Sulcal Pattern and Morphology of the Superior Temporal Sulcus

this paper, we demonstrated the possibility performing structural morphometry using automatic image analysis tools to extract the cortical folds (Mangin et al., 1995a,b). The fold labeling according to the sulcus nomenclature, however, was performed manually by a neurosurgeon. We have recently developed a system to perform this labeling automatically, using a set of 500 neural networks trained on a manually labeled database (Riviere et al., 2002). Each neural network is in charge of a local anatomical feature. This opens the door to automatic "Structural Based Morphometry" (SBM), that is to say a large-scale morphometric study similar to those performed with a voxel-based approach. We have also reported the advent of new visualization tools dedicated to the 3D shape of the folding that can be used to improve the current understanding of the folding patterns. These tools can be used to study the variability of the buried gyri (pli de passage) which often interrupt the usual sulci. Future work consists of the validation and automatic recognition of the foetal sulcal root patterns, which would improve the reliability of the morphometric attributes used for the automatic studies. This work will rely on longitudinal studies of the folding process and new mathematical methods to infer the buried gyri localization for the curvature of the cortical surface (Cachia et al., 2001