags have been removed from x-ray images, as have any DICOM overlays containing
unwanted information.
4.1.1 Note about pixel spacings in radiographs acquired and stored digitally:
95
Appendix

For radiographs acquired and stored digitally (Modality CR or DX stored in DICOM header), the pixel sizes
stored in the DICOM tag for PixelSpacing (0028,0030) and ImagerPixelSpacing (0018,1164) are the values written
by the acquisition equipment. It is important to note that the values specified may only be approximate. The pixel
size is always square with sides varying from 0.100mm to 0.200mm and depends on the acquisition site and type
of radiographic exam, although full-limb radiographs will have larger pixel sizes. There are two situations where
PixelSpacing and ImagerPixelSpacing may contain different values.
The first situation is when the precision used to write the value into the fields is slightly different an example of this
would be PixelSpacing (mm) containing 0.189\0.189 and ImagerPixelSpacing (mm) containing
0.188895\0.188895. In these situations, the values are effectively the same, but we suggest that you use the value
with the highest precision (0.188895 mm in this case).
The second situation occurs when the x-ray equipment has attempted to estimate both source to detector distance
and source to patient distance, in which case ImagerPixelSpacing will be larger than PixelSpacing. An example of
this would be if PixelSpacing (0028,0030) is 015796\0.15796 and ImagerPixelSpacing (0018,1164) is
0.168960\0.168960, and in these cases, the value for PixelSpacing has usually been estimated by the equipment
using the values from the DICOM header tags (0018,1110) DistanceSourceToDetector, DistanceSourceToPatient
(0018,1111) in combination with ImagerPixelSpacing. It is important to note that since DistanceSourceToPatient is
not well known and is only approximate, the value for pixel spacing does not necessarily relate to the actual size of
the anatomy being imaged.
For x-ray images in which DistanceSourceToDetector and DistanceSourceToPatient are recorded, the following
equation should relate those values with ImagerPixelSpacing and PixelSpacing tags:
[ImagerPixelSpacing] [DistanceSourceToDetector] [PixelSpacing]
[DistanceSourceToPatient]
Equipment manufacturers may or may not complete all 4 fields correctly, but the above equation shows the
relationship between the 4 values. Please note, that for most x-rays, no information is provided about the
distances from the source to either detector or patients and imager pixel spacing and pixel spacing contain
effectively the same information the spacing between the centers of each pixel at the detector plane.
It is important to note that pixel sizes for the anatomy being imaged are always difficult to estimate due to
magnification effects related to spacings between the relevant anatomy, the x-ray source and the position of the
detector. We encourage you to use the Synaflexer Frame described in Section 2.2.1 for cross calibration of fixed
flexion knee radiographs between visits and between participants.
Important Note: In particular, the distance between the metal beads should always measure to be larger than 1
(25.4mm). For some CR and DX radiographs, manufacturers put estimated values for pixel spacing, but those are
not calibrated in anyway, and may assume that the knee is much further from the detector than it actually is. In
those cases, incorrect use of the DICOM Tag PixelSpacing, rather than ImagerPixelSpacing may suggest that the
metal beads are closer than 1 apart.
4.1.2 Note about pixel spacings in radiographs acquired on film and then digitized:
For radiographs digitized from film (Modality RG stored in DICOM header), the 3 different film digitizers used were
set to a nominal pixel size of 0.100mm, but each of the digitizers used had the actual pixel size at that setting
estimated. The value stored in the DICOM Tags PixelSpacing (0028,0030) and ImagerPixelSpacing (0018,1164) in an
image is set to that actual pixel size for the digitizer used. For such OAI images the values are either 0.100\0.100,
0.101\0.101, or 0.102\0.102 mm. The same caveats mentioned in the previous section, regarding pixel spacings
and the separations between anatomy, film and x-ray source affecting measurements taken from x-rays, also apply to
images digitized from film.
4.1.3 Note about pixel spacings in radiographs acquired digitally, then printed on film:
A small subset of the X-ray images for participants from site E, which were obtained from approximately 04/01/2005
thru 03/15/2006, were acquired digitally, printed on film, and then digitized. The pixel spacings stored in the DICOM
headers represent the values from the digitization of the film. A total of about 140 images
96
Appendix

are of this type. These may get replaced by their CR image in future image releases. Once again, for fixed
flexion knee radiographs, the Synaflexer frame can be used to calibrate such images.
4.2 MRI DICOM Tags

Table 2: DICOM Header Tags for MRI Data
Number
(0008,0020) StudyDate Unchanged
(0008,0030) StudyTime Unchanged
(0008,0050) AccessionNumber Synarc Barcode, e.g. 016601234567
(0008,0090) ReferringPhysiciansName Empty
(0008,1030) StudyDescription OAI^MR^[visit name]^LEFT
OAI^MR^[visit name]^RIGHT
OAI^MR^[visit name]^THIGH
[visit name] is ENROLLMENT VISIT, 12 MONTH VISIT, 18
MONTH VISIT, 24 MONTH VISIT, 30 MONTH VISIT , or
36 MONTH VISIT, 48 MONTH VISIT, 72 MONTH VISIT
(0008,103E) SeriesDescription SAG_3D_DESS_RIGHT
COR_T1_3D_FLASH_RIGHT
SAG_T2_MAP_RIGHT
SAG_T2_CALC_RIGHT
SAG_IW_TSE_RIGHT
COR_IW_TSE_RIGHT
COR_MPR_RIGHT
AX_MPR_RIGHT
SAG_3D_DESS_LEFT
COR_T1_3D_FLASH_LEFT
SAG_T2_MAP_LEFT
SAG_T2_CALC_LEFT
SAG_IW_TSE_LEFT
COR_IW_TSE_LEFT
COR_MPR_LEFT
AX_MPR_LEFT
MP_LOCATOR_RIGHT
MP_LOCATOR_LEFT
AX_T1_THIGH
PRESCRIPTION_THIGH
MP_LOCATOR_THIGH
(0008,1050) PerformingPhysiciansName Empty
(0008,1060) NameOfPhysiciansReadingStudy Empty
(0010,0010) PatientsName [ParticipantID prefixed with OAI] e.g. OAI9007827

(0010,0020) PatientID [ParticipantID, no prefix] e.g. 9007827
(0010,0030) PatientsBirthDate Empty
(0010,0040) PatientsSex O (DICOM for other)
(0010,1010) PatientsAge Empty

(0012,0050) ClinicalTrialTimePointID Enrollment V00, 12 Month V01, 18 Month V02,
24 Month V03, 30 Month V04, 36 Month V05
48 Month V06, 72 Month V08
(0012,0051) ClinicalTrialTimePointDescription Enrollment Visit (Baseline), 12 month Annual Visit, 18 month
Interim Visit, 24 month Annual Visit, 30 month Interim Visit,
36 month Annual Visit, 48 month Annual Visit, 72 month Annual
Visit
(0018,0015) BodyPartExamined Unchanged
(0018,1030) ProtocolName Unchanged
(0020,0010) StudyID Unchanged
(0020,0011) SeriesNumber Unchanged
97
Appendix

Number
(0020,0060) Laterality Unchanged
All private tags, including those added during central image QA by Synarc have been cleared from the image files.
Manufacturer-defined (Siemens) private tags have been retained, since their contents are known and well defined.
5. DICOM VIEWING SOFTWARE AND RESOURCES
The Official DICOM website is:

http://medical.nema.org/
and there are many other useful websites, including the following:
http://www.dclunie.com/dicom-status/status.html
http://www.sph.sc.edu/comd/rorden/dicom.html
http://www.dcmtk.org/
5.1 Freely Available DICOM Image Viewing Software

A number of image viewers are available, some of which may be used for a trial period or with a limited set of
features, some which ask for a donation to be made, and some that are free altogether.
5.1.1 ImageJ
ImageJ is a program that is distributed by the National Institutes of Health, and is in the public domain. A recent
version is included on this disk. The latest version may be obtained from:
http://rsb.info.nih.gov/ij/
Written in the Java language, ImageJ runs well on the most popular computer operating systems, including
Microsoft Windows, Apple's MacOSX and Linux.
By itself, ImageJ can open the DICOM image files on this disk. With the addition of a free plugin from the Institut fuer
Telematik in der Medizin, you can use ImageJ to browse the disk. This plugin may be obtained from this address:
http://www.iftm.de/telemedizin/dcmimex.htm
As with ImageJ itself, this plugin works well under Microsoft Windows, Apple's MacOSX, and Linux. It requires
version 1.4 of the Java runtime.
5.1.2 Other DICOM viewers

The following are worthy of mention:
For MS Windows, there is DicomWorks by Philippe Puech and Loic Boussel:
http://www.dicomworks.com
and Osiris from Hopitaux Universitaires de Geneve:
http://www.sim.hcuge.ch/uin
For MacOSX version 10.3 or later, there is OsiriX, and also a version of Osiris for Mac OS 9, both available at:
http://www.sim.hcuge.ch/uin
This viewing software runs under java and should run any computer with java installed:
http://www.visus-tt.com/download/dicom-viewer.html
There is also an open source, freely available DICOM workstation and viewing package available at this site:
http://www.clearcanvas.ca
which runs on MS Windows.
5.2 Viewing Software Installation
5.2.1 DicomWorks (MS Windows NT, 2000, XP)

Quit all open applications.
Launch the file dicomworks_135b.exe in the VIEWERS folder.
Follow the instructions given by the installation program.
98
Appendix

To register, follow this link in your web browser: http://www.dicomworks.com and follow the instructions.
Registration is not required if you only wish to view the images on this DVD. But it will enable you to use features
of the program that are otherwise not available. If you find the program useful, consider sending the author a
donation. The suggested amount is $25. See the web site for details.
5.2.2 ImageJ with DICOM Im/Ex plugin

The DICOM Im/Ex plugin requires Java 1.4 or newer to be installed on your computer. Before it can be installed, you
must install ImageJ on your computer.
Detailed installation instructions for ImageJ can be found here:

http://rsb.info.nih.gov/ij/docs/install/
To obtain the most recent version of ImageJ, look here:
http://rsb.info.nih.gov/ij/download.html
ImageJ for Windows:

Launch the file ij133-jdk15-setup.exe found in the VIEWERS folder.
Select the destination for the program. If you already have another version of ImageJ installed, you may wish
to override the default. In any case, make a note of the location you select, as you will need this information
to install the DICOM Im/Ex plugin.
Complete the installation by answering the other questions asked by the installer program. Open
the file dcmie.zip found in the VIEWERS folder.
Extract the files to the directory where you chose to install ImageJ.
Move the folder Dcm to the plugins directory.
Quit ImageJ if it is running. Then launch ImageJ.
Look under the Plugins menu, and verify that the Dcm sub-menu now appears.
We suggest that you edit a file Run_ImageJ.bat in C:\Program Files\ImageJ (or whichever folder ImageJ is
installed into) to contain:
java.exe -Xms384m -Xmx384m -classpath .\lib\dcmie.jar;ij.jar;.\plugins ij.ImageJ
where 384m represents about of the physical memory on your PC (e.g., use 384m if you have 512MB of
memory, or 96m if you have 128MB of memory).
ImageJ for MacOS X:

Verify that you have Java version 1.4 or newer installed. You can do this from the Terminal application by
typing:
"java -version" (followed by Enter) at the prompt.
Double-click the file ij133.tar.gz to extract the "ImageJ (OS X)" folder.
Copy the ImageJ (OS X) folder to the /Applications directory.
Open the file dcmie.zip found in the VIEWERS folder.
Extract the files to the directory /Applications/ImageJ.
Move the folder Dcm to the plugins directory.
Quit ImageJ if it is running.
From the terminal window, launch ImageJ with the commands:
cd /Applications/ImageJ
java -Xms96m -Xmx96m -cp ./lib/dcmie.jar:ij.jar ij.ImageJ
Look under the Plugins menu, and verify that the Dcm sub-menu now appears.
You may wish to create an AppleScript to avoid typing this command. The syntax of the script is: tell
application "Terminal"
run
do script with command "cd /Applications/ImageJ
java -Xms96m -Xmx96m -cp ./lib/dcmie.jar:ij.jar ij.ImageJ"
end tell
We suggest that you change 96m to represent about of the physical memory on your computer (eg: 384m if your
computer has 512MB of memory).
99
Appendix

ImageJ for Linux (i386):
Change to the VIEWERS folder on the hard disk
Copy ij133-x86.tar.gz and dcmie.zip to a temporary place (e.g., /tmp):
cp ij133-x86.tar.gz dcmie.zip /tmp
Change folder to be where you want to install ImageJ:
cd /usr/package/bin
Extract files from the file ij133-x86.tar.gz:
tar xvfz /tmp/ij133-x86.tar.gz
Change to the ImageJ directory and unpack the dcmie.zip file:
cd ImageJ; unzip /tmp/dcmie.zip
Move the Dcm folder to the plugins subdirectory:
mv Dcm plugins/
Edit the "run" file as follows:
./jre/bin/java -Xms96m -Xmx96m -cp ./lib/dcmie.jar:ij.jar ij.ImageJ
we suggest that you change 96m to represent about of the physical memory on your computer (eg:
384m if your computer has 512MB of memory).
Execute the run script:
./run &
Verify that the sub-menu Dcm appears in the Plugins menu.
Remove ij133-x86.tar.gz and dcmie.zip:
rm /tmp/ ij133-x86.tar.gz /tmp/dcmie.zip
5.2.3 Osiris and OsiriX

Not included on this DVD, Osiris is nonetheless an excellent free program for viewing DICOM images. Versions exist
for MS Windows, MacOS 9 and MacOS X 10.3 (or later). All of these versions can be
easily found by starting from here:
http://www.sim.hcuge.ch/osiris/
5.3 Viewing the Images
5.3.1 DicomWorks
It is easy to accidentally make DicomWorks scan all images on the hard disk, which can slow down your
computer dramatically. Here is the recommended way to get started:
Launch the DicomWorks program.
From the Tools menu, choose Preferences. In the Preferences window, select the second "Display" option.
Choose the option "Only make a thumbnail of the middle image in a series" or un-check the box
"Automatically build series thumbnails (if selected)". Then close the preferences menu.
From the toolbar at the top of main window, click on DICOM Directory, and then navigate to the folder that
you wish to load images from (e.g., F:\0.C.1\9088841 if your hard disk is drive F: and you want to look at
baseline images for participant 9088841.
At the bottom of the window, you should see a status message that says "Scanning current directory In a
minute or so (depending upon your PC's speed) you will see a list of the participants images.
Double click a participant to browse their images.
5.3.2 ImageJ with DICOM Im/Ex Plugin

Launch the ImageJ program. Make sure that it is launched with the proper options, so that it will find the DICOM
Im/Ex plugin (instructions are found at http://www.iftm.de/telemedizin/dcmimex.htm).
From the Plugins menu, choose Dcm->Dcm Import.
Select the FILESYSTEM tab, if it does not appear by default.
Press the folder icon and navigate to the required participants folder on the hard disk (eg:
F:\0.C.1\9088841).
Double Click on Study Date to see series available from that date.
Double Click on the folder of a series you want to view.
Single click on one image from the list to preview an image.
Select one or more images (shift-click or drag) and press "Import images". The images will open in a
scrollable window in ImageJ.
Use the scroll bar at the bottom of the window to view the images one by one.
100
Appendix
5.3.3 Osiris 4.19 for Windows.

Launch Osiris.
In the File selection dialog box, navigate the required participants folder (eg: F:\0.C.1\9088841) and click
the Patient List button.
Select a study from the list of Studies and a Series from the list of Series.
Select from the thumbnails the images you wish to view, and press "Open".
6. ACKNOWLEDGEMENTS
The DICOM tags for all images were processed using the program "dcmodify", and the DICOMDIR data files were
created using the program "dcmmkdir". Both programs are part of the DCMTK DICOM ToolKit by the OFFIS
Computer Science Institute (http://www.dcmtk.org/)
7. REFERENCES
1. Buckland-Wright JC, Macfarlane DG, Williams SA, Ward RJ. Accuracy and precision of joint space width
measurements in standard and macroradiographs of osteoarthritic knees. Ann Rheum Dis 1995; 54:872-80.
2. Peterfy C, Li J, Zaim S, et al. Comparison of Fixed-Flexion Positioning with Fluoroscopic Semi-Flexed
Positioning for Quantifying Radiographic Joint-Space Width in the Knee: Test-Retest Reproducibility. Skeletal
Radiology 2003; 32:128-32.
3. Vignon E, Piperno M, Le Graverand MP, et al. Measurement of radiographic joint space width in the tibiofemoral
compartment of the osteoarthritic knee: comparison of standing anteroposterior and Lyon schuss views. Arthritis
Rheum 2003; 48:378-84.
101
Appendix
APPENDIX A: Details of Participants in Group A
A.1 Participants in Group A

200 participants were selected to be in image release Group A. Participants were eligible for Group A if they met
the following requirements:
Had all baseline clinical data (Initial Eligibility, Screening Visit, Enrollment Visit) fully entered and fully edited
in the data system
Met the minimum baseline imaging requirements (see Section 2 for imaging details):
o accepted fixed flexion knee radiograph
o accepted SAG 3D DESS WE in at least one knee
o if in Progression subcohort, accepted SAG IW TSE FS and COR IW TSE in the same knee as
accepted SAG 3D DESS WE
Met the minimum baseline biospecimens requirements
Had neither knee replaced
In the Progression or Incidence subcohort
For details of subcohort definitions and biospecimen requirements see OAI Online at:
http://www.oai.ucsf.edu/
From this pool, a stratified random sample of 200 participants was chosen. The strata for this random sample
were:
Gender
Subcohort assignment (Progression and Incidence)
Clinic (four Recruitment Centers)
There were 4 gender x subcohort strata within each Recruitment Center (16 strata in all), from each of which 13
participants were randomly picked, giving a total of 208 to choose from. The number of image types missing for each
participant was then calculated and 8 participants eliminated who had the most missing images, while keeping the
number of participants from each Recruitment Center equal (25 each), the subcohorts balanced (100 of each), and
the genders balanced (100 of each).
A.2 Participant Demographics for Group A

The following two tables show the age, sex and race distributions for the 200 participants in Group A.

White or Caucasian Non-White Total
Male 90 10 100
Female 79 21 100
Total 169 31 200
Age (years)
45 to 49 50 to 59 60 to 69 70 to 79 Total
Male 11 24 27 38 100
Female 11 29 35 25 100
Total 22 53 62 63 200
A.3 Clinical Data for Participants in Group A

Questionnaire, clinic measurement, and biomarker data are available in SAS datasets for the 200 participants in
Group A at the OAI Online website http://www.oai.ucsf.edu/. The 200 participants in Group A can be identified by the
variables in the ENROLLEES SAS dataset. The variables are V00IMAGESA, V01IMAGESA, V03IMAGESA,
V05IMAGESA, V06IMAGESA, V08IMAGESA, and will have non-zero values for Group A participants with knee
images at baseline, 12-month, 24-month, 36-month, 48-month, and 72-month visits respectively. To obtain the most
up-to-date clinical data on these participants, and to determine whether particular images are available for specific
participants, you will need to retrieve the latest version of the clinical datasets.
102
Appendix

APPENDIX B: Details of Participants in Group B
B.1 Participants in Group B

160 participants were selected to be in image release Group B if:
The participant was among the 1,383 for whom a clean set of baseline images was delivered to the
Coordinating Center from Synarc on 4/14/06.
The participant was assigned to the Progression subcohort
The participant had the following baseline characteristics:
o Accepted* fixed-flexion knee radiograph
o Accepted* three key MRI sequences (SAG 3D DESS WE, COR IW TSE, SAG IW TSE FS) in at least
one leg (see section 2 for imaging details)
o Minimum biospecimen sample, or at least a second try to obtain one o
No knee replacements
o Based on the clinics baseline knee radiograph reading, DID NOT have the combination of one knee with
end-stage OA (P01LXRKOA2=6 or P01RXRKOA2=6) and the other knee normal or nearly normal
(P01LXRKOA2 in (0,1,2) or P01LXRKOA2 in (0,1,2)
The participant had the following 12-month follow-up characteristics: o
Had completed a 12-month follow-up clinic visit as of 5/30/06
o Accepted* 12-month follow-up visit fixed-flexion knee radiograph
o Accepted* 12-month 3 key follow-up visit MRI sequences in at least one leg
For details of subcohort definitions and imaging and biospecimen requirements see OAI Online at:
From this pool of eligible participants, we chose a stratified random sample of 160, as balanced by clinical center and
gender as possible. Due to a lag in recruitment, one of the clinical centers was substantially under - represented in
this sample despite selecting all participants from the center who were in the eligible pool, so it was necessary to find
additional participants from this center who met the above criteria. Eleven additional eligible participants became
available from the most under-represented site by having their baseline images processed after the 1383 images
were transferred to the Coordinating Center, and most of these (those with the most useable images) were
substituted for participants with less optimal images from clinic/gender strata that were over-represented in the
original sample of 160. The final sample is still slightly unbalanced with not all cells containing 20 participants; under-
filled cells were balanced by over-filled cells from other clinics, as much as possible preserving the 50:50 sex ratio of
the sample. The final gender balance is 81 females to 79 males.
B.2 Participant Demographics for Group B

The following two tables show the age, sex and race distributions for the 160 participants in Group B.

Male 66 13 79
Female 66 15 81
Total 132 28 160

Age (years)
45 to 49 50 to 59 60 to 69 70 to 79 Total
Male 12 20 23 24 79
Female 14 29 20 18 81
Total 26 49 43 42 160
*
Either of acceptable quality based on central QA review or one that was not selected for central QA and was
accepted by default without QA review.
103
Appendix

B.3 Clinical Data for Participants in Group B
Group B at the OAI Online website http://www.oai.ucsf.edu/. The 160 participants in Group B can be identified by the
variables in the ENROLLEES SAS dataset. The variables are V00IMAGESB, V01IMAGESB, V03IMAGESB,
V05IMAGESB, V06IMAGESB, V08IMAGESB, and will have non-zero values for Group B participants with knee
images at baseline, 12-month, 24-month, 36-month, 48-month, and 72-month visits respectively. To obtain the most
up-to-date clinical data on these participants, and to determine whether particular images are available for specific
participants, you will need to retrieve the latest version of the clinical datasets.
104
Appendix
APPENDIX C: Details of Participants in Group C
C.1 Participants in Group C

The participants included in Group C image releases are essentially all OAI Study participants who were enrolled up
to and including 4/30/2005, along with any participants whose images were released in image releases 0.A.1, 0.B.1
or 1.B.1. We call this group of participants Group C and it comprises 2686 people. 815 of the participants are in the
progression subcohort, 1862 are in the incidence subcohort and 9 are non-exposed controls. For details of
subcohort definitions see OAI Online at http://www.oai.ucsf.edu/
For the convenience of investigators wanting to start analysis of images, OAI releases group C images, along with
their clinical data before data and images from the entire cohort are available. These images and data are NOT
representative of the study cohort as a whole (compare demographics below to those in Appendix E). It is important to
remember that results from analysis of images from Group C may not be representative of a similar analysis of
images taken from the entire cohort. Therefore care must be taken when considering publication of results of image
analyses of Group C participants. In particular, it should be noted that a random selection of Group C participants will
be under-representative of African-Americans when compared to the entire OAI cohort, and over-representative of
participants with radiographic plus symptomatic knee OA (the progression subcohort). So when selecting subgroups
of Group C participants for image analyses, there are certain situations where the selection criteria may become
complex and based on the demographics or other characteristics of the entire cohort.
These preliminary images are released with a version number of x.C.y. When images and data are available for the
entire cohort, group C alone are no longer available for request. Instead the investigator is required to request the
entire cohort (Image release E), to ensure validity of analysis.
C.2 Participant Demographics for Group C

The following two tables show the sex, race, and age distributions at enrollment for the participants in Group C.
Distribution of Race by Sex (2 missing values for race, 1 male, 1 female)

Male 878 111 989
Female 1440 255 1695
Total 2318 366 2684

Age (years)
45 to 49 50 to 59 60 to 69 70 to 79 Total
Male 131 317 251 291 990
Female 168 521 582 425 1696
Total 299 838 833 716 2686
C.3 Clinical Data for Participants in Group C

Questionnaire, clinic measurement, and biomarker data are available in SAS datasets for the participants in Group C
at the OAI Online website http://www.oai.ucsf.edu/ . Group C participants can be identified by the variables in the
ENROLLEES SAS dataset. The variables are V00IMAGESC, V01IMAGESC, V03IMAGESC, V05IMAGESC,
V06IMAGESC, V08IMAGESC which will have non-zero values for Group C participants with knee images at baseline,
12-month, 24-month, 36-month, 48-month, and 72-month visits respectively. To obtain the most up-to-date clinical
data on these participants, and to determine whether particular images are available for specific participants, you will
need to retrieve the latest version of the clinical datasets.
105
Appendix
APPENDIX D: Details of Participants in Group D
D.1 Participants in Group D

The participants in Group D comprise OAI Study participants who had an interim 18-month clinic visit. We call this
group of participants Group D. They are all from the Progression subcohort. For details of subcohort definitions see
OAI Online at http://www.oai.ucsf.edu/
D.2 Participant Demographics for Group D

The following two tables show the age, sex and race distributions for the participants with images in Group D.

Male 97 13 110
Female 142 35 177
Total 239 48 287

Age (years)
45 to 49 50 to 59 60 to 69 70 to 79 Total
Male 11 42 23 34 110
Female 23 35 69 50 177
Total 34 77 92 84 287
D.3 Clinical Data for Participants in Group D

Group D at the OAI Online website http://www.oai.ucsf.edu/. The 287 participants in Group D can be identified by the
variables in the ENROLLEES SAS dataset. The variables are V00IMAGESD, V01IMAGESD, V02IMAGESD,
V03IMAGESD, V05IMAGESD, V06IMAGESD, V08IMAGESD and will have non-zero values for Group D participants
with knee images at baseline, 12-month, 18-month, 24-month, 36-month, 48-month, and 72-month visits respectively.
To obtain the most up-to-date clinical data on these participants, and to determine whether particular images are
available for specific participants, you will need to retrieve the latest version of the clinical datasets.
106
Appendix
APPENDIX E. Details of Participants in the Entire Cohort
E.0 Image Group E vs. Image Release E

For the convenience of investigators wanting to start analysis of images, OAI releases group C images (see Appendix
C), along with their clinical data before data and images from the entire cohort are available. These images and data
are NOT representative of the study cohort as a whole (compare demographics below to those in Appendix C) and
care must be taken when considering publication of results of image analyses of Group C participants. These
preliminary images are released with a version number of x.C.y. When images and data are available for the entire
cohort, group C alone are no longer available for request. Instead the investigator is required to request the entire
cohort (Image release E). The version numbers of the images pertaining to group C remain the same (x.C.y), but the
complement of this group (called group E) have version numbers of x.E.y. Thus image release E and image group E
are not equivalent, as are all other image groups and releases.
E.1 Participants in the Entire Cohort

Image releases with an E designator, comprise the whole OAI Study Cohort. Images are available for all 4796 OAI
Study participants, of whom 1390 participants are in the Progression subcohort, 3284 participants are in the incidence
subcohort and 122 participants are non-exposed controls. For details of subcohort definitions see OAI Online at
The following two tables show the age, sex and race distributions at enrollment for the whole OAI Study cohort of
4796 participants:
White or Caucasian Non-White Unknown Total
Male 1666 323 3 1992
Female 2124 678 2 2804
Total 3790 1001 5 4796

Age (years)
45 to 49 50 to 59 60 to 69 70 to 79 Total
Male 259 739 495 499 1992
Female 290 923 968 623 2804
Total 549 1662 1463 1122 4796
E.2 Participant Clinical, Measurement and Biomarker Data

Questionnaire and clinic measurement data are available in SAS datasets at the OAI Online website
http://www.oai.ucsf.edu/ . To obtain the most up-to-date clinical data on these participants, you will need to retrieve
the latest version of the clinical datasets. The ENROLLEES SAS datasets contains variables that let you easily
determine whether a participant has knee MRI and/or knee x-rays available at specific visits. For a specific visit, the
sum of two variables (or value or one variable) is used. The following table explains their use:
VISIT Variables to use to check for No Knee Knee MRI Knee x-rays Knee MRI &
Knee MRI and X-rays Images Knee x -rays
Baseline V00IMAGESC(a)+V00IMAGESE 0 1 2 3
(a)
18-month V02IMAGESD 0 1 n/a(b) n/a(b)
24-month V03IMAGESC(a) +V03IMAGESE 0 1 2 3

30-month V04IMAGESG 0 1 n/a(b) n/a(b)

(a) Non -zero value for variables V00/V01/V03IMAGESC means that images for a participant were released in an
early image release of the specific visit prior to the release of images for the whole cohort
(b) MRI images only are acquired in subsets of participants at 18-month and 30-month interim visits.
107
Appendix
Variables associated with contacts prior to enrollment are identified by a prefix starting with P (e.g., P01 variables are
from the screening visit where the baseline fixed flexion knee radiographs are acquired). Variables associated with
enrollment visit are prefixed V00, variables at 12-month follow-up are prefixed V01, variables at 18-month interim visit
are prefixed V02, variables at 24-month follow-up are prefixed V03, variables at 30-month interim are prefixed V04,
variables at 36-month follow-up visit are prefixed V05, variables at 48-month follow-up visit are prefixed V06, and
variables at 72-month follow-up visit are prefixed V08.
To determine exactly which MRI pulse sequences are available in which knee(s) for participants at specific visits,
and to determine the availability of all types of x-ray images meta-data can be downloaded from OAI Online
(http://www.oai.ucsf.edu).
For each visit, meta-data describing the type, acquisition, and quality of the images in this data release can be found
in a pair of SAS datasets, XRAYxx and MRIxx, where xx is 00 for baseline, 01 for the 12-month, 02 for the 18-month
visit, 03 for the 24-month visit, 04 for the 30-month visit, 05 for the 36-month visit, 06 for the 48-month visit, and 08 for
the 72-month visit. For example, the dataset XRAY00 contains the x-ray meta-data from baseline and XRAY01
contains the x-ray meta-data from the 12-month follow-up visit. Please note the x-ray and MRI datasets have multiple
records per participant. For example, the dataset XRAY00 contains one record per baseline radiograph expected or
released for each of the participants in this image release. The dataset MRI00 contains one record per knee per
baseline standard OAI MRI pulse sequence (DICOM series), excluding MRI localizer sequences, for each of the
participants (section 2 of this document outlines the types of x-ray and MRI images that can be available for each
knee). Further information can be found in the zipped documentation that accompanies each SAS dataset.
Further details of SAS datasets can be found in the document Overview of Structure, Use, and Conventions for OAI
data at: http://www.oai.ucsf.edu/datarelease/DataClinical.asp and descriptions of the clinical and imaging data
version conventions can be found in the document OAI Data Release Version Conventions at:
http://www.oai.ucsf.edu/datarelease/docs/DataReleaseVersionConventions.pdf
108
Appendix
APPENDIX F: Details of Participants in Group F
F.1 Participants in Group F

The participants included in this image release comprise all OAI Study participants in the non-exposed control
subcohort. For details of subcohort definitions see OAI Online at http://www.oai.ucsf.edu/
F.2 Participant Demographics for Group F

The following two tables show the age, sex and race distributions for the participants in Group F.

Male 41 6 47
Female 72 3 75
Total 113 9 122

Age (years)
45 to 49 50 to 59 60 to 69 70 to 79 Total
Male 14 16 12 5 47
Female 24 36 13 0 75
Total 38 54 25 5 122
F.3 Clinical Data for Participants in Group F

Group F at the OAI Online website http://www.oai.ucsf.edu/ The 122 participants in Group F can be identified by
having V00COHORT=3 in the ENROLLEES dataset. The variables V00IMAGESF, V01IMAGESF, V03IMAGESF,
V05IMAGESF, V06IMAGESF, and V08IMAGESF (also in the ENROLLEES dataset) will have non-zero values for
Group F participants with knee images at baseline, 12-month, 24-month, 36-month, 48-month, and 72-month visits
respectively. To obtain the most up-to-date clinical data on these participants, and to determine whether particular
images are available for specific participants, you will need to retrieve the latest version of the clinical datasets.
109
Appendix
APPENDIX G: Details of Participants in Group G
G.1 Participants in Group G

The participants in Group G comprise OAI Study participants who had an interim 30-month clinic visit. We call this
group of participants Group G. They are all from the Progression subcohort. For details of subcohort definitions see
OAI Online at http://www.oai.ucsf.edu/
G.2 Participant Demographics for Group G

The following two tables show the age, sex and race distributions for the participants with images in Group G.

Male 198 45 243
Female 137 103 240
Total 335 148 483

Age (years)
45 to 49 50 to 59 60 to 69 70 to 79 Total
Male 28 92 68 55 243
Female 23 87 85 45 240
Total 51 179 153 100 483
G.3 Clinical Data for Participants in Group G

Group G at the OAI Online website http://www.oai.ucsf.edu/. The 483 participants in Group G can be identified by the
variables in the ENROLLEES SAS dataset. The variables are V00IMAGESG, V01IMAGESG, V03IMAGESG,
V04IMAGESG, V05IMAGESG, V06IMAGESG, V08IMAGESG, and will have non-zero values for Group G participants
with knee images at baseline, 12-month, 24-month, 30-month, 36-month, 48-month, and 72-month visits respectively.
To obtain the most up-to-date clinical data on these participants, and to determine whether particular images are
available for specific participants, you will need to retrieve the latest version of the clinical datasets.
110
Appendix
APPENDIX X: Knee MRI Pulse Sequence Parameters

MP_LOCATOR COR_IW_TSE SAG_3D_DESS SAG_IW_TSE_FS COR_T1W_3DFLASH_WE SAG_T2_MAP
Weighting T1W Int T2* Int T1W T2 Map

Plane 3-plane Coronal Sag Sagittal Coronal Sagittal
Fat Sat No No WE Yes WE No
Matrix (phase) 256 307 307 313 512 269
Matrix (freq) 512 384 384 448 512 384
No. of slices 21 41 160 37 80 21
FOV (mm) 200 140 140 160 160 120
Slice thickness (mm) 5 3 0.7 3 1.5 3
Skip (mm) 1 0 0 0 0 0.5
Flip Angle (deg) 40 180 25 180 12 n/a
TE/TI (ms) 5 29 4.7 30 7.57 10, 20, 30, 40, 50, 60, 70
TR (ms) 10 3850 16.3 3200 20 2700
BW (Hz/pixel) 250 352 185 248 130 250
Chemical Shift (pixels) 1.8 1.3 0 0 0 1.8
NAV (NEX) 1 1 1 1 1 1
Echo train length 1 7 1 5 1 1
Phase Encode Axis A/P, R/L R/L A/P S/I R/L A/P
Phase Partial Fourier (8/8 = 1) 1 1 1 1 1 1
Readout Partial Fourier (8/8 = 1) 1 1 1 1 1 1
Slice Partial Fourier (8/8 = 1) 1 1 0.75 1 0.75 0.75
Options: elliptical k- elliptical k-space elliptical k-space elliptical scanning, elliptical elliptical k-space filter and
space filter and filter, elliptical filter and large k-space filter, larger FOV large FOV filter
large FOV filter sampling, large FOV filter filter
FOV filter
Distance Factor (%) 50 0 0 0 0 16
Phase Oversampling 0 20 0 40 0 0
Slice Oversampling 0 0 10 0 0 0
Phase Resolution 50 80 80 70 100 70
Averaging Technique Short Term Short Term Short Term Short Term Short Term Short Term
Gradient Rise Time Fast Fast Fast Fast Fast Fast
RF Amplitude Normal Normal Fast Normal Fast Normal
X-Resolution (mm) 0.391 0.365 0.365 0.357 0.313 0.313

Y-Resolution (mm) 0.781 0.456 0.456 0.511 0.313 0.446
Scan time (min) 0.5 3.4 10.6 4.7 8.6 10.6
111
Appendix
112
Appendix
APPENDIX B: 3D Slicer Editor Effects

At Editor Module, Effects operate either by clicking the Apply button in the GUI
or by clicking and/or dragging in the slice view. For operations where you use the GUI,
the target label map is the one in the Red slice viewer. Otherwise the target is based on
which slice viewer you click in.
Threshold Effect
Select a label value.
Voxels in the source volume within the range will have the label value that is
selected.
When apply is clicked, the labels are written into the label map.
Paint
Pick the radius (in millimeters) of the brush to apply.
Left click to apply single circle.
Left click and drag to draw a curve.
A trace of circles is left which are applied when the mouse button is released.
The 'smudge' option can be used when working on the boundary between
two or more regions. In this mode, the label color is set to the value of the
label map where the first left click occurs. This mode allows you to quickly
switch between colors.
Draw
Left click to lay individual points of an outline.
Left drag to lay down a continuous line of points.
'X' key to delete the last point added.
Right click or 'a' key to apply outline to label map.
Wand
Click to fill connected regions with similar intensities.
Click multiple times to grow the region.
Tolerance option controls how similar adjacent pixels must be in order to be
filled.
Max pixels per click determines how quickly the region is filled.
113
Appendix
Fill Volume option allows for 3D segmentation.
Rectangle
Left click to start rubber band rectangle.
Release button to apply.
Level tracing
Moving the mouse defines an outline where the pixels all have the same
background value as the current background pixel.
Clicking the left mouse button applies that outline to the label map.
Identify Island
Use this tool to create a unique label value for each connected region in the
current label map. Connected regions are defined as groups of pixels which
touch each other but are surrounded by zero valued voxels.
If FullyConnected is selected, then only voxels that share a face are counted
as connected; if unselected, then voxels that touch at an edge or a corner are
considered connected.
Change Island
Changes color of an island, where Island is a group of connected pixels with
the same label. Does not change other voxels with the same label.
This tool is useful after thresholding, to separate anatomy further.
o Click Change Island
o Click New Label and select a color, or enter a color number
o Click any part of the island
o Pixels that are part of that island will receive the New Label value
Remove Island
Remove Islands: Automatically removes unlabelled voxels within a region.
Useful for removing small areas of noise inside a surrounding thresholded
area.
Two modes are supported:
o Connectivity mode removes unsegmented islands that are completely
enclosed with segmentation
o Morphology mode removes small regions in and out of the
segmentation even if they are partly connected to the larger regions.
114
Appendix
Save Island
Save Island retains the selected island, and removes disconnected pixels
with the same label. This is another way to clean up the results of a
thresholding operation.
Useful for separating a single connected structure from other structures.
Erode and Dilate

Add or remove a single layer of pixels from the current label value.
Change label
Changes every voxel in the target label map with the input value to the output
value.
Modelmaker
A simple interface to of the full CLI module.
The currently selected label number will be used to build.
Select smooth model for more pleasing results.
Just click apply and models of all label maps will be created.
Model building runs in the background and the model appears in the 3D view
when finished.
GrowCutSegment
GrowCut is a powerful algorithm that uses example segmentation to create a
full segmentation of the volume. You can use multiple label colors to define
regions that represent parts of anatomical structures and then let the GrowCut
algorithm seek out a best labeling for adjacent pixel so they match your
example.
To use this, you first define your example segmentation using the normal
paint or draw tools as described above. Define at least two structure classes
and then click Apply. If you are not satisfied with the result, use the Undo
button to go back and modify your example.
115
Appendix
WatershedFromMakers
Watershed From Markers a multi-lable segmentation tool which has similar
behavior to the GrowCuts algorithm. As input it expects the current label map
to be multiple labels or markers defining different objects. By applying the
algorithm, the labels are greedily grown with the minimal gradient value. So
object with well defined edges are desirable for good segmentations with this
method.
FastMarching
FastMarching is a statistics-based region growing 3D segmentation
algorithm. Paint to define seed voxels (one color only). Push "March" button
to initiate region growing. Use the marching slider to interactively examine the
front propagation history and choose the result you like.
116

Analisis de Soluciones Tecnologicas para El Tratamiento de La Osteoartritis

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Analisis de Soluciones Tecnologicas para El Tratamiento de La Osteoartritis

Caricato da

Copyright:

Formati disponibili

ESCUELA POLITCNICA SUPERIOR DE MONDRAGON

Trabajo presentado para la obtencin del ttulo de

GRADO EN INGENIERA MECNICA

Ttulo del Trabajo Lanaren izenburua

ANLISIS DE SOLUCIONES TECNOLGICAS PARA EL

Autor Egilea DIEGO MAURO POJMAEVICH

Curso Ikasturtea 2015/2016

ANLISIS DE SOLUCIONES TECNOLGICAS PARA EL

Nombre y apellidos del autor

Nombre y apellidos del/los director/es del trabajo

Lugar donde se realiza el trabajo

La osteoartritis es una enfermedad producida por el desgaste del

Este proyecto tuvo como objetivo obtener un modelo 3D de la rodilla de un

Luego se concluy en que es posible segmentar una rodilla con osteoartritis

Osteoarthritis is a disease caused by loss of cartilage tissue, which acts as a

This project aimed to obtain a 3D model of the knee of a patient affected by

Chapter 1: Introduction .............................................................................................. 1

Chapter 2: Knee anatomy and kinematics ................................................................ 7

Chapter 3: Osteoarthritis .......................................................................................... 21

Anlisis de soluciones tecnolgicas para el tratamiento de la osteoartritis

3.2.2. Weight Management ................................................................................. 27

Chapter 4: Knee prostheses ....................................................................................... 37

Anlisis de soluciones tecnolgicas para el tratamiento de la osteoartritis

4.1.4.2. Most used materials ........................................................................... 50

Chapter 5: Manual segmentation of a knee............................................................. 61

Chapter 7: References .............................................................................................. 79

Anlisis de soluciones tecnolgicas para el tratamiento de la osteoartritis

Anlisis de soluciones tecnolgicas para el tratamiento de la osteoartritis

Fig. 24. Complex replacement .................................................................................... 31

Anlisis de soluciones tecnolgicas para el tratamiento de la osteoartritis

Anlisis de soluciones tecnolgicas para el tratamiento de la osteoartritis

According to the World Health Organization (WHO),

Osteoarthritis is a degenerative joint disease, which mainly affects the articular

As it is known, cartilage is the slippery

By the year 2004, osteoarthritis globally causes moderate to severe disability in

1.2. Goals and project phases

1. Analysis of knee anatomy and kinematics.

2. DICOM image selection.

3. Segmentation and generation of the 3D model.

2.1. Anatomy of the knee joint

Functionally it is classified as a ginglymus or hinge-joint and according to its

It consists of three articulations in one:

Two condyloid joints, one between

One partly arthrodial, between the

Fig. 3. Knee anatomy

Also, the knee is composed of three functional compartments (Fig. 3):

Patellofemoral compartment, which consists of the patella, or "kneecap", and

2.1.1 Articular surfaces

Fig. 4. Lower extremity of the femur

2.1.1.2 Upper extremity of the tibia

Fig. 5. Upper extremity of the tibia

The menisci or meniscus are two crescent-shaped fibrocartilaginous structures

Each meniscus covers approximately the peripheral two-thirds of the

Fig. 6. Medial and lateral meniscus

The front surface (Fig. 7A) is divided into three parts:

The middle third has numerous vascular canaliculi.

The posterior surface (Fig. 7B) is divided into two parts.

Fig. 7. Patella. Anterior facet (A); Posterior facet (B).

2.1.2. Articular ligaments

Intracapsular ligaments (Fig.10)

Anterior Cruciate Ligament (ACL)

Posterior Cruciate Ligament (PCL)

Anterior meniscofemoral ligament or