w21-93Jr19~1uJI-0775 502.

wAJ

NORMATIVE MODEL OF HUMAN HAND

FOR BIOMECHANICAL ANALYSIS.

K. N. AN, E. Y. CHAO,W. P. COONEY,III and R. L. LINSCHEID

Orthopedic Biomechanics Laboratory, Mayo Clinic/Mayo Foundation, Rochester, Minnesota, U.S.A.

Abstract - A three-dimensional normative model of the hand was established, based on the averaged
anatomical structure of ten normal hand specimens. The joint and tendon orientations were defined from
biplanar X-ray films. The configurations of the hand at the joints were describe-d by the classic Eulerian
angles. Force potential and moment potential parameters were utilized to describe the contribution of each
tendon in the force analysis. The mean values of these two parameters were used to compute the designated
two points for each tendon at each joint in the normative model. With appropriate coordinate
transformations at the joints, the tendon locations and excursions under various functional configurations
can be commuted. This model can be used to nerform force and motion analyses for both normal and
pathological hands.

INTRODUCTION
The complexities of the function and anatomy of
human hand have long been recognized (Bunnell,
1948 : Eyler and Markee, 1954 ; Haines, 1951; Lands-
meer, 1949, 1955; Smith, 1974; Stack, 1962). From a
biomechanical standpoint, the human hand can be
considered as a linkage system of intercalated bony
segments. The joints between each phalanx are span-
ned by ligaments, tendons and muscles. With the
contraction of muscles, these joints can be moved in a
characteristic manner constrained by the interposing
soft tissues and the bony articulation. If the motion is
resisted, functional strength in the forms of pinch and
grasp can result. In the hand, most of the tendons span
the joint and continue their course over one or more
joints, thus forming a bi-articular or poly-articular
system.
The functional anatomy of the spatial relationships
between these tendons and muscles and their as-
sociated joints have been extensively studied by
Landsmeer (1955, 1961). Landsmeer also @posed a
series of models to represent the various nianners in
which tendons bridge the associated joints. Several
biomechanical models of various aspects .of hand
function have also been developed (Smith, 1964; Flatt,
1971). However, these studies show either a lack of
quantitative description or that the information is
restricted in only two dimensions. The objective of the
present study is to establish a workable model in a
three-dimensional manner, based on the direct and
careful measurements of 10 normal specimens. Such a
model can easily be utilized in the study of hand
motion or force analysis of hand under various
functional activities. Systematic variation of this
model based on clinical examinations can provide an
effective means to study the internal force in the hand
suffering pathological changes. The present model and
the associated data can help other investigators who
l Received 17 January 1978.
wish to pursue biomechanical analysis of the hand but
are limited by clinical facilities and resources to
perform detailed anatomical studies.
This model has two limitations. First, it does not
provide information concerning the instantaneous
center of motion pathway for all joints in the hand. As
a result, it can only be used to quantitate relative
motion between connecting segments. No articulating
surface motion pattern of the joint can be analyzed.
Secondly, the model does not include the detailed
ligamentous structure of hand joints, which disallows
force analysis at the capsular level. Such forces are
lumped together as joint constraints which can later be
decomposed through equipollent force analysis when
the quantitative information concerning the capsulo-
ligamentous complex becomes available.
EXPERIMENTAL METHOD
.In the study of the three-dimensional locations of
tendons with respect to the bony segments at the joint,
10 fresh cadaver specimens were used. The experimen-
tal techniques are the same as those described in the
previous paper of Chao (1976). By the use of standard
surgical incisions, a small segment of each tendon and
muscle near the joint was carefully exposed. Markers
made of different grades of surgical wire with varying
lengths were inserted into each tendon and muscle at
the sites immediately proximal and distal to the joint.
Finger and thumb specimens with wires embedded
were then subjected to biplanar X-ray exposure’on
small dental film. A specially designed holder for
alignment of the X-ray beam and film with the
dorsal-volar or radial-ulnar plane of the finger or
thumb specimen was used so that accurate. AP and
lateral X-rays could be obtained. The markers made
from different grades of surgical wire with varying
lengths were inserted into the center of each tendon
and muscle at the sites proximal and distal to the joint.
The proper position for the insertion of the markers

775
716 K. N. AN, E. Y. CHAO, W. P. COOXY,III and R. L. LINSCHEID

Table 1. Tendons and muscles involved in hand function In addition to the invasive method used. several
noninvasive techniques have been tested in order to
Hand element Joint Tendons and muscles provide means for in ciro studies. These methods were :
(1) high resolution and slow scanning ultrasound
Terminal extensor (TE)
DIP technique; (2) tomographic xerography technique:
Flexor profundus (FP)
and (3) EM1 Body Scanner. Cadaver hand specimens
Extensor slip (ES)
Radial band (RB) were first used to examine the resolution of each
Finger PIP
Ulnar band (UB) method. After the noninvasive study, the specimens
Flexor sublimis (FS) were sliced at the levels of scan for quantitative
Long extensor (LE) comparison. The equipment used for the ultrasound
Radial interosseous (RI) scan and the computerized tomography are illustrated
MP
Ulnar interosseous (UI) in Figs. 2 and 3.
Lumbrical (LU)

Flexor pollicis longus (FPL) FORMULATION OF THE MODEL

IP Extensor pollicis longus (EPL)
Abductor pollicis brevis (APB)
Flexor po&is brevis (FPB)
Thumb MCP
Adductor nollicis (ADD)
Extensor &llicis drevis (EPB)
Opponens pollicis (OPP)
CMC Abductor pollicis longus (APL)
was determined with reference to the pulley systems,
retinacular ligaments and transverse bands which
constrain the tendons and muscles at the joint. These
constraint ligaments and pulleys minimize bowstring-
ing with flexion and extension of the joints, such that a
straight line relationship exists along the tendon
markers and best represents the direction of the tendon
force. The use of different sizes and lengths of wires
facilitated easy identification of the markers on the X-
ray films. The selection of wire size depends on the
diameter of the tendon, in order to make insertion
possible. For flexor tendons, small incisions in the
tendon sheath were made for the insertion of wire-s so
that the pulley systems were not damaged. Careful
closure was then performed. This procedure was
repeated for all the tendons at their associated joints,
as defined in Table 1. The sizes and grades of Kirschner
wires used for this study as shown in Table 2. The
enlarged prints ofthe two X-ray films (Fig. 1) were then
utilized for the analysis of the tendon locations for each
joint.
In order to establish a mathematically workable
model, six Cartesian coordinate systems, as shown in
Fig. 4, were established. These coordinates were
established to define the locations and orientations of
tendons and can also be utilized to describe the joint
configuration. There are two coordinate systems for
both the middle and proximal phalanges and only one
system for both the distal phalanx and metacarpal.
The primary systkms (numhrs 2,4 and 6) are located
at the approximate center of rotation of the phal-
angeal and metacarpal heads, and the secondary
systems (numbers 1, 3 and 5) are a translation of the
proximal systems to the centers of the concave arti-
cular surfaces. The x-axis is projected along the
phalangeal or the metacarpal shaft, passing from the
center of rotation to the center of the concave articular
surface at the proximal end. The y-axis is projected
dorsally and the z-axis is projected radially for the
right hand and ulnarly for the left hand.
h 71
Y. I I PIP

Table 2. Grades and sizes of Kirschner wire marker for

tendon location study

Thumb Finger
Mu&e Muscle
and Size and Size
tendon Grade (mm) tendon Grade (mm)

EPL 28 10 TE 28 5
FPL 20 10 FP 30 5
ADD 23 10 ES 28 5
OPP 25 10 :! 23
26 5
DI 30 10
EPB 30 5 FS 10
FPB 20 5 LE :: 5
APB 23 5 RI 28 10
APL 25 5 UI 23 10
LU 10 10
Fig. 4. Coordinate systems used to define tendon locations.
Fig. I. Biplanar X-ray views of finger tendons and muscles located by surgical wires

Fig. 2. Ultrasonic scanning equipment of Stanford Research Institute (SRI).

777
 c
c _
.
1

Fig. 3. Hand specimen in the EMI CT-5000 Body Scanner.

778
Fig. 8. SRI Ultrasonic Scanner image of a cadaver hand specimen.

779
-5000 Body Scanner image of a cadaver hand at the level of the distal metacarpal head.
 Normative model of human hand for biomechanical analysis
With the assistance of these coordinate systems, the
tendon locations in three-dimensional space can be
determined and measured. At each joint, tendon is
represented by a straight line joining two points
designated by the markers ; one on the distal side of the
joint and the other on the proximal side of the joint.
The three-dimensional locations ofthese two points on
each tendon were expressed, based on the distal and
proximal coordinate systems.
In the analysis of data among 10 specimens, the
exact tendon locations with respect to each localized
coordinate system were expressed in dimensionless
terms by normalizing their coordinates according to
the length of the middle phalanx of a finger, to avoid
the anthropometric variations. In the case of a thumb,
the length of the proximal phalanx was used as a
common denominator.
The coordinates for each marker in the 10 cadaver
specimens were not used in the averaging process,
since not all of the markers in the tendons for each
specimen have identical positions, and this would
introduce significant variations in the final data.
Instead, two parameters used to describe the orien-
tation and location of each tendon for the final model
were determined first from raw da’ta. These two
parameters are ‘force potential’ and ‘moment poten-
tial’. The force potential is expressed in terms of the
directional cosine of a tendon with respect to the distal
system, and the moment potential specifies the mo-
ment arm of the tendon in regard to the joint center
and in the direction ofeach coordinate axis of the distal
system. Obviously, force potential provides the contri-
bution of a particular tendon in generating joint
constraint forces and the moment potential specifies
the functional moment of each tendon in rotating the
joint at three mutually perpendicular directions. The
mean values of these parameters among the 10 speci-
mens was then used to construct the normative model.
The application of force and moment potentials can
be demonstrated by the static force analysis of the
hand under isometric functions. In deriving the
equilibrium equations about the center of rotation of
the joint, the following forms are usually used:
YP
Fig. 5. Two points on each tendon of the joint have fixed relationships with their coordinate systems.
781
Force equations
Z:I,F,+C,+R,=O
~BiFi+C,+R,=O (1)
Z;‘iFifC*+R*=O
Moment equations
ZUiFi+ M,+ r,=O
ZbiFi+Myf Ty=O (2)
~Cifi + M; + r:=",
where
zi, fii, ‘ii =
force potential parameters,
oi, bi, ci =
moment potential parameters,
C,, C,, C, = unknown joint constraint forces,
M,, M,, M, = unknown joint constraint moments,
Fi =
unknown tendon or muscle forces,
R,, R,, R, = externally applied forces, and
7,, T7, 7, = externally applied moments.
Based on the parameters in the model, static force
analysis can easily be applied for the determination of
the unknown tendon forces and joint constraint forces
and moments. On the other hand, if kinetic analysis is
required, the kinematic and inertial contributions to
the problem can be conveniently defined, based on the
established coordinate system. The force components
can be formulated similar to that illustrated in equa-
tions (1) and (2). When externally applied forces are
included in the system, they can be incorporated into
the static or dynamic equation through proper coor-
dinate transformation.
The coofdinates of these two points of each tendon
at the joint, as shown in Fig. 5, were determined from
th&e averaged values of force and moment potential
parameters. Several steps were involved in this re-
duction procedure. First, the force potential para-
meters were normalized such that these parameters of
each tendon form a unit force vector. Second, the
average moment potential parameters were adjusted
minutely, if necessary, so that the moment arm and its
unit force vector of each tendon were orthogonal.
Finally, these normalized and adjusted force and
782 K. N. AS. E. Y. CHAO, W. P. COO%~Y,
III and R. L. LINSCHEID
: Floxion-Extodon (1st rotation- Zp)
8: Radlouhw DoAtioo (2nd rotation- 1’)
u V
$: Prwutkn-Swpkmtion (3rd rot&ion - 4)
Fig. 6. Eulerian angles used to define the orientations of
finger digits.
moment potential parameters were ready for calculat-
ing the coordinates of points for tendons. Two points
of each set, one on the distal side of the joint and the
other on the proximal side of the joint, were expressed
with respect to the distal and proximal coordinate
system at each joint, respectively. It was assumed that
the tendons were constrained to remain fixed in
relationship with the proximal and distal coordinate
systems by pulleys and ligaments. Therefore, the
results obtained at the neutral position could later be
used for any finger configuration, as shown in Fig. 5,
with the appropriate coordinate transformations.
When the finger is in the functiona configuration
other than the neutral position, the proximal coor-
dinate system at a particular joint can be related to the
distal system through simple rotational and trans-
lational transformation. The orientation between two
digits was defined by the classic Eulerian angles. These
angles are flexion-extension ($), radio-ulnar deviation
(0) and axial rotation (II/).The rotational sequence of
these angles, as related to both the distal (moving
system) and proximal (fixed system) coordinate SYS-
terns at a joint is illustrated in Fig. 6. With the above
described Eulerian angles, the coordinate of a tendon
point or the components of a force vector as defined in
either system can be related by the following transfor-
mation equation :
in which
X,,, Y”, Z,, = coordinates of a lendon point or com-
ponents of a vector measured with re-
spect to the distal system,
Xp, Y,+Z, = coordinates of a tendon point or com-
ponents of a vector measured with re-
spect to the proximal system,
X,,, YO,Z, = coordinates of the origin of the proximal
system expressed in the distal system,
s = sine, and
c = cosine.
The transformation equation was utilized in the
force analysis to transform the tendon points and
externally applied force vectors into the appropriate
coordinate system where the free body was taken.
RESULTS
The linkagejtructure of the bony elements was
related with th=e coordinate systems. The distance
between the origins of two coordinate systems, i.e. the
center of rotation and the center of the concave
articular surface of the joint, was used to specify the
location of the six coordinate systems. Figure 7 shows
the average and the standard deviation of these
distances. The dimensions are all normalized with
respect to the distance between O1 and 0, ; that is, the
distance between the center of rotation of the DIP joint
and the center of the concave surface of the PIP joint.
The normalized force potential and moment poten-
tial parameters of the hand model in neutral position
are tabulated in Tables 3-6 for index, long. ring and
little fingers. The sign convention used for the force
potential and moment potential in Tables 3-6 is
illustrated in Table 7.
The two representative points of each tendon at its
associated joint were summarized in the dimensionless
form in Tables 8-11. As mentioned previously, the
distal point and proximal point of a tendon are
measured with respect to the distal and proximal
systems at the joint, respectively.
The results of utilizing noninvasive methods to
provide detailed cross-sectional images of the hand are
shown in Figs. 8 and 9. The image of the whole hand in
the AP view, as shown in Fig. 8, was obtained by the
SRI Ultrasonic Camera. The hand specimens were
imbedded in the water tank during the scanning
process. Despite a high resolution and slow scanning
technique, the ultrasound scanning technique still did
not provide satisfactory results, simply because bone
tissue has a very high reelection rate compared with
tendons and muscles, and this causes the obscureness
in the image.
I
Figure 9 is an actual cross-sectional scanning image
of the hand specimen at the level of the metacarpophal-
angeal joint obtained by EMI Body Scanner. In this
picture, the flexor tendons on the volar side are
obvious. The image on the dorsal side of the joint is not
clear enough, however, to separate the extensor ten-
dons from the bone, since they are closely packed
 Normative model of human hand for biomechanical analysis 783

Table 3. Force and moment potential parameters of tendons in normal index finger under neutral position
(mean value of 10 specimens)

Force Moment
Joint Tendon z B Y a b C

DIP EXT 0.998 - 0.059 -0.001 -0.001 -0.021 - 0.202

PROF 0.977 -0.212 0.030 0.002 0.036 0.201
PROF 0.992 -0.126 0.032 - 0.007 0.036 0.359
RA-B 0.992 -0.046 0.120 0.035 0.250 -0.199
PIP UL-B 0.995 - 0.069 - 0.078 - 0.032 - 0.328 -0.115
SUBL 0.994 -0.113 -0.006 0.002 0.003 0.286
SLIP l.OQO 0.012 -0.011 -0.002 -0.037 - 0.276
PROF 0.953 -0.283 -0.104 0.050 -0.022 0.518
SUBL 0.964 -0.260 -0.061 0.015 - 0.084 0.605
RA-I 0.933 -0.353 0.067 0.170 0.490 0.222
MP LUMB 0.847 -0.495 0.193 0.181 0.501 0.488
UL-I 0.936 -0.350 -0.033 -0.156 - 0.447 0.325
LEXT 0.989 0.120 -0.085 - 0.036 - 0.036 -0.476

Table 4. Force and moment potential parameters of tendons in normal long finger under neutral position
(mean value of IO specimens)

Force Moment
Joint Tendon 2 B Y a b C

DIP EXT 0.997 0.074 0.026 0.006 - 0.026 -0.169

PROF 0.976 -0.216 -0.000 0.010 0.044 0.185
PROF 0.994 -0.111 0.008 -0.001 0.014 0.314
RA-B 0.997 - 0.070 0.035 0.02 1 0.246 -0.114
PIP UL-B 0.984 -0.168 -0.060 - 0.05 1 - 0.268 -0.092
SUBL 0.994 -0.104 -0.029 0.007 0.002 0.232
SLIP 0.999 - 0.027 0.025 0.006 -0.018 -0.243
PROF 0.987 -0.159 -0.033 0.015 0.010 0.406
SUBL 0.971 - 0.237 -0.044 0.029 0.026 0.507
MP RA-I 0.881 -0.431 0.192 0.126 0.414 0.351
LUMB 0.875 - 0.477 0.084 0.154 0.363 0.461
UL-I 0.96 1 - 0.276 -0.006 - 0.096 -0.337 0.152
LEXT 0.979 0.204 -0.022 -0.001 -0.038 -0.409
784 K. N. AN, E. Y. CHAO.W. P. Coo\~, III and R. L. LIXSCHEID

Table 5. Force and moment potential parameters of tendons in normal ring finger under neutral position
(mean value of 10 specimens)

Force Moment
Joint Tendon 1 B Y (I b C

EXT 0.999 - 0.024 0.049 0.006 - 0.020 -0.141

DIP -0.216 0.009 0.188
PROF 0.975 0.052 -0.008
PROF 0.998 - 0.045 0.040 -0.011 0.007 0.285
RA-B 0.982 - 0.096 0.164 0.038 0.226 - 0.098
PIP UL-B 0.984 -0.178 -0.007 - 0.042 -0.232 - 0.074
SUBL l.ooo - 0.023 0.005 -0.001 -0.006 0.231
SLIP 0.993 - 0.079 0.083 0.016 -0.OQO -0.188
PROF 0.969 - 0.248 0.014 0.003 0.036 0.409
SUBL 0.970 - 0.243 0.001 0.014 0.056 0.456
RA-I 0.994 -0.102 0.049 0.029 0.350 0.141
MP -0.310 0.117 0.303
LUMB 0.944 0.076 0.346
UL-I 0.938 - 0.347 - 0.024 -0.116 -0.328 0.218
LEXT 0.140 0.034 0.025 -0.375

Table 6. Force and moment potential parameters of tendons in normal littlefinger under neutral position
(mean value of 10 specimens)

Force Moment
Joint Tendon a B 7 a b C

DIP EXT 0.977 0.060 0.044 0.005 - 0.068 - 0.206

PROF 0.99 1 -0.102 0.089 -0.021 - 0.034 0.200
PROF 0.988 -0.146 0.048 -0.007 0.070 0.366
RA-B 0.99 I -0.113 0.066 0.044 0.293 -0.166
PIP UL-B 0.999 -0.016 0.044 0.001 -0.309 -0.124
SUBL 0.990 -0.138 0.012 O.O@l 0.029 0.290
SLIP 0.994 -0.059 0.087 0.023 -0.006 - 0.268
PROF 0.974 -0.124 0.191 - 0.065 0.249 0.492
SUBL 0.944 - 0.309 0.112 0.002 0.199 0.530
MP RA-I 0.972 -0.216 0.092 0.099 0.474 0.070
LUMB 0.949 -0.313 0.032 0.136 0.441 0.29 I
UL-I 0.958 - 0.262 -0.113 -0.104 -0.495 0.27 I
LEXT 0.969 0.198 0.146 0.064 0.057 - 0.499

Table 7. Sign convention used for anatomical. variation in fingers

Component Sign Force or moment direction

Axial + Axial compression

Constraint + Dorsal subluxation
Dorsovolar - Volar subluxation
force
variation + Radial dislocation
Radioulnar - Ulnar dislocation

•F Radial twist (pronation)

Axial - Ulnar twist (supination)
Moment arm + Radial bending
Radioulnar - Ulnar bending
variation
+ Flexion
Flexion-extension - Extension
 Normative model of human hand for biomechanical analysis 785

Table 8. Tendon locations of index finger

Distal point Proximal point

Joint Tendon X Y Z X 1 Z

TE 0.000 0.217 -0.019 0.000 0.205 - 0.0’0

DIP 0.000 -0.159 0.028 0.300 - 0.270 0.056
FP
FP -0.210 - 0.305 0.027 0.400 -0.375 0.039
RB -0.110 0.2 19 0.207 0.100 0.202 0.261
PIP UB -0.110 0.149 -0.303 0.100 0.116 - 0.347
FS -0.210 -0.240 0.005 0.400 - 0.294 0.003
ES -0.040 0.293 -0.036 O.CiJO 0.280 - 0.04
FP -0.100 - 0.380 0.300 - 0.605 -0.033
FS -0.100 - 0.483 -0.033 0.300 - 0.685 - 0.082
RI -0.300 0.020 0.478 0.400 -0.476 0.6 16
MP -0.741 0.679
LU -0.300 -0.146 0.444 0.400
UI -0.300 -0.067 -0.462 0.400 -0.4-W - 0.477
LE 0.000 0.455 -0.013 0.000 0.506 - 0.021

Table 9. Tendon locations of middle finger

Distal point Proximal point

Joint Tendon X Y Z X Y Z

DIP TE -0.05 0.154 -0.022 0.000 0.168 -0.015

FP -0.05 -0.141 0.032 0.300 - 0.254 0.033
FP -0.28 -0.273 -0.004 0.400 -0.266 -0.004
RB -0.18 0.172 0.227 0.100 0.1 0.242
PIP UB -0.18 0.161 -0.247 0.100 0.076 -0.279
FS -0.28 - 0.207 0.001 0.400 -0.217 -0.016
ES -0.03 0.247 - 0.024 0.000 0.236 -0.019
FP -0.18 -0.330 0.023 0.300 - 0.470 0.012
FS -0.18 -0.393, 0.039 0.300 -0.574 0.019
MP RI -0.38 -0.038 0.331 0.400 -0.565 0.471
LU -0.38 -0.140 0.328 0.400 -0.623 0.422
UI -0.38 0.036 -0.357 0.264 -0.358
LE -0.08 0.322 -0.018 0.393 - 0.039

Table 10. Tendon locations of ring finger

Distal point Proximal point

Joint Tendon X Y Z X Y Z

TE - 0.070 0.144 - 0.030 0.000 0.141 - 0.025

DIP
FP - 0.070 -0.145 0.001 0.300 -0.253 0.032
FP - 0.290 - 0.278 -0.016 0.400 -0.291 O.blO
RB -0.190 0.140 0.163 0.100 0.110 0.237
PIP UB -0.190 0.141 -0.228 0.100 0.072 - 0.246
FS - 0.290 - 0.234 -0.009 0.400 - 0.228 -0.00-t
ES -0.040 0.208 -0.016 0.000 0.202 0.006
FP -0.210 - 0.278 0.006 0.300 -0.463 0.012
FS -0.210 -0.338 0.038 0.300 -0.517 0.036
MP RI -0.410 0.000 0.275 0.400 -0.211 0.283
LU -0.410 - 0.059 0.165 0.400 -0.415 0.213
UI -0.410 0.012 -0.199 0.400 -0.330 - 0.207
LE -0.110 0.326 0.002 O.OQO 0.388 O.O!c!
 786 K. N. AN, E. Y. CHAO,W. P. COO>~Y,
III and R. L. LINSCHEID

Table 11. Tendon locations of little linger

Distal point Proximal Point

Joint Tendon X Y Z X Y Z

DIP TE -0.010 0.199 -0.132 0.000 0.187 -0.119

FP -0.010 -0.187 -0.075 0.030 -0.236 - 0.020
FP -0.210 -0.317 0.017 0.400 -0.359 0.058
RB -0.110 0.203 0.184 0.100 0.182 0.224
PIP UB -0.110 0.110 -0.321 0.100 0.113 - 0.293
FS -0.210 -0.271 0.004 0.400 -0.310 -0.008
ES -0.040 0.272 -0.012 0.000 0.257 0.006
FP -0.060 -0.413 0.076 0.300 -0.584 0.112
FS - 0.060 - 0.489 0.117 0.300 -0.653 0.129
MP RI - 0.260 0.066 0.430 0.400 -0.145 0.473
LU - 0.260 - 0.056 0.423 0.400 -0.363 0.469
UI -0.260 -0.100 -0.502 0.400 -0.304 - 0.590
LE 0.040 0.414 0.035 0.000 0.477 0.111

together anatomically. In addition to the radial and ulnar lateral movement

DISCUsslOw
Tendon locations and orientations in the finger were
obtained by experimental studies of 10 normal hand
specimens. The results presented are the average of
these raw data. The measurements of tendon locations
were normalized with respect to the middle phalanx
length (0203) of the corresponding finger before
averaging. This normalization concept was based on
the assumption that normal hands maintain similar
anatomical structure and dimensional proportion,
regardless of their physical size. With this normali-
zation, the anthropometric variation was avoided. The
small variation in the normalized data justifies this
process.
The results show small variations in those pre-
dominant components, while significant variation in
secondary or minor components was seen among the
specimens. This indicates that tendons and muscles
have more than one function. In addition to their
predominant functions, each muscle also contributes
to other functions. For example, in addition to flexion
and extension, finger flexors and long extensor ten-
dons are also adductor muscles for all the fingers at the
MP joints. However, these secondary contributions of
function vary largely among the specimen population.
at the MP joint, the intrinsic muscles contribute to
flexion at the MP joint. For all four fingers, the
lumbricals have more Rexion moment potential than
the interossei. Various amounts of moment poten-
tial for axial rotation were also noticed within each
individual tendon and muscle. However, they were not
as significant as those for flexion-extension and
abduction-adduction.
For the purpose of verification. two-dimensional
models (Fig. IO), as proposed by Landsmeer (1961).
were used to compare with the results produced by the
present model. Model I describes the situation of
tendons which follows the curvature of the articular
surface. This model is particularly designed for the
extensors. Model II states that the tendon runs
through a loop which is freely movable around the
axis, and the system is assumed to be free of friction.
Therefore, the two parts of the tendon run parallel with
the long axis ofthe bone and then the sling will take up
a position along the bisection of the joint angle. This
model describes the intrinsic muscle system. The last
model, Model III, is such that the tendon runs in a
tendon sheath which holds it firmly in a constant
position against the shaft of the bone, but allows the
tendon to curve smoothly in the joint angle. In both
Model II and Model III, the symmetries of the tendon
pulley systems at the joints were assumed in order to

[MODEL II [MODEL II] [MOM1 Ill]

Fig. 10. Landsmeer’s two-dimensional model of tendons bridging a joint.

 Normative model of human hand for biomechanical analysis 787

derive the relationships of tendon excursions and . .7

. *
moment arms. In the present normative model, ten-
. l . l.
dons at each joint were represented by two points. . *
. .
Between the two points, the simple ‘bow string’ . . .
_ -9.6
conditions were assumed. This model allows us to
derive the tendon excursions and moment arm in a
three-dimensional sense without considering the
symmetry.
Tendon excursions and moment arms calculated
based on the normative model in the range of finger
I .
joint motion were compared with those obtained by I. . ?
Landsmeer’s models. Four parameters. namely x and y . .+
. EXCUMION~
coordinates of the distal and proximal points, were .
.

used in the normative model to specify tendon lo- ] *

.

cations. With various combinations and selection of i *

.30 _
.
these four parameters, the results obtained could be
1
.
approximately either similar to or different from those .
..

? .
of Landsmeer’s models. Figure 11 illustrates the !!!?E4_!m?
comparison of tendon excursions and moment arms of ? . I(oI*IITIVE
YOOEL
I LANOSlAEEl'S L
the extensor slip at the PIP joint obtained from the

I
MODEL Y
8
?
normative model with those of Landsmeer’s Model 1.
In the range of 60” flexion-30” extension, the tendon
excursions were almost the same. However, the mo-
ment arms were not quite the same. Those obtained
.20

I’
I ’
I
.

I I , , I ,
i]

from the normative model were comparably smaller -10. 10. M 50. m a&”
with the joint at more than 20’ flexion or extension. fLEXION Of PIP JOINT (DEGREE)

From these comparisons, it implies that the use of a Fig. 17. Comparison of tendon excursion and moment arms
bowstring model between two fixed points to represent of flexor profundus at PIP joint from the normative model
certain extensor tendons is not optimum. In the future with those of Landsmeer’s Model 11 (in the range of 90’
refinements, the proximal point of the bowstring Rexion-10” extension).

.I model for the extensors will be allowed to shift along

.

. . . . . . . . . . . . . . . .\.
. .**.
.
.
l

1 .1
the articular surface with the joint angles. The com-
parison of tendon excursions and moment arms of the
flexor profundus tendon at the PIP joint obtained
.’
1
i
from the normative model with those of Landsmeer’s
Model III is shown in Fig. 12. The tendon excursions
obtained from both models are similar, in the range of
90” flexion-1O” hyperextension. The moment arms
obtained from the normative model were smaller as
the joint went into extension or slight flexion pos-
itions; however, they were larger at larger flexion
angles. On the whole, the tendon excursions obtained
by the normative model were similar to those obtained
from Landsmeer’s model. However, moment arms
obtained by the normative model were frequently
smaller than those from Landsmeer’s model. Con-
. sequently, the joint and muscle forces estimated, based
.
. on the normative model, will usually be larger than
. !E?QO!__~S
. those from Landsmeer’s model.
. AOWATIVE
NOM1
. . . LAoA~LN~EU’S Finally, one interesting point should be mentioned
.
here. At the proximal interphalangeal joint, the con-
joined lateral band shifted at the dorsolateral sides of
the base of the middle phalanx (Harris, 1972; Smith,
I I I I 1974). When both interphalangeal joints flexed, the
-40. -20. 20.
conjoined lateral band shifted several millimeters
EXTENStON Of PIP J&4, ( DEGREE)
volarly. When the proximal interphalangeal joint and
Fig. 11. Comparison of tendon excursions and moment arms
the distal interphalangeal joint extend, the lateral
of extensor slip at PIP joint obtained from the normative
model with those of Landsmeer’s Model I (in the range of60” bands shift dorsally. This volar-dorsal tendon shift at
flexion-30” extension). the base of the middle phalanx, as shown in Fig. 13.
788 K. N. AN. E. Y. CHAO, W. P. COOWY, III and R. L. LINSCHEID
Fig. 13. Volar-dorsal shift of conjoined lateral band at PIP
joint was illustrated by two-point technique of normative
model.
was perfectly simulated by the two-point technique of
the normative model.
In the use ofnoninvasive methods to identify tendon
locations in the intact hand, all soft tissue falls within
the limited range of resolution allowable by the
equipment used. Based on the results obtained in
noninvasive scanning, the following conclusions were
made :
(a) The high-performance ultrasonic camera de-
veloped by Stanford Research Institute does not have
the resolution required to identify tendons accurately
in the hand.
(b) The identification of the wire markers in the
hand sp&imens through tomograms and xeroradiog-
raphy were quite clear in both the anteroposterior and
lateral views. However, the tendon itself could not be
differentiated from other soft tissue without the aid of
metal wire.
(cl The use of the EM1 CT-5000 Body Scanner, as
shown in Fig. 3, provides certain promise.
A possible modification in the use of computerized
tomography is to embed the hand in a liquid-filled
container to further improve the image contrast. With
improved computer programming technique, the over-
ail resolution may be improved. Substantial modifi-
cation in both the equipment and the software is
required, however, before it can be used to identify
fibrous tissues in detail. No acceptable result has been
obtained to date.
SUMMARY
pi three-dimensional mathematical workable model
of human hand has been derived, based on the
quantitative anatomical analysis of 10 fresh cadaver
specimens. Two parameters, ‘force potential’ and ‘mo-
ment potential’, were used to analyze the constraint of
each tendon to the associated joint. Two points for
each tendon were then obtained to represent the
tendon as it passes through the joint. This model can
be used for both motion and force analysis of the
normal hand. Systematic variations of these data
based on clinical examinations can provide a simu-
lated study of hand in pathological conditions.
Acknowledgement - This study was funded by NIH Grant
AM 17172. Earlier contributions by Mr. F. E. Axmear and
Dr. J. D. Opgrande should be acknowledged.
REFERENCES
Bunnell, S. (1948) Surgery of the Hand, 2nd Edn. Lippincott,
Philadelphia.
Chao, E. Y., Opgrande, J. D. and Axmear, F. E. ( 1976)Three-
dimensional force analysis of finger joints in selected
isometric hand functions. J. Bfomechonics 9. 387-396.
Eyler, D. L. and Markee, J. E. (1954) The anatomy and
function of the intrinsic musculature of the fingers. J. Bone
Jr Surg. (A) 36, 1-9.
Flatt, A. E. (1971) The pathomechanics of ulnar drift. Social
and Rehabilitation Services, Final Report. Grant No. RD
2226M.
Haines, R. W. (1951) The extensor apparatus of the finger. J.
Anat. 85, 251.
Harris, C. and Rutledge, A. L. (1972) The functional anatomy
of the extensor mechanism offinger. J. Bone Jt Surg. (A) 54,
713-726.
Landsmeer. J. M. F. (1949) The anatomy of the dorsal
aponeurosis of the human finger and its functional signific-
ance. Anat. Rec. 104, 31.
Landsmeer, J. M. F. (1955) Anatomical and functional
investigations on the articulation of human fingers. Acta
Anat., Suppl. 24.
Landsmeer, J. M. F. (1961) Study in the anatomy of
articulation 1. The equilibrium of the ‘intercalated’ bone.
Acta morph. neerl.-stand. 3, X7-303.
Smith, E. M., Juvinall, R. C., Bender, L. F. and Pearson, J. R.
(1964) Role of the finger flexors in rheumatoid deformities
of the metacarpophalangeal joints. Arrhritis Rheum. 7,
467-480.
Smith, R. J. (1974) Balance and kinetics of the finger under
normal and pathological conditions. Clin. Orthop. Rel. Res.
104,92-111.
Stack, H. G. (1962) Muscle function in the finger. J. Bone Jr
Surg. (B) 44, 899.