Statistical Bone Shape Analysis for Image Free Surgery
Mauricio Reyes, González-Ballester, Marius George Linguraru · Bern Open Repository and Information System (University of Bern) · 2007
Statistical shape analysis techniques employed in the medical imaging community, such as Active Shape Models or Active Appearance Models, rely on Principal Component Analysis (PCA) to decompose shape variability into a reduced set of interpretable components. Our model uses point distribution models (PDM) for the representation of shapes. The point association is initialized using shape-based surface correspondence and optimized with the Minimum Description Length (MDL) criterion. The model fitting algorithm is formulated as a least square error minimization regularized by the Mahalanobis distance of the predicted model. We present quantitative and qualitative results obtained for the case of statistical shape analysis of bones, with applications to image free surgery.