Cortical Surface Registration Using Texture Mapped Point Clouds and Mutual Information

An inter-modality registration algorithm that uses textured point clouds and mutual information is presented within the context of a new physical-space to image-space registration technique for imageguided neurosurgery. The approach uses a laser range scanner that acquires textured geometric data of the brain surface intraoperatively and registers the data to grayscale encoded surfaces of the brain extracted from gadolinium enhanced MR tomograms. Intra-modality as well as inter-modality registration simulations are presented to evaluate the new framework. The results demonstrate alignment accuracies on the order of the resolution of the scanned surfaces (i.e. submillimetric). In addition, data are presented from laser scanning a brain's surface during surgery. The results reported support this approach as a new means for registration and tracking of the brain surface during surgery.

• Type: Conference
• Year: 2002
• Where: MICCAI
• Authors: Tuhin K. Sinha, David M. Cash, Robert J. Weil, Robert L. Galloway, Michael I. Miga
• Keywords: Brain Surface, Medical Imaging, MICCAI 2002, Textured Geometric Data, Textured Point Clouds