3-D Reconstruction of Injured Brain Structures

摘要

Initial research and development efforts were made to improve methods for 3-D computer graphics reconstruction of damaged brain tissues in living subjects. Some of the clinically more intriguing VHIA tapes were analyzed in order to establish whether it is feasible with such material to obtain accurate and reliable 3-D reconstructions. Four main issues evolved. 1. The only conclusive check on accuracy of computer graphics modeling is obtainable by making an accurate 3-D reconstruction following death, by reconstructing 3-D images from tissue sections, postmortem. 2. Automatic boundary detection is needed for potentially more accurate, and certainly more reliable, objective and essentially indefatigable detection methods and to provide the ability to reconstruct a sufficient number of cases to establish the database required for comparison between injured and normal neuroanatomic structures. 3. Superimposition of model of normal neuroanatomical structures with injured structures was accomplished, though a warping algorithm to accommodate 3-D warping is not available at this time. 4. Proper reconstruction of injured brain structures was not possible for reason of the narrow range of Hounsfiled numbers that depict intracranial structures, using CT Scans of that vintage.