Imaging recognition of morphological variants at the midcarpal joint.

PURPOSE: To compare the imaging methods for identifying the various morphological variations of the articular surfaces at the midcarpal joint.
METHODS: Thirteen cadaveric wrists were examined by plain neutral anteroposterior radiographs; 2-dimensional computed tomography (CT); 3-dimensional CT reconstruction, and 3-tesla magnetic resonance imaging (MRI). Carpal measurements were performed, and the parameters that defined the scaphoid, lunate, hamate, and capitate morphological types were investigated, with dissection being used as the definitive measure of morphology. The dissection findings were compared to the results of each imaging technique to determine the accuracy of morphological determination from each technique.
RESULTS: Lunate type was the most accurately identified morphological variant amongst all imaging techniques. Lunate type was most accurately determined from coronal MRI. A lunate with a small, cartilaginous ulnar facet (intermediate type) could be differentiated only by coronal MRI and dissection. Scaphoid type could not be determined accurately using any of the imaging modalities described. Capitate type was most accurately determined from coronal MRI. However, flat and spherical-type capitates could not be routinely differentiated from V-shaped capitates. Hamate type was most accurately determined from 3-dimensional CT reconstruction.
CONCLUSIONS: Accurate identification of carpal bone morphology is required to improve our understanding of carpal mechanics and pathology. Not all morphological features can be identified radiographically. Direct visualization is required to differentiate types of scaphoid, and to differentiate V-type capitates. MRI provides the most accurate identification of lunate type, and 3-dimensional CT provides the best method of differentiating hamate types.