Clinical Radiology 70 (2015) 235e242

Contents lists available at ScienceDirect

Clinical Radiology
journal homepage: www.clinicalradiologyonline.net

Technical Report

The 10 key steps for radiographic analysis of

adolescent idiopathic scoliosis
J.R. Panchmatia a, *, A. Isaac b, T. Muthukumar c, A.J. Gibson c, J. Lehovsky c
a
Department of Orthopaedic Surgery, Johns Hopkins Hospital, 601 N Caroline Street, Baltimore, MD 21287, USA
b
Kings College Hospital, Denmark Hill, London SE5 9RS, UK
c
The Royal National Orthopaedic Hospital, Brockley Hill, Stanmore, Middlesex HA7 4LP, UK

art icl e i nformat ion

Article history:
Received 4 December 2013
Accepted 19 November 2014

Introduction With an estimated prevalence of between 2 and 9%,

Scoliosis, defined as a coronal deformity of the spine,
was first described as a clinical entity by Hippocrates.1 In
order for a diagnosis of scoliosis to be made, a curve
greater than 10
in magnitude needs to be present.2
Scoliosis arising between 11e18 years can therefore be
broadly divided into idiopathic scoliosis and scoliosis with
a known underlying osteogenic or neurological cause.3
Idiopathic scoliosis in turn can be further subdivided
based on the age of onset. Adolescent idiopathic scoliosis,
a term first coined by Professor J. I. P. James of the Royal
National Orthopaedic Hospital, refers to scoliosis arising
between the age of 11 and 18 years4e6 without a known
underlying cause.
* Guarantor and correspondent: J.R. Panchmatia, Department of Ortho-
paedic Surgery, Johns Hopkins Hospital, 601 N Caroline Street, Baltimore,
MD 21287, USA. Tel.: þ1 347 925 2115.
E-mail address: jaykarpanchmatia@hotmail.com (J.R. Panchmatia).
adolescent idiopathic scoliosis is a significant cause of
morbidity amongst young people. These include back
pain, negative self-image and, in the case of large curves,
impaired pulmonary and cardiac function. Ultimately,
however, fewer than 10% of patients with adolescent
idiopathic scoliosis require treatment.7,8 The radiographs
of patients with scoliosis are frequently reviewed in both
general hospitals and specialists centres. Referrals to the
radiology department for the radiographic assessment of
scoliosis occur as direct referral from primary care or from
paediatric and general orthopaedic and neurosurgical
services, in addition to spinal surgeons.9 In spite of the
high prevalence of this condition and its general ubiquity,
there remains a lack of clarity regarding how to analyse
and best describe the radiographic features of
scoliosis.10e12
This suboptimal radiographic assessment of scoliosis is
of grave clinical concern as the radiographic evaluation of
scoliosis is central not only in deciding whom to treat but
also in deciding what treatment to offer.13 For example, one
of the key indications for surgery is the risk of curve pro-
gression. This is predicted by radiological findings such as

http://dx.doi.org/10.1016/j.crad.2014.11.013
0009-9260/Ó 2015 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
236 J.R. Panchmatia et al. / Clinical Radiology 70 (2015) 235e242
Table 1
The Risser staging system.
Risser stage Degree of ossification
0 None
1 1e25%
2 26e50%
3 51e75%
4 76e100%
5 Complete fusion to the iliac crest
curve magnitude and indicators of skeletal maturity.14
Moreover, once the decision has been made to proceed
with surgery, radiological features, such as whether a curve
is flexible, play a central role in determining surgical
approach and the extent of spinal fusion.15 It is for this
reason that we propose a simple algorithm that may be
easily used by radiologists and clinicians when reviewing
the radiographs of a patient with adolescent idiopathic
scoliosis (Table 1).
Posteroanterior radiograph
1. Confirm the radiograph shows the whole spine and
enumerate the vertebrae (Fig 1).
Figure 1 An adequate posteroanterior radiograph including the cer-
vical spine and both iliac crests in a patient with six lumbar vertebrae.
Before analysing a posteroanterior radiograph of the
spine, its adequacy must be confirmed. The radiograph
must be that of the standing patient and must include the
cervical spine, sacrum, and both iliac crests to identify
congenital anomalies and associated syndromes, such as
Klippel Feil syndrome and spinal dysraphism. The reason
for obtaining a posteroanterior radiograph as opposed to an
anteroposterior radiograph is to minimize radiation expo-
sure to the breasts.16
The correct enumeration of each vertebra is essential
when describing a radiograph of the spine, partly
because it is an important means of reducing the likeli-
hood of wrong level surgery.17 Counting cephalad from
the presumed fifth lumbar vertebra risks overlooking
variations in lumbosacral anatomy. This is an issue of
considerable importance given that only around 90% of
people have five lumbar vertebrae. We instead recom-
mend identifying the second cervical vertebra and
counting caudally.18
2. Count the number of curves present (Fig 2).
Typically the number of curves present does not exceed
three.19
Figure 2 A radiograph demonstrating two lateral curves.
 J.R. Panchmatia et al. / Clinical Radiology 70 (2015) 235e242 237

3. Identify the apical vertebra and describe the location of

each curve (Fig 3).

The apical vertebra is defined by the Scoliosis Research

Society as that which is most laterally deviated from the
vertical axis. The vertical axis is a line passing through
the centre of the sacrum, also known as the central sacral
line. Having determined the apical vertebra, the curve
can be described as being a left-sided curve or a right-
sided curve. If the apex is located between T2 and the
T11/12 disc the curve can be further described as
thoracic, if the apex is located between T12 and L1 it can
be described as thoracolumbar, and if it is located be-
tween the L1/2 disc and L4 the curve can be described as
lumbar.19

4. Identify the end vertebrae of each curve and quantify

the magnitude of each curve (Fig 4).

Each curve has two end vertebrae. The cephalad/su-

perior end vertebra is defined as the first vertebra in a

Figure 4 A right-sided thoracic curve with a Cobb angle of 85.3
.

cephalad direction from the curve apex whose superior

surface is tilted maximally toward the concavity of the
curve. The caudad/inferior end vertebra is the first
vertebra in a caudad direction from a curve apex whose
inferior surface is tilted maximally toward the concavity
of the curve. These two end vertebrae are used to calcu-
late the Cobb angle, a measure of a curve’s magnitude. The
Cobb angle is defined as the angle subtended by a line
parallel to the superior endplate of the cephalad end
vertebra and a line parallel to the inferior endplate of the
caudad end vertebra. Interobserver, intra-observer, and
diurnal variabilities have been recorded. Although there
are inherent limitations in the accuracy of measurements
of the Cobb angle, it remains one of the standard mea-
surements for primary evaluation and follow-up of the
scoliosis curves. Multiple measurements and serial com-
parisons are therefore performed.9

5. Determine which curve is the major curve (Fig 5).

The major curve is defined as the curve with the largest

Cobb angle measurement. All other curves are minor
Figure 3 A patient with a right thoracic and a left lumbar curve. curves.
238 J.R. Panchmatia et al. / Clinical Radiology 70 (2015) 235e242
Figure 5 A patient with a major thoracic curve and a minor lumbar
curve.
6. Assess for rotation (Fig 6).
Scoliosis is not a two-dimensional condition; rather, it is
a three-dimensional condition associated not only with a
lateral curvature but also with a degree of vertebral rota-
tion. The greater the degree of displacement of the convex
pedicle in relation to the centre of vertebral body, the
greater the degree of rotation.20 Several methods have
been described for radiographic assessment of rotation,
including the Nash and Moe, Pedriolle, and Vidal
methods.21 Assessment of vertebral rotation is performed
clinically or more accurately on cross-sectional imaging
(CT/ MRI).
7. Assess coronal balance (Fig 7).
Patient symptoms are in part a reflection of truncal
imbalance. Truncal imbalance has two components: cor-
onal and sagittal. Global and regional coronal imbalance
can be assessed using a posteroanterior radiograph.
Global coronal imbalance is quantified by measuring the
Figure 6 The pedicles on the convexity of the rotated thoracic curve
gradually disappear from view, whereas the pedicles on the convexity
of the non-rotated lumbar curve remain visible throughout.
distance between the C7 vertebral body plumb line and
the central sacral vertical line. Normal global coronal
balance is when the distance between the C7 vertebral
body plumb line and the central sacral vertical line is less
than 1 cm.22 Regional imbalance, on the other hand,
tends to focus on shoulder asymmetry. This is in part
because shoulder asymmetry has been shown to be a
source of psychological distress in patients with scoliosis.
Moreover, correction of shoulder imbalance strongly in-
fluences patient satisfaction with surgical outcomes.
Shoulder asymmetry can be quantified by measuring the
clavicle angle. This is the angle subtended by a line con-
necting the distal aspect of both clavicles and the
horizontal.23
8. Assess skeletal maturity (Fig 8).
Skeletal maturity can be estimated by assessing the
degree of ossification of the iliac crest apophysis using the
 J.R. Panchmatia et al. / Clinical Radiology 70 (2015) 235e242
Figure 7 A patient with a clavicle angle of 4.4
reflecting mild
shoulder asymmetry but with normal global coronal balance.
Risser staging. The Risser stage is a strong predictor of
curve progression.14 However, if in doubt, the bone age
with a plain radiograph of the left hand should be
considered. In boys, Risser staging is considered less reli-
able as ossification of the apophysis commences earlier
than in girls.9 It is also important to state whether the
triradiate cartilage is fused.
Lateral bending views
9. Determine whether a curve is structural or non-
structural (Fig 9)
Curves can be categorized as either structural or non-
structural. A structural curve is one that does not correct
Figure 8 A child with fused triradiate cartilages at Risser Stage 3.
239
past 0
on lateral bending towards the side of the curve’s
apex. It is typically the largest curve present. On the other
hand, a non-structural curve corrects past 0
on lateral
bending. A non-structural curve can be regarded as a flex-
ible, compensatory curve maintaining truncal balance. In
time, however, flexible non-structural curves can evolve
into inflexible structural curves. This is due to the short-
ening of ligaments, bony degenerative changes, and muscle
atrophy.
Lateral radiograph
10. Assess sagittal balance globally and regionally (Fig 10).
Sagittal imbalance culminates in pain, functional
disability, concerns about self-image, and impaired social
interaction due to an inability to maintain horizontal gaze.
As such, quantifying a patient’s sagittal profile with a view
to it being corrected surgically is a key component of the
radiographic evaluation of scoliosis.
Given that sagittal balance refers to the relationship of
the head to the pelvis, a lateral radiograph must include
the whole spine and both femoral heads. The present
authors advocate assessing sagittal balance both globally
and regionally. The global assessment involves drawing a
vertical plumb line from the centre of the vertebral body
of C7 and measuring its horizontal offset from the post-
erosuperior corner of S1.24 A plumb line that falls anterior
to the posterosuperior corner of S1 is indicative of a
positive sagittal balance. Likewise, a plumb line that falls
posterior to the posterosuperior corner of S1 is indicative
of a negative sagittal balance. There is a lack of consensus
regarding the value at which a negative or positive
sagittal balance switches from being physiological to
pathological, with values ascribed including 2.5 and
5 cm.25
The regional assessment of sagittal balance involves
analysing both spinal and pelvic parameters. The present
authors suggest measuring two spinal parameters: thoracic
kyphosis and lumbar lordosis. Thoracic kyphosis can be
calculated by measuring the angle subtended by a line
parallel to the cephalad endplate of T4 and a second line
parallel to the caudad endplate of T12. Lumbar lordosis is
the angle subtended by a line parallel to the cephalad
endplate of L1 and a second line parallel to the caudad
endplate of L5. Thoracic kyphosis is typically between
20e50
, whereas lumbar lordosis is typically between
31e79
(Fig 11).24,26
As low as reasonably achievable (ALARA) and
radiological considerations for the use of
ionizing radiation
Although outside the remit of this review, the importance
of minimizing radiation exposure to patients with adolescent
idiopathic scoliosis should be reiterated. The association be-
tween cumulative irradiation in young people and an
240 J.R. Panchmatia et al. / Clinical Radiology 70 (2015) 235e242

Figure 9 Lateral bending views confirming the presence of structural thoracic and thoracolumbar curves. These images have been flipped in the
horizontal plane so that they mimic the patient lying prone on the operating table. This is the norm at many spinal units. This ensures that
pre- operative imaging can easily be correlated with intra- operative surgical and radiological findings.
 J.R. Panchmatia et al. / Clinical Radiology 70 (2015) 235e242
Figure 10 A patient with a positive sagittal balance of 7.4 cm.
increased risk of neoplasia is well established. The breasts and
thyroid are especially sensitive to the cumulative effects of
radiation. Multiplicative risk models, calculating death risk
rate after exposure to ionizing radiation at ages 5, 25, and 45
years, show that for patients exposed to ionizing radiation at
age 5 years, the risk of death is two-times higher than for
patient exposed at age 25 years, and five-times higher than
for patient exposed at age 45 years.27 As such, posteroanterior
radiographs, collimation, ultra-high-speed films and grids,
and gonadal and breast shields should be used to minimize
radiation exposure. Also, whilst primary imaging of the spine
is often performed at the local hospital, further assessment
and any follow-up imaging should be performed at a
specialist spinal unit following clinical evaluation. This
multidisciplinary approach optimizes the availability of the
radiographic expertise and the infrastructure of tools
required for acquiring the images with relatively low-dose
techniques, as well as ensuring the availability of serial im-
ages on the same PACS system, and therefore, reduces the
radiation burden.
241
Figure 11 A patient with a thoracic kyphosis of 33.1
and a lumbar
lordosis of 62
.
Conclusion
By following these 10 simple steps, radiographs of pa-
tients suffering with scoliosis can be analysed and described
with clarity, thereby facilitating communication between
physicians (Table 2). The first eight of these commandments
are reliant on posteroanterior radiographs, the ninth is
determined using lateral bending views, whereas the final
commandment is dependent on lateral radiographs.
Table 2
A summary of the 10 key steps.
1 Confirm that the radiograph shows the whole spine and enumerate
the vertebrae
2 Count the number of curves present
3 Identify the apical vertebra and describe the location of each curve
4 Identify the end vertebrae of each curve and quantify the
magnitude of each curve
5 Determine which curve is the major curve
6 Assess for rotation
7 Assess coronal balance
8 Assess skeletal maturity
9 Determine whether a curve is structural or non-structural
10 Assess sagittal balance globally and regionally
242 J.R. Panchmatia et al. / Clinical Radiology 70 (2015) 235e242
References
1. Marketos SG, Skiadas P. Hippocrates. The father of spine surgery. Spine
1996;24:1381e7.
2. Weiss H, Negrini S, Rigo M, et al. Indications for the conservative
management of scoliosis (guidelines). Scoliosis 2006;1:5.
3. Kleinberg S. The operative treatment of scoliosis. Arch Surg
1922;5:631e45.
4. Anderson DG, Vaccaro AR. Decision making in spinal care. New York:
Thieme Medical Publishers; 2006.
5. James JI. Idiopathic scoliosis; the prognosis, diagnosis, and operative
indications related to curve patterns and the age at onset. J Bone Joint
Surg (Br) 1954;36:36e49.
6. Miller MD. Pediatric orthopaedics. In: Review of orthopaedics. 5th ed.
Philadelphia: Saunders; 2008. p. 215.
7. Asher MA, Burton DC. Adolescent idiopathic scoliosis: natural history
and long term treatment effects. Scoliosis 2006;31:1e10.
8. Lonstein JE. Scoliosis: surgical versus nonsurgical treatment. Clin Orthop
Relat Res 2006;443:248e59.
9. Cassar-Pullicino VN, Eisenstein SM. Imaging in scoliosis: what, why and
how? Clin Radiol 2002;57:543e62.
10. Goldberg MS, Poitras B, Mayo NW, et al. Observer variation in assessing
spinal curvature and skeletal development in adolescent idiopathic
scoliosis. Spine 1988;13:1371e7.
11. Morrissy RT, Goldsmith GS, Hall EC, et al. Measurement of the Cobb
angle on radiographs of patients who have scoliosis. Evaluation of
intrinsic error. J Bone Joint Surg (Am) 1990;72:320e7.
12. Beauchamp M, Labelle H, Grimard G, et al. Diurnal variation of Cobb angle
measurement in adolescent idiopathic scoliosis. Spine 1993;18:1581e3.
13. Aubin CE, Labelle H, Ciolofan OC. Variability of spinal intrumentaion
configurations in adolescent idiopathic scoliosis. Eur Spine J
2006;16:57e64.
14. Lonstein JE, Carlson JM. The prediction of curve progression in untreated
idiopathic scoliosis during growth. J Bone Joint Surg (Am)
1984;66:1061e71.
15. Suk AI, Kim JH, Kim SS, et al. Pedicle screw instrumentation in adoles-
cent idiopathic scoliosis. Eur Spine J 2012;21:13e22.
16. Ron E. Ionizing radiation and cancer risk: evidence from epidemiology.
Pediatr Radiol 2002;32:232e7.
17. Mody MG, Nourbakhsh A, Stahl DL, et al. The prevalence of wrong level
surgery among spine surgeons. Spine 2008;33:194e8.
18. Paik NC, Lim CS, Jang HS. Numeric and morphological verification of
lumbosacral segments in 8230 consecutive patients. Spine
2013;38:E573e8.
19. Lenke LG, Betz RR, Harms J, et al. Adolescent idiopathic scoliosis: a new
classification to determine extent of spinal arthrodesis. J Bone Joint Surg
(Am) 2001;83:1169e81.
20. Nash Jr CL, Moe JH. A study of vertebral rotation. JBJS (Am)
1969;51:223e9.
21. Pedriolle R, Vidal J. Morphology of scoliosis: three-dimensional evolu-
tion. Orthopedics 1987;10:909e15.
22. Qiu X, Ma W, Li W, et al. Discrepancy between radiographic shoulder
balance and cosmetic shoulder balance in adolescent idiopathic scoli-
osis patients with double thoracic curve. Eur Spine J 2009;18:45e51.
23. Malfair D, Flemming AK, Dvorak MF, et al. Radiographic evaluation of
scoliosis: review. AJR Am J Roentgenol 2010;194:S8e22.
24. Morvan G, Mathieu P, Vuillemin V, et al. Standardized way for imaging
of the sagittal spinal balance. Eur Spine J 2011;20:602e8.
25. Harding IJ. Understanding sagittal balance with a clinical perspective.
Eur J Phys Rehbil Med 2009;45:571e82.
26. Nnadi C, Roussouly P. Sagittal plane deformity: an overview of inter-
pretation and management. Eur Spine J 2010;19:1824e36.
27. ICRP. 1990 Recommendations of the International Commission on
Radiological Protection. ICRP Publication 60. Ann ICRP 1991;21(1e3).