This study illustrates that differences in meniscal geometry cannot be explained by scaling only, but that different meniscal shapes can be distinguished. The cluster analysis resulted in four clusters, yet these clusters represented two statistically different meniscal shapes, as differences between cluster 1, 2 and 4 were only present for principal component 1. Significant differences between male and female menisci were only found for principal component 1, which predominantly reflected size differences. Shape changes were most prominent in the cross-sectional plane, rather than in the transverse plane. Of the variation in medial meniscus geometry, 53.8% was found to be due to primarily size-related differences and 29.6% due to shape differences. This allowed the comparison of male and female menisci, and a cluster analysis to determine distinct morphological meniscus groups.

Each meniscus from the original dataset was then reconstructed as a linear combination of principal components. A principal component analysis was performed to determine the most important modes of geometry variation and the characteristic changes per principal component were evaluated. A statistical shape model was created, containing the meniscus geometries of 35 subjects (20 females, 15 males) that were obtained from MR images. Also we performed a cluster analysis to identify distinct morphological groups of medial menisci and assessed whether meniscal geometry is gender-dependent. Therefore, the aim of this study was to quantify 3D meniscus geometry and to determine whether variation in medial meniscus geometry is size- or shape-driven. The geometry-dependent functioning of the meniscus indicates that detailed knowledge on 3D meniscus geometry and its inter-subject variation is essential to design well functioning anatomically shaped meniscus replacements.

Shapes geometry a l