advances in the in vivo measurement of carpal kinematics

ADVANCED TECHNIQUES IN THE MANAGEMENT OF WRIST TRAUMA 0030-5898/01$15.00 + .OO

ADVANCES IN THE IN VIVO MEASUREMENT OF NORMAL AND ABNORMAL CARPAL KINEMATICS

Joseph J. Crisco, PhD, Scott W. Wolfe, MD, Corey P. Neu, MSc, and Sandi Pike, MSc

PURPOSE

The purpose of this article is to present the development of an in vivo, three-dimensional methodology using markerless bone registration (MBR) for examining the normal and abnormal kinematics of the wrist carpal bones.12 The resulting descriptions of three- dimensional kinematics from normal patients and patients with documented unilateral scapholunate interosseous ligament (SLIL) injuries are briefly presented.

CARPAL BONE MOTION PAlTERNS: ROW VERSUS COLUMN

Wrist motion depends in large part on the complex motions of the constituent carpal bones. To facilitate understanding of the normal motion of the wrist, the carpal bones traditionally have been described by grouping bones with similar motions into various patterns, including rows, columns, and more complicated combinations. De Lange et all4 agreed with the concept of dividing the carpal bones into proximal and distal rows. In this

This work was funded by grant no. NIH ARM005 and the RIH Orthopaedic Foundation.

"fixed-row" concept, the proximal and distal rows were each thought to rotate as a rigid group around fixed axes, with the scaphoid bridging the two rows.14 Berger et a13 investigated individual carpal bone motions and also supported the fixed-row concept. Ruby et a146 concluded that the wrist functions as two carpal rows, with the distal row of bones rel- atively tightly bound to one another and the proximal row bones less so but still moving together. Gilford et all7 and Talei~nik~~ proposed that the wrist joint performs as a system of three longitudinally linked columns. N a ~ a r r o ~ ~ originally proposed the column theory in 1919 to explain certain types of carpal instability. Talei~nik~~ modified this theory by including the trapezium, trapezoid, and hamate with the central column of the wrist. Lichtman et aP8,30 returned to a modification of the original row theory by noting the critical ligamentous con- nections at either end of the proximal carpal row and the fact that instability can arise because of a break at any point within this ring of carpal bones.

The results of more recent investigations report that the existing row and column theories may oversimplify carpal motion.', l o r z 5 Craigen and Stanley' reported that carpal kinematics

From the Department of Orthopaedics, Brown University School of Medicine, Rhode Island Hospital (JJC); Division of Engineering, Brown University (JJC, CPN, SP), Providence, m o d e Island; The Hospital for Special Surgery and Weill Medical College of Cornell University, New York, New York (SWW)

ORTHOPEDIC CLINICS OF NORTH AMERICA

VOLUME 30 NUMBER 2 - APRIL 2001 219

220 CRISCO et a1

cover a spectrum from the row theory to the column theory and that women are more likely to have a column type wrist. They proposed a "CR index" so that the tendency of a wrist to- ward row or column theory can be quantified. Recent work by Crisco and Neu'O showed that in vivo carpal bone rotation patterns do not follow traditional functional theories of rows and columns in wrist flexion, extension, and ulnar deviation. Rather, in vivo carpal bones exhibit more complex patterns that depend on the direction of wrist motion. Thus, discrepancies in the description of carpal bone motion patterns motivate the advancement of new technologies to appropriately describe wrist kinematics.

CARPAL BONE MOTION PATTERNS: EFFECT OF SCAPHOLUNATE INTEROSSEOUS LIGAMENT TEARS

Tear of the SLIL is a common ligamentous injury in the carpus. There is controversy among investigators concerning whether and how a SLIL tear affects carpal motion, including several reports of no effect at a11.2,21,35,55 Others have assigned a major role to this structure in the prevention of SL disassociation and carpal collapse.* Among researchers who agree that the SLIL has a key stabilizing role, there is still disagreement as to the specific changes that take place following a tear of this ligament. Thus, if consistent changes do result from SLIL injury, further research is needed to move to- ward agreement on the details.

Several researchers believe that SLIL injury does not affect carpal stability or kinematics. Berger et a12 measured little change in SL kinematics with isolated sectioning of the SLIL in cadaveric specimens, concluding that the pri- mary scapholunate (SL) joint stabilizer is the ra- dioscapholunate ligament (RSL), not the SLIL.2

Other researchers agree that SLIL tears affect carpal kinematics. Using sonic digitization to detect alterations in carpal bone position, Ruby et a145 concluded that sectioning the SLIL significantly changes carpal motion of the scaphoid and lunate. Their study also supported the idea that the dorsal portion of the SLIL is an im- portant structure in maintaining a normal SL relationship under these conditions. This find-

*References 1, 4, 16, 20, 21, 24, 26, 30, 35, 36, 40, 44, 45, 51,53,56, and 57.

ing is in agreement with Ka~er ,2~ Palmer et a1,4O and Howard et a1.20

The repair of an isolated dorsal tear or a complete disruption is crucial to the normal function of the SL joint. Left untreated, the resultant alteration in carpal kinematics may lead to the characteristic and progressive arthritic pattern referred to as SL advanced collapse (SLAC).18,29 Surgical methods of treating SL ligament tears have included tendon weaves, repair of the na- tive ligament with or without dorsal capsu- lodesis, limited wrist arthrodesis, and proximal row carpectomy.52

Anatomic and kinematic studies have shown that the SL joint has a dorsal axis of r o t a t i ~ n . ~ ~ , ~ This is particularly pertinent because hyperextension and intercarpal supination initiate most carpal injuries.32 Mechanically, the SL joint has a specialized design that enhances stability but also is believed to allow load- dampening during excessive loads applied to the wrist in hyperextension and intercarpal supination. The controversy over SLIL injury- related findings further motivates the advancement of new technologies to appropriately describe wrist kinematics.

SUMMARY OF PREVIOUS CARPAL KINEMATIC MEASUREMENT METHODOLOGIES

Carpal kinematics have been investigated using several methodologies, including

phy60; stereoph~togrammetry~~, 21, 42, 49 , * gonio-

tracking; and magneti~, '~ ~patial,2~, 31 and sonic digitization.2, 45

Radiography has been used for more than 100 years to describe the motion of the carpal bones.5 More recent cadaveric studies typically used implanted radiopaque markers or transcutaneous pins with biplanar radiography to track the individual carpal bones. Using a minimally invasive stereophotogrammetric (biplanar radiographic) technique, Kobayashi et a126 studied normal carpal kinematics. In these studies, markers were implanted completely within the bones, so their direct in- fluence on surrounding anatomical structures was Biplanar radiography has also been used by Ishikawa et a123 and Craigen and Stanley9 who analyzed radiographs to

8,9,13,26,35,39,45,48,54; rentgenogra-

2553 and video50, 61 metry6, 40,47,51,56;

ADVANCES IN THE IN VIVO MEASUREMENT OF CARPAL KINEMATICS 221

compare scaphoid shortening and ulnar translation of the scaphoid during different wrist motions. However, many of these radiographic techniques required disruption of the soft tissue envelope for marker placement. To what extent the use of implantable markers and other invasive techniques alter normal carpal kinematics is unclear.

NEED FOR MEASURING IN VlVO CARPAL KINEMATICS

In examining the results and methodologies of previous carpal kinematics studies, it is im- portant to consider the limitations of the methods used by many of these studies. First, data from in vitro studies using cadaveric specimens harvested postmortem may not accu- rately reflect normal in vivo results. Second, in vitro studies have largely used invasive techniques, such as the implantation of radiopaque markers or rods that may have disrupted soft tissues or interfered with normal kinematics. Third, the methodology required to overcome the obstacles inherent in the measurement of in vivo carpal kinematics (e.g., the small size and tight articular spacing of the bones), has only recently become available.12 Finally, wrist motion was simulated by the application of estimated muscular forces across the wrist. These studies did not account for the possibility that small changes in wrist loading may have profound effects on wrist kinematic^.^^ Many in vivo studies examined only gross wrist motion and did not specify motion of the individual carpal bones. In addition, some previous studies were based on two-dimensional kinematic analyses, and documenting three- dimensional kinematics is essential because of the complex and coupled motions of the carpus.

To overcome many of these previous limitations and examine the complex three- dimensional kinematics of the carpal bones, an in vivo methodology using MBR was developed." Because this technique is noninvasive, applicable in vivo, and capable of pro- viding unique data of bone geometric proper- ties, it may greatly extend our understanding of the carpus. This methodology has the poten- tial to help answer questions that previous invasive methods were unable to address, such as how carpal instability progresses or what

the long-term effects of surgical intervention may be.

AN OVERVIEW OF THE METHODS FOR MARKERLESS BONE REGISTRATION

Measuring the rigid body motion of a carpal bone requires tracking the bone in multiple positions. As described previously, carpal motion has typically been measured by tracking the displacements of a minimum of three specific markers (e.g., radiopaque spheres) delib- erately imbedded in or attached to the bones. In contrast, MBR uses the inherent features and shape of the bone surface as the markers.12 The bone surfaces are defined at each position using the same methodology and then are registered to determine motion. Importantly, if the shape of the carpal bones were ideal spheres, MBR would fail, because a sphere contains no dis- tinguishing features. Luckily, the carpal bones and surrounding anatomy are rich with shape features that permit the description of their detailed kinematics.

Wrists are first scanned with CT in a neutral position, defined by aligning the back of the hand with the back of the forearm. Addi- tional positions include 30" and 60" of flexion, 30" and 60" of extension, 20" and 40" of ulnar deviation, and 20" of radial deviation, as determined by protractor readings on a positioning jig. Continuous 1-mm image slices from the distal radius and ulna to the proximal heads of the metacarpals are acquired at each position (typically 60 image slices for each wrist position; voxel size = [0.2-0.912 x 1 mm3).

The contours of each bone are then segmented from the CT volume images using Analyze AVW software (Mayo Foundation, Rochester, MN), reconstructed using custom MATLAB software (Mathworks, Natick, MA), and used to calculate bone physical proper- ties and motions. Carpal volumes, centroid locations and principal axes of orientation are determined from bone contour data at each wrist position using previously developed equations."

The posture of the carpal bones in the neutral position is defined using the orientation of the first principal inertial axis of each bone (defined as the principal axis about which the minimum bone inertia occurs and shown as I

Figure 1. Orientation and position of the carpal bones were defined using their centroids and inertial axes. Those of the capitate (the first principal axis is labeled I) are shown here along with the anatomic coordinate system fixed to the radius.

for the capitate in Fig. 1) from the neutral bone contour data. In neutral, the orientation of the first principal axis (0 of the carpal bones is de- composed into two fixed angles relative to an anatomic coordinate system (described subse- quently) and is found to be in flexion or extension and radial deviation (described subse- quently in the section on results). The flexion or extension posture is defined as the initial fixed angle between the first principal axis and the coronal (XY) plane of the anatomic coordinate system (positive and negative for flexion and extension, respectively). Radial deviation posture is defined as the initial fixed angle between the XY projection of the first principle axis and the Z axis.

Motions of the carpal bones are determined in a sequence of steps (Fig. 2). First, motion of the carpal bones and radius for each wrist position are determined relative to a fixed CT- based coordinate system. Specifically, carpal bone rigid body motion transformations from the neutral wrist position to subsequent wrist positions are calculated by registering the respective principle axes of inertia and centroids

of the capitate. This technique is valid only for bones whose surfaces are consistently imaged in subsequent positions because principal axes orientations and centroid locations vary with the amount of the imaged bone surface. Be- cause the scanned length of the radius varies and is typically less than 2 cm, registration relative to a fixed CT-based coordinate system is calculated for both of these bones by minimiz- ing the root mean square distances between bone surfaces.43

Second, kinematic variables for each carpal bone are described relative to an orthogo- nal distal radius-based coordinate system defined by anatomic features to facilitate motion descriptions.

Finally, carpal bone motions from neutral to subsequent wrist positions are reported using helical axis of motion (HAM) variables, defined as a rotation about and a translation along a unique helical axis (e.g., Panjabi et al4I). In addition, carpal rotations are typically ex- pressed about X (supination-pronation), Y (flexion-extension), and Z (radioulnar deviation) anatomic axes as determined by multiply- ing the total rotation by the square of respective HAM axis orientation components.

Carpal motions have been determined for three groups of subjects: (1) normal or healthy subject wrists (n = 20), (2) injured subject wrists following an SLIL injury (n = 8); and (3) uninjured contralateral wrists (uninjured wrists of the injured subjects; n = 8). Importantly, the accuracy of MBR for measuring carpal kinematics has been fully evaluated using a cadaveric model,3* with rotation errors of less than 0.5" for the capitate and scaphoid and rotation errors of generally less than 2" for the other bones. Translation errors of the bones generally were less than 1 mm. Thus, markerless registration methods, such as that described herein, provide accurate measurements of carpal kinematics and can be used to study the noninvasive, three-dimensional in vivo kinematics of the wrist and other skeletal joints.

NORMAL CARPAL MOTION: ROW VERSUS COLUMN

In vivo MBR measurements limited to flexion-extension and ulnar deviation suggest that the row and column models are too sim- plistic. In flexion (Table 1 and Fig. 3A), the


Figure 2. Two registration steps are used to calculate carpal bone motion. The neutral radius (solid surface) is first defined as the reference bone (Al). Then the radius at any given position (point surface) is registered to the neutral radius, which accounts for all forearm motion (A2). This transformation is then applied to each carpal bone, in this case the scaphoid (A2, 61). The motion of the scaphoid, relative to the radius, is then defined by the transformation that registers the scaphoid from any given position to its neutral position (62).

Table 1.CARPAL BONE ROTATIONS AS A PERCENTAGE OF CAPITATE ROTATION (? VALUES) IN THREE DIRECTIONS OF WRIST MOTION. BONE ROTATIONS THAT WERE NOT SIGNIFICANTLY DIFFERENT ARE GROUPED. THIS GROUPING IS ALSO ILLUSTRATED IN FIGURE 3. VALUES ARE FOR BOTH WRISTS OF 5 MALES AND 5 FEMALES (n = 20)

Flexion Extension Ulnar Deviation

Carpal Bone Value Group Value Group Value Group

Hamate (H)

Capitate (C)

Trapezoid (Td)

Trapezium (Tp)

Triquetrum (Tq)

Lunate (L)

Scaphoid ( S )

Pisiform (P)

102% (0.83) 100% (1.0) 108% (0.85) 96% (0.94) 50% (0.54) 46% (0.63) 73%

(0.85) 54%

(0.68)

ff

ff

ff

ff

B

91% (0.87) 100% (1.0) 95% (0.86) 92% (0.89) 63% (0.69) 68 %

(0.61) 99%

(0.91) 75 %

(0.80)

ff

ff

ff

95 % (0.94) 100% (1.0) 90% (0.51) 140% (0.72) 24% (0.17) 16%

(0.22) 16%

(0.15) 20% (0.11)

~

ff

ff

ff

4J

224 CHSCO et a1

Figure 3. The pattern of carpal bone rotations varied with the direction of wrist motion (n = 20; both wrists of 5 females and 5 males) for flexion (A), extension (B), and ulnar deviation (C). Bones that rotated together are shaded similarly, and these bones rotated significantly different from differently shaded bones (see Table 1). These images are a palmar view of the right wrist in the neutral position reconstructed from a typical CT volume image.

bones of the distal row (hamate, trapezoid, and trapezium) did not rotate differently from the capitate (P = 0.4, 0.02, 0.07; Bonferroni P = 0.004 for significance). Similarly, in the proximal row the triquetrum did not rotate differently from the lunate ( P = 0.28), but scaphoid rotations were different from the rotations of the proximal and distal rows ( P < 0.0001). In extension (Fig. 3B), the scaphoid rotated with the distal row ( P = 0.7; Bonferroni P = 0.0018), whereas the other bones in the proximal row did not rotate differently from one another. In ulnar deviation (Fig. 3 0 , the trapezium rotated significantly more than did the distal row ( P < 0.0001; Bonferroni P = O.OOlS), whereas the scaphoid rotated with the proximal row. Pisiform rotations and translations were unique.

The authors’ study demonstrates that, in wrist flexion and extension and ulnar deviation, carpal bone rotation does not follow the traditional row and column- theories of wrist motion. In vivo, carpal bone rotation is complex, and it varies with the direction of wrist motion. Because rotation patterns are different for sagittal and coronal plane rotations, it remains to be seen what carpal rotation patterns occur during combined wrist motions. One of the aims of the authors’ current work involves determining carpal bone motion for the complete range of in vivo wrist motion.

The complexity of carpal motion also can be appreciated by studying the direction of rotation for each carpal bone. The coupled motions of the carpus, those motions that are inherently linked but occur in different directions, are per-

haps best visualized by the orientation of the respective HAM axes (Fig. 4).

NORMAL CARPAL MOTION: CAPITATE KINEMATICS

A more detailed look at individual carpal bone motion revealed that the capitate can be used as a measure of wrist motion because it articulates closely with the third metacarpal. The capitate and third metacarpal were determined to rotate together during wrist flexion- extension and radioulnar deviation. In wrist flexion-extension, the mean difference in the rotation of the capitate and third metacarpal was -0.7 f 4.3” and was not significantly different from zero ( P = 0.279). In wrist radial and ulnar deviation, the mean difference in the rotation of the capitate and third metacarpal was -0.3 f 4.4“ and was also not significantly different from zero (P = 0.684); however, gross protractor readings did not correlate well with capitate motions; this calls into question the utility and validity of clinical protractor or goniometric measurements for carpal posture and motion. The mean difference between protractor readings and capitate rotation measurements has been found to be 9.0” (P < 0.0001). This discrepancy in protractor readings and rotation measurements can be distributed over a range of wrist positions.

In vivo wrist motion does not simply occur about a single pivot point, as seen in a typical universal joint. This is demonstrated by non- intersecting rotation axes for different capitate motions (Fig. 5). The motion axis location


Figure 4. The complexity of carpal bone motion during ulnar deviation is shown here by the location and orientation of the carpal rotation axes (HAM axes). For example, the orientation of the HAM axes demonstrates that the rotation of the scaphoid and lunate are in flexion-extension while the capitate (and wrist) is in ulnar deviation. These HAM axes are the vector average of 20 measurements (both wrists of five females in two ulnar deviation positions). Each specific axis is labeled with the carpal bone abbreviation as in Table 1.

for all wrist motions indicates that the ulnar deviation motion axis is the most distal axis and the flexion and extension motion axes are the most proximal axes. The distance between flexion and ulnar deviation axes was 3.9 f 2.0 mm,

and the distance between extension and ulnar deviation was 3.9 f 1.4 mm.

Capitate rotation axes for males tended to be located more distally than axes for females; however, the authors believe that this result

Figure 5. Average (n = 20; both wrists of 5 females and 5 males) capitate rotation axes (HAM axes) for flexion (F), extension (E), radial (R), and ulnar (U) deviation shown from three views.

226 CRISCO et a1

was related to subject size and not to gen- der. Interestingly, the capitate bone volumes of the male subjects were significantly greater ( P < 0.0001) than those of the female subjects; one possible explanation for more distally located male axes. The mean capitate volumes for our male and female subjects were 3231 f 542 mm3 (range: 2467-3998 mm3) and 2220 f 205 mm3 (range: 1987-2503 mm3), respectively. Not surprisingly, scaphoid and lunate volumes tended to correlate with capitate volume (Fig. 6). Also, the distance of the capitate centroid to the origin of the anatomic coordinate system in the neutral position in- creased with the mean capitate volume, further

A

n

"E E W

W Y

suggesting a correlation between the location of distal axes and carpal bone size.

NORMAL CARPAL MOTION: SCAPHOID AND LUNATE KINEMATICS

Furth.er investigation of individual carpal bone motion indicated that scaphoid and lunate rotations were linearly related to capitate rotation (Fig. 7). Importantly, rotations of the scaphoid and lunate differed for flexion and for extension. In flexion, the scaphoid contributed 73% (r2 = 0.85) of wrist motion, whereas the lunate contributed 46% (1 = 0.63).

B

1500 2000 2500 3000 3500 4000

Capitate Volume (mm3)

Figure 6. Carpal bones were typically larger in males than in females. Increasing capitate volume tended to correlate with increasing scaphoid (A) and lunate (B) volumes.


H Scaphoid - ScaphoidFit - - Capitate

0 Lunate - Lunate Fit

-80 -60 -40 -20 0 20 40 60 80 extension flexion

Capitate Rotation (degrees)

Figure 7. Normal flexion-xtension rotation of the scaphoid and lunate plotted as a function of capitate rotation.

In extension, the scaphoid contributed 99% (1-2 = 0.91) of wrist motion, while the lunate contributed 68% (9 = 0.60). These findings indicate "engagement" of the scaphoid and capitate during wrist extension (Fig. 8). The linear relationship seems strong in the data but may be

an artifact of averaging few positions over several subjects. Only more exhaustive data col- lection will be able to fully determine whether relative carpal rotations remain linear through- out the range of wrist motion. There was minimal HAM translation in flexion, whereas in

Figure 8. In extension, the capitate seems to engage the scaphoid sulcus (arrows) and limit the relative motion between the scaphoid and capitate (A). The capitate disengaged from the scaphoid in flexion (6).

228 CRISCO et a1

extension all three carpal bones moved radially approximately 4 mm at 60" of wrist extension.

ABNORMAL CARPAL MOTION: SLIL INJURED VERSUS UNINJURED KINEMATICS

The aim of this study was to test the hypothesis that SLIL tears alter in vivo carpal bone kinematics. The three-dimensional kinematics of the carpal bones were measured using our noninvasive CT-based method. The kinematics of the injured (Injured), and the uninjured contralateral wrist (Uninjured), were compared with the existing database of healthy male and female volunteers (Normals) to test the authors' hypothesis.

Both wrists (Injured and Uninjured) of eight subjects with arthroscopically verified unilateral traumatic SLIL tears (7 males, 1 female; average age, 38 y [range, 20-541) were studied. The kinematic values were compared with the data from healthy wrists of 10 volunteers (n = 20) defined as Normal (5 male, 5 female; average age, 26 y [range, 21-471). IRB approval and informed consent was obtained for all subjects. The posture and kinematic analyses focused on the capitate, scaphoid, and lunate bones. Carpal bone posture was described using the first principal axis of inertia of each carpal bone relative to the distal radial coordinate system. Carpal bone kinematics were described using helical axis of motion (HAM) variables, consist- ing of a rotation about and translation along a unique axis in three-dimensional space, with respect to the distal radial coordinate system. Bone rotations were plotted as a' function of capitate rotation and fit with linear regression lines to quantify the contribution of carpal rotation to wrist flexion and extension. Capitate rotation was used as a measure of wrist rotation.

The significance of the differences, if any, between the Injured (n = 8>, Uninjured (n = 8), and Normal (n = 20) were examined for neutral posture and wrist flexion and extension. A one- way ANOVA and with Dunnett multiple com- parisonpost-hoc test ( P < 0.05) wasused to determine if SLIL tears significantly altered carpal bone posture in the neutral wrist position. Stu- dent's t-tests with a Bonferroni correction factor for multiple comparisons ( P < 0.017) were

used to determine whether the slopes of the linear regression lines describing carpal bone rotation differed.

The neutral posture of the lunate was significantly ( P < 0.01) more extended in the Injured wrists (56 f 13") than in the Normal wrists (83 + 8"). Surprisingly, but consistent with our other findings, the contralateral Uninjured lunate (53 f 7") was also significantly different from Normals, yet did not differ from the Injured. The Injured (34 + 16") and the Unin- jured (27 f 20") scaphoids were not different, but both were significantly (P < 0.01) more extended than the Normals (49 + 11"). Align- ing the dorsum of the third metacarpal with the dorsum of the forearm to define the neutral wrist position was not precise, but the neutral postures of the capitate bones did not differ in the Injured (-50 f lY), Uninjured (-46 + 23"), and Normal wrists (-45 f 90).

During wrist extension, the Injured and Un- injured lunates rotated 35% and 26% as much as the capitate, respectively, each rotating significantly less ( P < 0.001) than the Normal lunate, which rotated 69% as much as the capitate. Extension of the lunate in the Injured and Uninjured wrists were not different ( P = 0.77). During wrist flexion, the Injured and Unin- jured scaphoids rotated 80% and 89% as much as the capitate, respectively, each rotating significantly more ( P < 0.01) than the Normal scaphoid, which rotated 73% as much as the capitate. Flexion of the scaphoid in the Injured and Uninjured wrists were also not different from each other ( P = 0.53).

The authors' results support the hypothesis that SLIL tears alter carpal bone kinematics; however, in the group of eight patients the authors studied, their asymptomatic, uninjured contralateral wrist also was found to have abnormal carpal bone kinematics. These results were based on in vivo measurements made using a novel markerless CT-based methodology whose accuracy has been examined in vitro to be better than 2" and 2 mm for the scaphoid, h a t e and capitate. This method utilizes CT scans of the wrist in multiple static positions, so the kinematics are calculated between static postures. Whether an actual dynamic method would yield similar findings is unknown.

Previous reports lend support to the authors' findings. Other authors have demonstrated


bilateral soft tissue defects after unilateral wrist injury in 60% to 100% of the patient^.^,^^ In patients with a range of soft and hard tissue injuries, Feipel et all5 found no significant kinematic differences between the injured and the contalateral, asyptomatic wrist, but there was a significant difference between both wrists of the injured patients and the wrists of the uninjured volunteers in that study. The authors’ findings are in direct agreement with this finding. Feipel et all5 found individual variations in normal carpal bone motion, which is in contrast to the authors’ finding of minimal kinematic variability between normal wrists. The finding by Feipel et all5 of normal variation might be attributed to their motion tracking algorithm which has a lower accuracy than the authors’ algorithm.

The findings of the current study do not explain why carpal kinematics are abnormal in both wrists of patients with unilateral SLIL tears. The effect of age on carpal kinematics is not known, but the fact that the authors’ injured cohort was, on average, 12 years older than the healthy volunteer group may be a factor. Unilateral injury also may result in a complex neuromuscular response that affects carpal mechanics in both limbs. Also, a population of individuals may be predisposed to ligament injury becuase of abnormal carpal kinematics. Why there were none of these individuals in the authors’ healthy volunteer group is unclear. Further research is needed to confirm and understand the finding of bilateral abnormal carpal bone posture and kinematics in subjects with unilateral wrist ligament injuries.

SUMMARY AND MAJOR FINDINGS

The ability to measure in vivo three- dimensional skeletal kinematics noninvasively provides clinicians and scientists with a pow- erful new technique to study the carpus. This technique and others like it will permit the measurement of the entire carpus. and en- hance understanding of normal carpal kinematics. Such techniques also will permit the novel study of the injured and surgically reconstructed carpus. The authors’ approach has led to the development of a MBR algorithm. The noteworthy limitations of the authors’ ap-

proach are the use of CT and the intensive computational analysis. With this algorithm, the authors have evaluated initially the normal and abnormal kinematics of the carpus in several wrists positions. Their findings suggest that the mechanics of the carpus cannot be described completely by the row or column theories but by complex combinations of each that are dependent on the direction of wrist motion, at least and perhaps other yet undeter- mined factors, such as muscular activity and age. Also, the authors’ measurements of wrist motion have been limited to the main motions of flexion-extension and radioulnar deviation. The authors’ initial study of patients with unilateral scapholunate tears indicates that patients with these tears do have abnormal carpal kinematics, but that these abnormal kinematics may have existed before trauma. Hopefully, further studies by the authors’ group and others will permit a better understanding of these surprising findings.

References

1. Armstrong G: Rotational subluxation of the scaphoid. Can J Surg 11:306-314,1968

2. Berger RB, WF, Crowninshield RD, et a1 The scapholunate ligament. J Hand Surg 787-91,1982

3. Berger RC, Crowninshield RD, Flatt AE: The three- dimensional rotational behaviors of the carpal bones. Clin Orthop Re1 Res 167303-310,1982

4. Blevens AD, Light TR, Jablonsky WS, et al: Radiocarpal articular contact characteristics with scaphoid instability. J Hand Surg 14:781-790,1989

5. Bryce T On certain points in the anatomy and mechanism of the wrist-joint reviewed in light of a series of roentgen ray photographs of the living hand. J Anat Physiol31:59-79,1896

6. Burgess RC: The effect of a simulated scaphoid malu- nion on wrist motion. J Hand Surg 12(5 Pt 1):774-776, 1987

7. Cantor RM, Stern PJ, Wyrick JD, et al: The relevance of ligament tears or perforations in the diagnosis of wrist pain: An arthrographic study. J Hand Surg Am

8. Cautilli GP, Wehb6 MA: Scapho-lunate distance and cortical ring sign. J Hand Surg Am 16:501-503,1991

9. Craigen MA, Stanley JK Wrist kinematics: Row, column or both? J Hand Surg Br 20:165-170,1995

10. Crisco J, Neu C: In vivo carpal bone rotations linearly depend on the direction of wrist rotation. In Kobayashi M, Garcia-Elias M, Nagy L, et a1 (eds): OK, 2000, Florida

11. Crisco JJ, McGovern R D Efficient calculation of mass moments of inertia for segmented homogeneous three- dimensional objects. J Biomech 31:97-101,1998

12. Crisco JJ, McGovern RD, Wolfe SW Noninvasive technique for measuring in vivo three-dimensional carpal bone kinematics. J Orthop Res 179&100,1999

19~945-953,1994

230 CRISCO et a1

13. Dagum AB, Hurst LC, Finzel KC: Scapholunate dissociation: An experimental kinematic study of two types of indirect soft tissue repairs. J Hand Surg Am 223714- 719,1997

14. de Lange A, Kauer JMG, Huiskes R: Kinematic behav- ior of the human wrist joint: A roentgent stereophotogrammetric analysis. J Orthop Res 3:56-64,1985

15. Feipel V, Rooze M, Louryan S, et al: Bi- and three- dimensional CT study of carpal bone motion occur- ring in lateral deviation. Surg Radio1 Anat 14341-348, 1992

16. Frankel V The Terry-Thomas sign. Clin Orthop

17. Gilford W, Bolton R, Lambrinudi C: The mechanism of the wrist joint. Guy's Hospital Report 92:52-59,1943

18. Green D: Carpal dislocations and instabilities. In Green D (ed): Operative Hand Surgery. New York, Churchill Livingstone, 1993, pp 861-928

19. Horii E, Garcia-Elias M, An KN, Bishop AT, et a1 A kinematic study of luno-triquetral dissociations. J Hand Surg 16:355362,1991

20. Howard F, Fahey T, Wojeik E: Rotatory subluxation of the navicular. Clin Orthop 1043134-139,1974

21. Hurkmans HLP, Kooloos JGM, Meijer RS: Scapho- lunate dissociation and arthrodesis: An experimental study with lesions of the interosseous ligament and fu- sions with K-wires. Clinical Biomechanics 11:220-226, 1996

22. Imaeda T, An KN, Cooney WD: Functional anatomy and biomechanics of the thumb [review]. Hand Clin

23. Ishikawa J, Niebur GL, Uchiyama S, et al: Feasibility of using a magnetic tracking device for measuring carpal kinematics. J Biomech 30:1183-1186,1997

24. Kauer JM: Functional anatomy of the wrist. Clin Or- thop Re1 Res 149:9-20,1980

25. Kobayashi M, Berger A, Richard L: Normal kinematics of carpal bones: A three-dimensional analysis of carpal bone motion relative to the radius. J Biomech 30:787- 793,1997

26. Kobayashi MKGEM, Nagy L, Ritt MJPF, et al: Ax- ial loading induces rotation of the proximal carpal row bones around unique screw-displacement axes. J Biomech 30:1165-1167,1997

27. Li G, Rowen B, Tokunaga D, et al: Carpal kinematics of lunotriquetral dissociations. Biomed Sci Instrum

28. Lichtman DM, Scheider JR, Swafford AR, et al: Ulnar midcarpal instability: Clinical and laboratory analysis. J Am Hand Surg 6:515-523,1981

29. Linscheid R, Dobyns J: Treatment of scapholunate dissociation. Hand Clin 8:645452,1992

30. Linscheid RL, Dobyns JH, Beabout JW, et al: Traumatic instability of the wrist: Diagnosis, classification, and pathomechanics. J Bone Joint Surg Am 54:1612-1632, 1972

31. Logan SE,'Vannier MW, Bresnia SJ, et al: Wrist kinematic analysis using a 6 degree of freedom digitizer. In the Seventh Annual Conference of the IEEE/Engineer- ing in Medicine and Biology Society. 1985

32. Mayfield J, Mechanism of carpal injuries. Clin Orthop

33. Mayfield J: Pathogenesis of wrist ligament instability. In Lichtman D (ed): The Wrist and its Disorders. Philadelphia, WB Saunders, 1988, pp 53-73

34. Mayfield J, Johnson R, Kilcoyne R: Carpal dislocations: Pathomechanics and progressive perilunar instability. J Hand Surg 5:226-241,1980

35. Meade TD, Schneider LH, Cherry K Radiographic analysis of selective ligament sectioning at the carpal

129:321-322,1977

89-15,1992

27273-281,1991

149~45-54,1980

scaphoid: A cadaver study. J Hand Surg Am 15:855- 862,1990

36. Mior SA, Dombrowsky N: Scapholunate failure: A long term clinical follow-up [see comments]. J Manipulative Physiol Ther 15:255-260,1992

37. Navarro A: Luxaciones del carpo. Anales de la facultad de medicina 6:11%141,1921

38. Neu C, McGovern R, Crisco J: Kinematic accuracy of three surface registration methods in a three- dimensional wrist bone study. J Biomech Eng 122528- 533,2000

39. Nielsen PT, Hedeboe J: Posttraumatic scapholunate dissociation detected by wrist cineradiography. J Hand Surg Am 9:135-138,1984

40. Palmer A, Dobyns J, Linscheid R: Management of posttraumatic instability of the wrist secondary to ligament rupture. J Hand Surg 3:507-532,1978

41. Panjabi M, Krag M, Goel V A technique for the measurement and description of three-dimensional six degree of freedom motion of a body joint with application to the human spine. J Biomech 14:447460,1981

42. Patterson RM, Nicodemus CL, Viegas SF, et al: Normal wrist kinematics and the analysis of the effect of various dynamic external fixators for treatment of distal radius fractures. Hand Clin 13:129-141,1997

43. Pelizzari CA, Chen GT, Spelbring DR, et al: Accurate three-dimensional registration of CT, PET, and/or MR images of the brain. J Comput Assist Tomogr 13:20-26, 1989

44. Rominger M, Bernreuter W, Kenney P, et a1 MR imag- ing of anatomy and tears of wrist ligaments. Radio- Graphics 13:1233-1246,1993

45. Ruby LK, An KN, Linscheid RL, et al: The effect of scapholunate ligament section on scapholunate motion. J Hand Surg Am 12(5 Pt 1):767-771,1987

46. Ruby LK, Cooney WP, An KN, et al: Relative motion of selected carpal bones: A kinematic analysis of the normal wrist. J Hand Surg Am 13:l-10,1988

47. Ryu JY, Cooney WP, Askew LJ, et al: Functional ranges of motion of the wrist joint. J Hand Surg 16:409419, 1991

48. Sarrafian SK, Melamed JL, Gashgarian GM: Study of wrist motion in flexion and extension. Clin Orthop Re1 Res 126:153-159,1977

49. Savelberg H: Wrist joint kinematics and ligament Be- havior. The Netherlands, Nijmegen University, 1992

50. Sennwald GR, Zdravkovic V, Kern HP, et al: Kinematics of the wrist and its ligaments. J Hand Surg Am 18:805- 814,1993

51. Seradge HS, PT, Seradge E, et a1 Segmental motion of the proximal carpal row: Their global effect on the wrist motion. J Hand Surg Am 15a:236-239,1990

52. Shin SS, Moore DC, McGovern RD, et al: Scapholunate ligament reconstruction using a bone-retinaculum- bone autograft: A biomechanic and histologic study. J Hand Surg Am 23:216-221,1998

53. Short WH, Werner FW, Fortino MD, et al: A dynamic biomechanical study of scapholunate ligament sectioning. J Hand Surg Am 20:986-999,1995

54. Smith DK, An KN, Cooney WD, et al: Effects of a scaphoid 'waist osteotomy on carpal kinematics. J Orthop Res 7:590-598,1989

55. Taleisnik J: The ligaments of the wrist. J Hand Surg

56. Tang JB, Ryu J, Kish V Scapholunate interosseous ligament sectioning adversely affects excursions of radial

1:110-118,1976

wrist extensorand flexo; tendons. J Hand Surg Am

57. Tanz S: Rotation effect in lunar and perilunar dislocca- 22~720-725,1997

tions. Clin Orthop 57147-152,1968


58. Valero-Cuevas FJ: CF, Load dependance in carpal kinematics during wrist flexion in vivo. Clin Biomech 12154159,1997

59. Yin YM, Evanoff B, Gilula LA, et a1 Evaluation of selective wrist arthrography of contralateral asymptomatic wrists for symmetric ligamentous defects. AJR Am J Roentgen01 1661067-1073,1996 1979

60. Youm Y, McMurthy RY, Flatt AE, et al: Kinematics of the wrist: I. An experimental study of radial-ulnar deviation and flexion-extension. J Bone Joint Surg Am 60:423431,1978

61. Youm Y, Yoon YS: Analytical development in investigation of wrist kinematics. J Biomech 12:613-621,

Address reprint requests to

Joseph J. Crisco, PhD Rhode Island Hospital

Orthopedic Research-SWP3 593 Eddy Street

Providence, RI 02903

advances in the in vivo measurement of carpal kinematics

Documents