The influence of anatomical shape variations of wrist bones on bone orientation values in CT scans.

INTRODUCTION

Four-Dimensional Computed Tomography (4DCT) shows promise in diagnosing scapholunate ligament (SL) lesions. Wrist motion analysis requires local coordinate systems (LCS) for carpal bones, which might be affected by bone shape variations. These variations can affect the extraction of scapholunate angle (SLA) and capitolunate angle (CLA), indicative for SL lesions. This study characterizes the impact of anatomical shape variations on LCS determination and subsequent SLA and CLA estimates.

METHODS

A statistical shape model (SSM) was created for the scaphoid, lunate, and capitate using CT scans from a 4DCT dataset. The quality of the SSM was assessed using a leave-one-out approach by calculating the root mean squared error (RMSE). Subsequently, LCSs per bone were assigned to the SSM while 3000 shape variations were introduced. The rotational deviation of the LCSs was calculated by a combined rotation around the three axes. Finally, the SLAs and CLAs were calculated.

RESULTS

The SSM was created using 106 scans with an RMSE of 0.38 mm. The 95th percentile of the rotational deviations of the LCSs was below 5°. The resulting range due to anatomical variations in the calculation of the SLA and CLA was 5.7°and 6.8°, respectively.

CONCLUSION

Bone shape variations have a minimal influence on the determination of LCSs and consequently on the resulting SLA and CLA estimates. Based on the naturally occurring variability of carpal angles found in literature, the ranges in the variation of the SLA and CLA are probably not clinically relevant but should be taken into account when performing inter-subject comparisons.