Journal of Biomechanics 44 (2011) 2162–2167

Contents lists available at ScienceDirect

Journal of Biomechanics
journal homepage: www.elsevier.com/locate/jbiomech
www.JBiomech.com

Short communication

An upper extremity inverse dynamics model for pediatric Lofstrand

crutch-assisted gait$
Brooke A. Slavens a,b,n,1, Neha Bhagchandani c,2, Mei Wang c,d,e,2,3,4,
Peter A. Smith b,1, Gerald F. Harris b,c,d,e,1,4
a
Department of Occupational Science and Technology, University of Wisconsin-Milwaukee, P.O. Box 413, Milwaukee, WI 53201-0413, USA
b
Shriners Hospital for Children, 2211N. Oak Park Ave, Chicago, IL 60707, USA
c
Department of Biomedical Engineering, Marquette University, P.O. Box 1881, Milwaukee, WI 53201-1881, USA
d
Department of Orthopaedic Surgery, Medical College of Wisconsin, 9200W. Wisconsin Ave., Milwaukee, WI 53226, USA
e
Orthopaedic and Rehabilitation Engineering Center, 735N. 17th St, Suite 105, Milwaukee, WI 53233, USA

a r t i c l e i n f o a b s t r a c t

Article history: The objective of this study was to develop an instrumented Lofstrand crutch system, which quantifies three-
Accepted 8 May 2011 dimensional (3-D) upper extremity (UE) kinematics and kinetics using an inverse dynamics model. The
model describes the dynamics of the shoulders, elbows, wrists, and crutches and is compliant with the
Keywords: International Society of Biomechanics (ISB) recommended standards. A custom designed Lofstrand crutch
Biomechanics system with four, six-degree-of-freedom force transducers was implemented with the inverse dynamics
Crutches model to obtain triaxial UE joint reaction forces and moments. The crutch system was validated statically
Gait and dynamically for accuracy of computing joint reaction forces and moments during gait. The root mean
Inverse dynamics square (RMS) error of the system ranged from 0.84 to 5.20%. The system was demonstrated in children with
Upper extremity
diplegic cerebral palsy (CP), incomplete spinal cord injury (SCI), and type I osteogenesis imperfecta (OI). The
greatest joint reaction forces were observed in the posterior direction of the wrist, while shoulder flexion
moments were the greatest joint reaction moments. The subject with CP showed the highest forces and the
subject with SCI demonstrated the highest moments. Dynamic quantification may help to elucidate UE joint
demands in regard to pain and pathology in long-term assistive device users.
& 2011 Elsevier Ltd. All rights reserved.

1. Introduction
According to the latest NIDRR mobility device report, there are
over 6.8 million assistive device users, with an estimated 566,000
crutch users (Kaye et al., 2000). Recognition of upper extremity
(UE) pain and pathology associated with long-term assistive
device usage has been reported in the current literature (Opila
et al., 1987; Sie et al., 1992; Lal, 1998; Sala et al., 1998). Joint
forces during Lofstrand, or forearm, crutch-assisted gait have
been shown to lead to pain and approach levels of injury,
especially in long-term usage. It has been shown that the crutches
support 6% to 50% of body weight (BW), (Melis et al., 1999;
Requejo et al., 2005; Haubert et al., 2006; Slavens et al., 2009). The
$
All authors were fully involved in the study and preparation of the manu-
script. The material within has not been and will not be submitted for publication
elsewhere.
n
Corresponding author at: Department of Occupational Science and Technology,
University of Wisconsin-Milwaukee, P.O. Box 413, Enderis Hall 983, Milwaukee,
WI 53201-0413, USA. Tel.: þ 1 414 229 6933; fax: þ 1 414 229 5100.
E-mail address: slavens@uwm.edu (B.A. Slavens).
1
Tel.: þ773 622 5400; fax: þ773 385 5459.
2 placed on the UEs during crutch-assisted gait. This system may
Tel.: þ414 288 3375; fax: þ 414 288 7938.
3
Tel.: þ414 805 7456; fax: þ414 805 7488.
4
Tel.: þ414 288 0697; fax: þ414 288 0713.
high end of these reported loads may be of concern with regards
to potential for pain and injury in long-term crutch users.
Although no prior studies directly quantify all UE joints (i.e.,
shoulder, elbow, and wrist), previous endeavors have examined
UE dynamics during Lofstrand crutch-assisted gait to some
extent. Requejo et al. (2005) presented a system with sensors
around the crutch handle, which estimated the cuff forces
through derivation. Recently, a standardized UE inverse dynamics
model was developed and applied to a pediatric population
(Slavens et al., 2010). This system used a single force transducer
located at the tip of each crutch to acquire joint reaction forces
and moments during gait. Crutch handle and forearm cuff forces
and moments were not captured.
This study presents a significant improvement over the pre-
vious inverse dynamic model, using a four-sensor crutch system
to directly measure crutch cuff kinetics and fully quantify wrist
dynamics in addition to elbow, and shoulder. This model deter-
mines all UE joint dynamics and crutch contributions during gait.
The model is presented with a demonstration in a pediatric
population of cerebral palsy (CP), spinal cord injury (SCI), and
osteogenesis imperfecta (OI) to gain insight into the demands
prove useful for UE dynamics quantification of ambulation and to
develop therapeutic gait strategies for long-term crutch usage.

0021-9290/$ - see front matter & 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jbiomech.2011.05.012
 B.A. Slavens et al. / Journal of Biomechanics 44 (2011) 2162–2167
The investigation of the force demands placed on the UE may
have significant impact on rehabilitation protocols, injury pre-
vention strategies, and crutch design.
2. Methods
2.1. Kinematic model
The UEs were defined according to the International Society of Biomechanics
(ISB) standards using 18 markers (Wu et al., 2005). The seven rigid body segments
modeled were the thorax, upper arms, forearms, and hands, following similar
methods of Slavens et al. (2010). The crutches were segmented into the handle,
lower crutch, and cuff segments (Fig. 1; Table 1).
2.2. Kinetic model
The kinetic Lofstrand crutch system was composed of four, six-degree-of-
freedom, FS6 force transducers (2 per crutch; AMTI, Watertown, MA) and Walk
Easy (Walk Easy, Inc., Delray Beach, FL) crutches. Force transducers (0.10 kg) were
placed above and below the crutch handle to directly measure triaxial forces and
moments. The segmental kinetics follow inverse dynamics Newton–Euler meth-
odology (Zatsiorsky, 2002; Figs. 2–4; Table 2).
2.3. System evaluation
Initialization was conducted before every testing session to obtain a baseline
force. The procedure involved aligning the transducers vertically and horizontally
using a custom built calibration device, which secured the crutches at the handle
above the ground.
Static evaluation was performed by applying known loads to the instrumented
crutches while supported by the calibration device. Two pound loads were applied
Fig. 1. Crutch segment definitions and axes orientation. Each crutch segment
consists of the handle, lower crutch, and cuff segments. The coordinate systems of
the segments follow x-axis: anterior, y-axis: superior, and z-axis lateral. The cuff is
offset from the lower crutch by a 151 angle. Markers were placed on the cuff
segment, right and left sides of the lower load cell, and on the anterior and
posterior crutch tip.
Table 1
Three-dimensional kinematic equations of the crutch segment origins and axes, where msubscript is the respective marker.
Segment Origin X-axis
Handle lc c ¼ mlateral þ mmedial mlateral mmedial
2163
along the three primary axes of the crutch at the shaft and the forearm cuff to
independently evaluate each 6-axis transducer. Five trials were performed for
each loading condition. Percentage root mean square (RMS) error and standard
deviation (SD) were computed to determine system accuracy and precision during
static loading conditions.
Dynamic evaluation of the system was then performed having a normal
subject walk with the crutches over a force plate (Requejo et al., 2005; Slavens
et al., 2010). Forces in three dimensions were measured from the crutch
transducers and force plate. The resultant forces of the lower crutch transducers
and the force plate were compared. Percentage RMS error and SD from five trials
was computed to determine dynamic accuracy and precision.
The transducers were fully calibrated, including calibration and sensitivity
assessments for all four 6-axis sensors. Each sensor was also independently
calibrated to industry standards by AMTI. This involved each transducer calibrated
in eight different load locations to provide data for Fx, Fy, Fz, Mx, My, Mz,  Mx, and
–My. The primary sensitivities are defined in Table 3; the crosstalk was less than
2% in all cases.
All transducers were also tested dynamically to assure accuracy of the
preceding calibration procedure. The four 6-axis transducers and crutch system
met these criteria during dynamic evaluation.
2.4. Subject demonstration
Written parental consent and subject assent was obtained in compliance with
Shriners Hospitals for Children-Chicago’s Institutional Review Board (IRB) require-
ments. The UE system was applied to three Lofstrand crutch users (mean age:
11.3 yr; mean height ¼ 1.3 m; mean weight¼29.7 kg). The pathologies included
diplegic CP, incomplete SCI (Level T6), and type I OI.
Fig. 2. Kinetic model of the lower crutch and crutch handle segments. Lower
crutch: F ctip and Mctip are the force and moment occurring at the crutch tip; mctip
and actip are the mass and the linear acceleration of the lower crutch segment; g is
the acceleration due to gravity; F lowerLC and M lowerLC are the known force and
moment at the lower load cell; rdist is the distance from the crutch tip and rprox is
the distance from the center of the lower load cell to the CoM of the lower
segment. Handle: F handle and M handle are the force and moment occurring at the
point of contact between the hand the crutch handle; mhandle and ahandle are the
mass and the linear acceleration of the handle segment; H_ handle is the rate of
change of angular momentum of the handle; F upperLC and M upperLC are the known
force and moment at the upper load cell; r dist is the distance from the CoM of the
lower loadcell, r prox is the distance from the Com of the upper load cell and r handle
is the distance from the point of contact of the hand on the crutch handle to the
CoM of the handle segment.
Y-axis Z-Axis
lc c cvirtualpoint  X Handle  Y Handle
2 jmlateral mmedial j  Y Handle lc c cvirtualpoint 


Lower crutch lc c þ cvirtualpoint mlateral mmedial
 Y LowerCrutch lc c Ctip c  X LowerCrutch  Y LowerCrutch
Ctipc ¼ 2 jmlateral mmedial j lc c Ctip c 
 
Cuff lc c þ 0:09  Y temp X Handle cosð151ÞY Handle sinð151Þ X Handle sinð151Þ þ Y Handle cosð151Þ Z Handle
2164 B.A. Slavens et al. / Journal of Biomechanics 44 (2011) 2162–2167

Fig. 3. Kinetic model of the crutch cuff and hand segments. Cuff: F cuff and M cuff are the unknown force and moment occurring at the point of contact of the cuff; mcuff and
acuff are the mass and the linear acceleration of the cuff segment; F upperLC and M upperLC are the known force and moment seen at the upper load cell; r dist is the distance from
the CoM of the upper load cell and rprox is the distance from the point of contact at the cuff to the CoM of the cuff segment. Hand: F wrist and M wrist are the unknown force
and moment occurring at the wrist joint; mwrist and awrist are the mass and the linear acceleration of the wrist segment; H_ wrist is the rate of change of angular momentum of
the wrist; F handle and Mhandle are the known force and moment at the point of contact between the hand and crutch handle; r dist is the distance from the point of contact
between the hand and crutch handle and rprox is the distance from the wrist joint center to the CoM of the wrist segment.

Fig. 4. Kinetic model of the forearm and upper arm segments. Forearm: F elbow and M elbow are the unknown force and moment occurring at the elbow joint; melbow and aelbow
are the mass and the linear acceleration of the forearm segment; H_ elbow is the rate of change of angular momentum of the elbow; F wrist and M wrist are the known force and
moment at the wrist joint; F cuff and M cuff are the known force and moment at the point of contact of the cuff; r dist is the distance from the wrist joint center, r prox is the
distance from the elbow joint center and r cuff is the distance from the point of contact at the cuff to the CoM of the elbow segment. Upperarm: F shoulder and M shoulder are the
unknown force and moment occurring at the shoulder joint; mhumreus and a humerus are the mass and the linear acceleration of the upper arm segment; H_ shoulder is the rate of
change of angular momentum of the shoulder; F elbow and M elbow are the known force and moment at the elbow joint; r dist is the distance from the elbow joint center and
r prox is the distance from the shoulder joint center to the CoM of the upper arm segment.

Table 2
Three-dimensional joint force and moment equations for the crutch, wrist, elbow, and shoulder, where F and M are the force and moment of the respective joint, m and a
are the mass and linear acceleration of the segment, g is the acceleration due to gravity, H_ is the rate of change of angular momentum, r is the distance from the distal
dist
part of the segment to the center of mass (CoM) of the segment and r prox is the distance from the proximal part of the segment to the CoM of the segment.

Joint Force Moment

Crutch tip F ctip ¼ mctip ðactip þ gÞF lowerLC M ctip ¼ H_ ctip M lowerLC r dist  F ctip r prox  F lowerLC
Crutch handle F handle ¼ mhandle ðahandle þ gÞ þ F lowerLC F upperLC M handle ¼ H_ handle þ M lowerLC M lowerLC M upperLC þ r dist  F lowerLC rproxt  F upperLC rhandle  F handle
Crutch cuff F cuff ¼ mcuff ðacuff þ gÞ þ F upperLC M cuff ¼ H_ cuff M upperLC þ r dist  F upperLC r proxt  F lcuff
Wrist F wrist ¼ mhand ðahand þ g Þþ F handle M wrist ¼ H_ wrist þ M handle þ rdist  F handle r prox  F wrist
Elbow F elbow ¼ mforearm ðaforearm þ gÞ þ F wrist þ F cuff M elbow ¼ H_ elbow þ M wrist þ Mcuff þ r dist  F wrist þ r cuff  F cuff r proxt  F elbow
Shoulder F shoulder ¼ mhumerus ðahumerus þ gÞ þ F elbow M shoulder ¼ H_ shoulder þ M elbow þ r dist  F elbow r proxt  F shoulder

Table 3 2.5. Data collection and analysis

Mean primary sensitivities of the transducers.

The instrumented Lofstrand crutches were adjusted to match the height and
Axis Primary sensitivity
cuff size of the subject’s current crutches. The subjects walked at a self-selected
pace and pattern with the crutches on a six-meter walkway while a 14-camera
Fx 2.4992 mV/Voltexc-N
Vicon motion analysis system (Vicon, Oxford, England) captured the data at
Fy 2.4973 mV/Voltexc-N
120 Hz. The gait cycle was normalized from heel strike to heel strike and averaged
Fz 0.6299 mV/Voltexc-N
over six trials.
Mx 134.1216 mV/Voltexc-N-m
Joint range of motion (ROM) was computed from the difference of the
My 134.6988 mV/Voltexc-N-m
maximum and minimum motions. Peak forces and peak moments were defined
Mz 90.7636 mV/Voltexc-N-m
as the absolute maximum throughout the gait cycle. Joint reaction forces were
 B.A. Slavens et al. / Journal of Biomechanics 44 (2011) 2162–2167 2165

normalized to percent body weight (%BW). Joint reaction moments were normal-
ized to percent body weight multiplied by height (%BWnH).
3. Results
moment. The largest SD of 0.55 N occurred during dynamic
evaluation of the right crutch resultant force. The right crutch
presented greater error and SD than the left crutch for dynamic
validation against the force plate (Table 4).

3.1. System evaluation 3.2. Kinematics

The % RMS error and SD of the four transducers were most The subject with OI presented the highest cadence (103 steps/
notable for the sagittal plane inferior (vertical) force and fore tilt min) and walking speed (0.78 m/s). The subject with SCI pre-
(flexion) moment during static validation (Table 4). The greatest sented the lowest walking speed (0.53 m/s), shortest stride length
error of 5.20% RMS was in the lower left transducer fore tilt (0.64 m) and longest stance duration (69%). The subject with CP
demonstrated the lowest cadence (78 steps/min), longest stride
length (0.95 m) and shortest stance duration (56%).
All subjects used a two-point reciprocal gait pattern for
Table 4
ambulation. Among all subjects, the crutches were titled laterally
Static and dynamic evaluation of accuracy and standard deviation (SD) of the
crutch system.
when the crutch support phase started. The wrists and elbows
remained extended and flexed, respectively, throughout the gait
Evaluation method Root mean square Standard cycle. The shoulder presented the greatest flexion motion during
error (%) deviation (N) the start of crutch support, which was accompanied by abduction
to assist in clearance of the crutch tip.
Right Left Right Left
The elbow had the greatest ROM and the wrist had the
Static evaluation smallest ROM of the UE joints. The subject with CP demonstrated
Upper transducer: inferior force 4.06 1.11 0.11 0.04 the largest ROMs among the shoulder, elbow, and crutches in
Upper transducer: fore tilt moment 4.09 4.76 0.001 0.05
three-dimensions, except that the subject with SCI showed the
Lower transducer: inferior force 0.84 0.90 0.05 0.05
Lower transducer: fore tilt moment 3.74 5.20 0.03 0.001 largest ROMs in the axial plane of the shoulder and the sagittal
plane of the crutches. The subject with SCI demonstrated the
Dynamic evaluation
Resultant force 2.81 1.43 0.55 0.29
greatest wrist joint ROM in the coronal and axial planes, and the
subject with OI showed the largest wrist ROM in the sagittal

Fig. 5. Wrist, elbow, and shoulder joint reaction forces for the primary weight bearing extremity. Forces are normalized to percent body weight (%BW). Black solid: subject
with CP, gray solid: subject with SCI, and black dashed-dot: subject with OI.
2166 B.A. Slavens et al. / Journal of Biomechanics 44 (2011) 2162–2167
plane. The subject with OI also demonstrated the smallest ROMs
among the UE joints in all planes, except for the sagittal plane of
the wrist and crutches, and the coronal plane of the wrist.
3.3. Kinetics
Each subject presented a primary weight bearing extremity,
which showed higher joint reaction forces and moments than the
contralateral extremity (Figs. 5 and 6).
3.3.1. Forces
The greatest forces were observed in the posterior direction of
the wrist (Fig. 5). The subject with CP experienced the highest
forces in all planes for the majority of joints of the UEs and
crutches. Exceptions included the subject with SCI demonstrating
the largest forces at the cuff in the anterior direction and the
subject with OI having the largest shoulder forces in the posterior
direction.
3.3.2. Moments
The greatest moments were observed in the shoulder during
flexion (Fig. 6). The subject with SCI presented the greatest
moments for all planes of the shoulder, elbow extension, and
crutch tip extension moments. The subject with CP displayed the
highest moments for all other joints, including all planes of the
crutch handle, cuff, and wrist.
Fig. 6. Wrist, elbow, and shoulder joint reaction moments for the primary weight bearing extremity. Moments are normalized to percent body weight multiplied by height
(% BWnH). Black solid: subject with CP, gray solid: subject with SCI, and black dashed-dot: subject with OI.
4. Discussion
This study presents an inverse dynamics model for biomechani-
cal assessment of UE kinetics during Lofstrand crutch-assisted gait.
Results of this study support the use of this technically validated, ISB
compliant system to evaluate ambulation patterns. Key features of
the model included validation and increased capability for quantify-
ing the wrist, elbow, and shoulder dynamics based on measurement
of crutch handle and cuff forces and moments.
Inferior force and fore tilt moment were the most notable
accuracy metrics. Error magnitudes were within range of prior
studies of crutch-assisted gait kinetics (Requejo et al., 2005;
Slavens et al., 2010).
The model was demonstrated with subjects with CP, SCI, and
OI to quantify UE joint dynamics. The kinematic model was
unique while incorporating important aspects, such as ISB recom-
mendations, from previous models (Poppen and Walker, 1976;
Veeger et al., 1997; Schmidt et al., 1999; Rab et al., 2002; Roux
et al., 2002; Nguyen and Baker, 2004; Wu et al., 2005; Slavens
et al., 2010). The model was capable of detecting distinctive joint
kinetic profiles among the three pediatric pathologies.
The subject with OI presented the smallest superiorly acting
compression forces and flexion moments at the shoulder. The
subject with CP showed the lowest cadence and highest forces at
the shoulder. The subject with SCI demonstrated the highest
moments at the shoulder. At the current stage, however, no
generalizations between pathologies can be made.
This study may form the basis for future studies linking crutch
kinetics to injury. The clinical motivation for this study is directly
linked to injury prevention, improved function and longer term
 B.A. Slavens et al. / Journal of Biomechanics 44 (2011) 2162–2167
transitional care of crutch users, such as children with CP, SCI, OI,
and myelomeningocele. Large compressive forces quantified in
this study are a concern for long-term assistive device users, as it
has been shown that high joint demands may lead to UE pain and
pathology (Opila et al., 1987; Sie et al., 1992; Lal, 1998; Sala et al.,
1998; Mercer et al., 2006; Collinger et al., 2008). Further applica-
tion of this system may offer valuable insight for crutch prescrip-
tion, therapeutic walking strategies, and long-term usage effects.
Conflict of interest statement
There are no personal or financial conflicts of interest asso-
ciated with this work.
Acknowledgments
This work was supported in part by the Dr. Ralph and Marian
Falk Medical Trust and NIDRR Advanced Rehabilitation Research
Training (ARRT) Grant H133P080005. We would like to thank
Adam Graf, MS, Joe Krzak, PT, and Kathryn Reiners for their
contributions to this study.
References
Collinger, J.L., Boninger, M.L., Koontz, A.M., Price, R., Sisto, S.A., Tolerico, M.L.,
Cooper, R.A., 2008. Shoulder biomechanics during the push phase of wheel-
chair propulsion: a multisite study of persons with paraplegia. Archives of
Physical Medicine & Rehabilitation 89 (4), 667–676.
Haubert, L.L., Gutierrez, D.D., Newsam, C.J., Gronley, J.K., Mulroy, S.J., Perry, J., 2006.
A comparison of shoulder joint forces during ambulation with crutches versus
a walker in persons with incomplete spinal cord injury. Archives of Physical
Medicine & Rehabilitation 87, 63–70.
Kaye, H.S., Kang, T., LaPlante, M.P., 2000. Mobility Device Use in the United States.
Disability Statistics Report (14). Washington, D.C., U.S. Department of Educa-
tion, National Institute on Disability and Rehabilitation Research.
Lal, S., 1998. Premature degenerative shoulder changes in spinal cord injury
patients. Spinal Cord 36 (3), 186–189.
2167
Melis, E.H., Torres-Moreno, R., Barbeau, H., Lemaire, E.D., 1999. Analysis of
assisted-gait characteristics in persons with incomplete spinal cord injury.
Spinal Cord 37 (6), 430–439.
Mercer, J.L., Boninger, M., Koontz, A., Ren, D., Dyson-Hudson, T., Cooper, R., 2006.
Shoulder joint kinetics and pathology in manual wheelchair users. Clinical
Biomechanics 21 (8), 781–789.
Nguyen, T.C., Baker, R., 2004. Two methods of calculating thorax kinematics in
children with myelomeningocele. Clinical Biomechanics 19 (10), 1060–1065.
Opila, K.A., Nicol, A.C., Paul, J.P., 1987. Upper limb loadings of gait with crutches.
Journal of Biomechanical Engineering 109 (4), 285–290.
Poppen, N.K., Walker, P.S., 1976. Normal and abnormal motion of the shoulder.
Journal of Bone and Joint Surgery, America 58 (2), 195–201.
Rab, G., Petuskey, K., Bagley, A., 2002. A method for determination of upper
extremity kinematics. Gait & Posture 15 (2), 113–119.
Requejo, P.S., Wahl, D.P., Bontrager, E.L., Newsam, C.J., Gronley, J.K., Mulroy, S.J.,
Perry, J., 2005. Upper extremity kinetics during Lofstrand crutch-assisted gait.
Medical Engineering & Physics 27 (1), 19–29.
Roux, E., Bouilland, S., Godillon-Maquinghen, A.P., Bouttens, D., 2002. Evaluation of
the global optimization method within the upper limb kinematics analysis.
Journal of Biomechanics 35 (9), 1279–1283.
Sala, D.A., Leva, L.M., Kummer, F.J., Grant, A.D., 1998. Crutch handle design: effect
on palmar loads during ambulation. Archives of Physical Medicine & Rehabi-
litation 79 (11), 1473–1476.
Schmidt, R., Disselhorst-Klug, C., Silny, J., Rau, G., 1999. A marker-based measure-
ment procedure for unconstrained wrist and elbow motions. Journal of
Biomechanics 32 (6), 615–621.
Sie, I.H., Waters, R.L., Adkins, R.H., Gellman, H., 1992. Upper extremity pain in the
postrehabilitation spinal cord injured patient. Archives of Physical Medicine
and Rehabilitation 73 (1), 44–48.
Slavens, B.A., Sturm, P.F., Bajorunaite, R., Harris, G.F., 2009. Upper extremity
dynamics during Lofstrand crutch-assisted gait in children with myelomenin-
gocele. Gait and Posture 30, 511–517.
Slavens, B.A., Sturm, P.F., Harris, G.F., 2010. Upper extremity inverse dynamics
model for crutch-assisted gait assessment. Journal of Biomechanics 43 (10),
2026–2031.
Veeger, H.E., Yu, B., Rozendal, R.H., K.N., 1997. Parameters for modeling the upper
extremity. Journal of Biomechanics 30 (6), 647–652.
Wu, G., van der Helm, F.C., Veeger, H.E., Makhsous, M., Van Roy, P., Anglin, C.,
Nagels, J., Karduna, A.R., McQuade, K., Wang, X., Werner, F.W., Buchholz, B.,
2005. ISB recommendation on definitions of joint coordinate systems of
various joints for the reporting of human joint motion—part II: shoulder,
elbow, wrist and hand. Journal of Biomechanics 38 (5), 981–992.
Zatsiorsky, V.M., 2002. Joint torques and forces: the inverse problem of dynamics.
In: Robertson, L.D. (Ed.), Kinetics of Human Motion, vol. 2. Human Kinetics,
Champaign, IL, pp. 365–454.