The in vivo three-dimensional kinematics of the midcarpal joint in the
right wrists of twenty-four healthy volunteers was studied with a markerless
bone-registration technique. We separately investigated the kinematics during
the dart-throwing motion in twelve volunteers and the kinematics during the
flexion-extension motion in the other twelve. The average age was twenty-six
years (range, twenty to thirty-two years) for the volunteers included in the
analysis of the dart-throwing motion and twenty-five years (range, twenty to
thirty-two years) for those included in the analysis of the flexion-extension
motion. Nine men and three women participated in the study of the
dart-throwing motion, and eight men and four women participated in the study
of the flexion-extension motion. All volunteers consented to be included in
the study.
Magnetic resonance images were acquired with the same method as was used in
our previous
studies7,8.
To immobilize the elbow and the wrist at a specific angle during the
dart-throwing and flexion-extension motions, a special posture device with a
grip bar and goniometers was
used9. This device
has three motion axes, which all cross perpendicularly at the wrist joint and
enable the wrist to move along any specific plane.
For five of the twelve wrists in the group performing the dart-throwing
motion, magnetic resonance images were acquired in six positions from 60°
of radial deviation/extension to 40° of ulnar deviation/flexion in 20°
increments in the dart-throwing plane. For five of the twelve wrists in the
group performing the flexion-extension motion, magnetic resonance images were
acquired in seven positions from 60° of extension to 60° of flexion in
20° increments in the sagittal plane. For the other seven patients in each
group, magnetic resonance images were acquired in three positions (a neutral
position and two extreme positions). We set the dart-throwing motion as an
oblique wrist motion relative to the sagittal plane of the wrist in an attempt
to reproduce the motion of the throwing of a dart or the hammering of a
nail.
The contours of each bone were mapped from magnetic resonance volume
images, and three-dimensional surface models of the bones were constructed
with use of a software program (Virtual Place-M; Aze, Tokyo,
Japan)7. The
kinematic variables were calculated by registering the bone in one position
and comparing it with another. The volume registration technique employed in
this study was performed with use of the same software program.
The radius and all of the carpal bones except the pisiform were registered,
and the relative motion of each bone was calculated. Three-dimensional
animations of the relative motions were created by gradually rotating bones
around their own axis of rotation between two adjacent positions and
connecting them with use of originally developed software (Orthopedics Viewer;
Osaka University, Osaka, Japan). We created animations of all twenty-four
wrists and compared the animations of the ten wrists subjected to incremental
motion analysis with the animations of the fourteen wrists that were studied
in only three positions. We observed a marked similarity between them; thus,
we calculated the kinematic variables from the combined data of all
wrists.
The coordinate system was constructed with use of osseous landmarks with
the wrist in neutral position. The z axis (the supination-pronation axis) was
defined as the long axis of the distal part of the radius. The x axis (the
flexion-extension axis) was defined as a line passing through the center of
the capitate body at a right angle to the z axis and parallel to a line
connecting the radial and ulnar styloids. The y axis (radioulnar deviation
axis) was defined as a line perpendicular to the other two axes. In our
definition of the anatomical planes, we considered the sagittal plane to equal
the flexion-extension plane, the coronal plane to equal the radioulnar
deviation plane, and the axial plane to equal the supination-pronation
plane.
Direction of Global Wrist Motion
The direction of global wrist motion was defined as a line connecting two
centroids of the capitate at the two extreme wrist positions on the basis of
the assumption that there is little motion between the third metacarpal and
the capitate10. The
angle between the direction of global wrist motion and the x axis of the
coordinate system was calculated as viewed in the axial plane.
Midcarpal Motion
Midcarpal motion was investigated by assessing capitate motion relative to
the scaphoid, lunate, and triquetrum as well as by analyzing scaphoid, lunate,
and triquetrum motions relative to the capitate. If the capitate is used as a
reference to determine the kinematics of the proximal row, one can observe all
of the relative motions in the midcarpal joint at the same time because there
is very little motion between the four bones of the distal
row2-4.
The axes of
rotations11 between
the two extreme positions of the capitate relative to the scaphoid, lunate,
and triquetrum were the only ones calculated. The range of motion of the
capitate was calculated as an angle of rotation around each axis of
rotation.
Motion Between the Scaphoid and Lunate and Between the Lunate and
Triquetrum
The ranges of motion of the scaphoid and triquetrum relative to the lunate
were also calculated as an angle of rotation around each axis of rotation
relative to the lunate. The angle was defined as positive when the scaphoid or
triquetrum rotated palmarly relative to the lunate in the wrist motion from
radial deviation/extension to ulnar deviation/flexion or from extension to
flexion.
Relationship Between the Midcarpal Kinematics and the Radiocarpal
Kinematics
To compare the midcarpal kinematics with the radiocarpal kinematics, the
axes of rotations of the scaphoid and lunate relative to the radius were also
investigated. Relationships between the axes of rotations of the midcarpal and
radiocarpal joints were analyzed by comparing the location of the axes of
rotations of the radiocarpal joint with the primary axis of rotation of the
midcarpal joint at the neutral position of the wrist.
Statistical Analysis
All data were expressed as the mean and standard deviation. Statistical
analysis of differences was performed with use of the Student t test, with p
< 0.05 considered to be significant.
Direction of Global Wrist Motion
The angles between the direction of the global wrist motion and the wrist
flexion-extension axis in the axial plane were 59° ± 9° in the
dart-throwing motion and 91° ± 8° in the flexion-extension
motion. Of the twelve wrists studied in flexion-extension motion, four showed
an angle of <90° (mean, 82° ± 4°; range, 78° to
87°), which means that the plane of global wrist motion inclined slightly
toward a dart-throwing-motion plane (the dart-throwing-motion-inclined
flexion-extension-motion group). The other eight wrists showed an angle of
>90° (mean, 96° ± 4°; range, 91° to 103°), which
means that the plane of global wrist motion inclined slightly toward an
opposite-dart-throwing-motion plane (the
opposite-dart-throwing-motion-inclined flexion-extension-motion group).
Midcarpal Motion
Motion of the Distal Row Relative to the Scaphoid
The scaphoid-based animation showed the direction of capitate motion
relative to the scaphoid to be always similar: it was oblique and it extended
from radiodorsal to ulnopalmar in radioulnar deviation, in the dart-throwing
motion, and in the dart-throwing-motion-inclined flexion-extension motion.
From wrist radial deviation, radial deviation/extension, or extension to wrist
ulnar deviation, ulnar deviation/flexion, or flexion, respectively, the
capitate always moved from radiodorsal to ulnopalmar
(Fig. 1, A, B, and
C; see Appendix for videos).
The axes of rotation between the scaphoid and capitate in radioulnar
deviation, in the dart-throwing motion, and in the
dart-throwing-motion-inclined flexion-extension motion were located closely in
space and obliquely penetrated the neck of the capitate from a radiopalmar to
an ulnodorsal direction. In the axial plane, the axes of rotation of the
capitate relative to the scaphoid in radioulnar deviation, in the
dart-throwing motion, and in the dart-throwing-motion-inclined
flexion-extension motion formed a radially and palmarly opening angle of
43° ± 7° (as previously
reported7), 38°
± 10°, and 31° ± 9°, respectively, with the wrist
flexion-extension axis (Tables
I and
II). There were no significant
differences between these angles.
In the opposite-dart-throwing-motion-inclined flexion-extension motion, the
capitate seldom moved and the scaphotrapeziotrapezoid joint appeared to be
"locked" (Fig. 1,
D; see Appendix for videos). The axes of rotation between
the scaphoid and capitate in the opposite-dart-throwing-motion-inclined
flexion-extension motion varied considerably (mean angle with the wrist
flexion-extension axis, 5° ± 53°) so that a consistent pattern
of the axis of rotation was not detectable. In some wrists (Cases 20 and 24),
the intercarpal joints between the trapezium, trapezoid, and capitate showed
irregular motions; the trapezium and trapezoid translated distalward relative
to the trapezoid and capitate, respectively, when the wrist moved from
extension to flexion.
As the global wrist motion changed from radioulnar deviation to the
opposite-dart-throwing-motion-inclined flexion-extension motion, the range of
the capitate rotation relative to the scaphoid gradually decreased. The ranges
of rotation of the capitates around their own axes of rotation were 41°
± 10° in radioulnar deviation (as previously
reported7), 37°
± 10° in the dart-throwing motion, 27° ± 5° in the
dart-throwing-motion-inclined flexion-extension motion, and 9° ±
15° in the opposite-dart-throwing-motion-inclined flexion-extension motion
(Tables I and
II). There were significant
differences between the values in radioulnar deviation and the
dart-throwing-motion-inclined flexion-extension motion (p < 0.05), between
those in the dart-throwing motion and the dart-throwing-motion-inclined
flexion-extension motion (p < 0.05), and between those in the
dart-throwing-motion-inclined flexion-extension motion and the
opposite-dart-throwing-motion-inclined flexion-extension motion (p <
0.05).
Motion of the Distal Row Relative to the Lunate and Triquetrum
When the global wrist motion changed from radioulnar deviation to the
opposite-dart-throwing-motion flexion-extension motion, the directions of the
capitate motions relative to the lunate and triquetrum inclined in a similar
way, which was clearly different from the capitate motion relative to the
scaphoid. The capitate motions relative to the lunate and triquetrum in the
midcarpal joint were essentially similar and synchronous with each other
regardless of the type of wrist motion.
The axes of rotation of the capitate relative to the lunate and triquetrum
in the dart-throwing motion and flexion-extension motion ran more transversely
in the axial plane than did those during radioulnar deviation
(Fig. 2). The axes of rotation
of the capitate relative to the lunate and triquetrum formed, with the wrist
flexion-extension axis, a radially and palmarly opening angle of 41°
± 11° and 42° ± 14°, respectively, in radioulnar
deviation (previously
reported7), 23°
± 10° and 20° ± 11° in the dart-throwing motion,
13° ± 7° and 14° ± 12° in the
dart-throwing-motion-inclined flexion-extension motion, and -18° ±
18° and -16° ± 12° in the
opposite-dart-throwing-motion-inclined flexion-extension motion (Tables
I and
II). There were significant
differences in the values between radioulnar deviation and the dart-throwing
motion (p < 0.0005), between radioulnar deviation and the
dart-throwing-motion-inclined flexion-extension motion (p < 0.0005), and
between the dart-throwing-motion-inclined flexion-extension motion and the
opposite-dart-throwing-motion-inclined flexion-extension motion (p <
0.005).
The ranges of motion of the capitate relative to the lunate and triquetrum
around their own axes of rotations were 44° ± 10° and 33°
± 6°, respectively, in radioulnar deviation (as previously
reported7), 54°
± 14° and 44° ± 12° in the dart-throwing motion,
48° ± 7° and 33° ± 8° in the
dart-throwing-motion-inclined flexion-extension motion, and 41° ±
12° and 38° ± 9° in the
opposite-dart-throwing-motion-inclined flexion-extension motion (Tables
I and
II). The values did not differ
significantly among the different types of wrist motion.
Global Capitate-Based Midcarpal Motion
Regardless of the type of wrist motion, the animation of the capitate-based
midcarpal motion showed that the loci of the displacement of all of the joint
surfaces of the midcarpal joint were located within a midcarpal ovoid space
whose major axis coincided with the typical axis of rotation of the scaphoid,
running from a radiopalmar to an ulnodorsal direction (Figs.
3 and
4; see Appendix for videos). A
line connecting the centers of the joint surfaces of the midcarpal joint could
be schematized as a letter "C" entwining the midcarpal ovoid
(Figs. 3, B, and
5, A).
We also found a specific pattern of midcarpal motion for each wrist motion
(Fig. 4). We previously showed
that the directions and ranges of motion of the lunate and triquetrum were not
significantly different from those of the scaphoid in radioulnar
deviation7. In the
dart-throwing motion, the directions of the lunate and triquetrum slightly
inclined toward a wrist flexion-extension-motion plane and the ranges of
motion did not change significantly compared with those in radioulnar
deviation, while the direction of the scaphoid motion was not significantly
different from that in radioulnar deviation but the range of scaphoid motion
decreased. When the wrist was moved in a dart-throwing-motion-inclined
flexion-extension-motion plane, these tendencies became more pronounced: the
directions of the lunate and triquetrum inclined more, while the range of
scaphoid motion decreased more. Finally, when the wrist was moved in an
opposite-dart-throwing-motion-inclined flexion-extension-motion plane, the
scaphoid seldom moved but the lunate and triquetrum moved along the plane of
the scapholunate joint, with the range of motion not significantly changing.
The lunate and triquetrum always moved separately, but the directions of the
motions of the lunocapitate and triquetrohamate joints were not significantly
different from each other.
In summary, as the global wrist motion changed from radioulnar deviation to
the opposite-dart-throwing-motion flexion-extension motion, the range of
motion of the scaphoid in the midcarpal joint gradually decreased without the
direction changing significantly and the ranges of motion of the lunate and
triquetrum did not significantly change but the direction gradually inclined
from a dart-throwing-motion plane to an opposite-dart-throwing-motion-inclined
flexion-extension-motion plane.
Motion Between the Scaphoid and Lunate and Between the Lunate and
Triquetrum
The scaphoid-based animation showed that as the wrist moved from radial
deviation, radial deviation/extension, or extension to ulnar deviation, ulnar
deviation/flexion, or flexion, respectively, the lunate always rotated
dorsally relative to the scaphoid; the relative position between the scaphoid
and lunate was similar to the dorsiflexed intercalated segment instability
posture (Fig. 1; see Appendix
for videos).
As the global wrist motion changed from radioulnar deviation to the
opposite-dart-throwing-motion-inclined flexion-extension motion, the range of
rotation of the lunate relative to the scaphoid gradually increased, averaging
8° ± 6° in radioulnar deviation (as previously
reported7), 21°
± 11° in the dart-throwing motion, 25° ± 4° in the
dart-throwing-motion-inclined flexion-extension motion, and 29° ±
9° in the opposite-dart-throwing-motion-inclined flexion-extension motion
(Tables I and
II). There was a significant
difference in the values between radioulnar deviation and the dart-throwing
motion (p < 0.001), radioulnar deviation and the
dart-throwing-motion-inclined flexion-extension motion (p < 0.0001), and
radioulnar deviation and the opposite-dart-throwing-motion-inclined
flexion-extension motion (p < 0.0001).
The lunotriquetral joint also moved in all types of wrist motion; however,
we did not find a significant difference in the range of motion of the
triquetrum relative to the lunate between radioulnar deviation, dart-throwing
motion, and flexion-extension motion (Tables
I and
II).
Relationship Between Midcarpal and Radiocarpal Kinematics
The axes of rotation of the scaphoid and lunate relative to the radius were
almost parallel to each other and passed through the neck of the capitate in
all types of wrist motion. The relationship between the major axis of the
midcarpal ovoid and the wrist flexion-extension axis was opposite the
relationship between the axes of the radiocarpal joint and the wrist
flexion-extension axis in radioulnar deviation, the relationships were similar
in the dart-throwing motion, and the relationships were midway between those
in radioulnar deviation and the dart-throwing motion in the flexion-extension
motion.
In the axial plane, the axes of rotation of the scaphoid and lunate
relative to the radius formed radially and palmarly opening angles with the
wrist flexion-extension axis of -43° ± 20° and -39°
± 14°, respectively, in radioulnar deviation (as previously
reported7); 23°
± 18° and 44° ± 27° in the dart-throwing motion;
-1° ± 6° and 1° ± 3° in the
dart-throwing-motion-inclined flexion-extension motion; and -2° ±
5° and -2° ± 7° in the
opposite-dart-throwing-motion-inclined flexion-extension motion (Tables
I and
II). There were significant
differences in these values between radioulnar deviation and the dart-throwing
motion (p < 0.0001) and between the dart-throwing motion and the
dart-throwing-motion-inclined flexion-extension motion (p < 0.05).
We previously investigated the scaphotrapeziotrapezoid joint in cadavers
both anatomically12
and
kinematically13,
and those studies suggested that the scaphotrapeziotrapezoid joint has
essentially a single degree of freedom, or is a uniaxial joint. The results of
this current in vivo kinematic study further revealed that the range of motion
of the scaphoid in the midcarpal joint gradually decreases as the global wrist
motion changes from radioulnar deviation to flexion-extension motion. We
noticed that the convex joint surfaces of the scaphoid articulating with the
trapezium and the trapezoid can approximate an ovoid whose major axis runs
from a radiopalmar to an ulnodorsal direction and coincides with the axis of
rotation of the capitate relative to the scaphoid. Moreover, the concave joint
surface of the scaphoid articulating with the capitate can also be a part of a
smaller ovoid whose axis is the same as the major axis of the midcarpal ovoid
(Fig. 3, B). Actually,
the axis of rotation of the joint between the scaphoid and the distal carpal
row is not rigid, probably because of minor mobility occurring among the
trapezium, trapezoid, and
capitate14 and a
lack of ligamentous constraint between the proximal part of the scaphoid and
the capitate. Thus, this joint is essentially uniaxial, but it also has an
adaptive mechanism that allows preservation of articular congruity of the
midcarpal joint when distortional force is applied.
We have also found that the skeletal constraint of the triquetrohamate
joint is relatively weak because the joint is essentially an ellipsoidal joint
with two degrees of freedom, or is a biaxial
joint8. The current
study further revealed that the motion of the lunate was always similar to
that of the triquetrum. Our three-dimensional analysis showed that most of the
joint surfaces of the lunocapitate and triquetrohamate joints are also part of
the midcarpal ovoid whose major axis runs in a radiopalmar to an ulnodorsal
direction (Fig. 3, B).
Thus, most of the joint surfaces in the midcarpal joint are contained within a
midcarpal ovoid; the carpal bones might be moving within this volume, but they
still have distinct motions relative to each other within it.
We consider midcarpal motion to be essentially the combined motion of three
types of joint systems: (1) the uniaxial joint between the scaphoid and the
distal row, the axis of which runs in a radiopalmar to an ulnodorsal
direction; (2) the biaxial and ellipsoidal joint between the lunate and
triquetrum and the distal row; and (3) the intercarpal joints of the proximal
row, which have an adaptive mechanism that accommodates the above-mentioned
two types of joint systems in the midcarpal joint. We believe that the primary
motion plane in the midcarpal joint is a dart-throwing plane, which is defined
by anatomical constraints of the joint between the scaphoid and the distal
row; however, adaptive intercarpal motions, mainly in the scapholunate joint,
and relatively weak constraints of the lunocapitate and triquetrohamate joints
allow the global midcarpal joint to move in the flexion-extension or even the
opposite-dart-throwing-motion-inclined flexion-extension-motion planes as
well.
This study raises concerns about a self-stabilizing mechanism of the carpus
under load. When the trapezium is axially loaded against the scaphoid, the
scaphoid tends to rotate into flexion; this flexion moment is constrained by
the extension moment experienced by the triquetrum, and stable equilibrium is
achieved15,16.
As a modification of this concept, we advocate use of an "ovoid/C"
concept to explain the carpal self-stabilizing mechanism. The
three-dimensional configuration of a line connecting the centers of the joint
surfaces of the midcarpal joint can be schematized as a letter "C"
entwining a midcarpal ovoid (Figs. 3,
B, and 5,
A). On the radiograph of a semisupinated wrist
(Fig. 5, B), which is
almost compatible with an axial radiograph of the ovoid, the midcarpal joint
displays a C-shaped
outline7. The
ulnopalmar rotational moment of the scaphoid generated by the inclination of
both the scaphotrapeziotrapezoid and the scaphocapitate joint competes and
maintains balance with the radiodorsal rotational moment of the triquetrum
generated by the inclination of the triquetrohamate joint
(Fig. 5, B). Gilula
and Weeks reported that three fairly smooth radiographic arcs can be drawn to
define the normal carpal relationship, and disruption of any one of these
lines may indicate a major carpal
derangement17. We
believe that our C-shaped outline on the radiograph of a semisupinated wrist
could provide additional information about carpal derangement because this
view is essentially a lateral view of the midcarpal joint.
In terms of the kinematic relationship between the midcarpal joint and the
radiocarpal joint, the question arises of how the oblique midcarpal motion
accommodates itself to radiocarpal motion. It is well known that, during
radioulnar deviation, the three proximal carpal bones move synergistically
relative to the radius from a flexed and radially deviated position in wrist
radial deviation to an extended and ulnarly deviated position in wrist ulnar
deviation5,18-20.
This finding is consistent with our observation that the axes of rotation of
the proximal row relative to the radius run in a radiodorsal to an ulnopalmar
direction. The relationship between the primary axes of the midcarpal joint
and the radiocarpal joint during radioulnar deviation is reciprocal,
effectively canceling rotation in the flexion-extension plane of the hand on
the forearm. During the dart-throwing motion, the axes of the two joints are
similar and synergistic with each other, which may explain why the range of
wrist motion achieved in this plane is greater than that achieved in
radioulnar deviation and why the dart-throwing plane is the usual plane of
utilization of the
wrist21. During
flexion-extension motion, the axis of the radiocarpal joint is almost the same
as the wrist flexion-extension axis and the axis of the midcarpal joint is
oblique, findings that are consistent with the fact that, normally, extension
of the wrist is associated with radial deviation and flexion is associated
with ulnar
deviation18.
Our current kinematic technique has some limitations, the greatest of which
is that the study was based on static three-dimensional views of the carpus in
the limited number of wrist positions examined. The static measurement does
not include any inertial or functional effects that might occur during normal
wrist motion. The angles of the axis of rotation would not be fixed angles for
the entire range of motion; they would have varied if we had used more
increments of motion. It would be reasonable to perform dynamic studies to
supplement this study. We tried to establish the wrist dart-throwing motion as
an oblique wrist motion relative to the sagittal plane, but there is no
consensus about what constitutes the true plane of the wrist dart-throwing
motion22. However,
the technique that we used in this study allowed us to obtain new information
on in vivo three-dimensional midcarpal kinematics without subjecting the
volunteers to radioactive exposure. Our findings suggest that the proximal
carpal row with all its related joints has a unique mechanism that contributes
to both the stability and the mobility of the wrist. Hopefully, this new
information on the ovoid/C-shape perspective of the anatomy and the kinematics
of the midcarpal joint will assist clinicians in obtaining a better
understanding of the wrist joint and some of its disorders.