Medical operation training system using patient models restored automatically from medical images and knowledge

在传统的医疗手术模拟系统中，由于器官模型是手工制作的，因此难以实现根据患者的手术。随着高分辨率数字图像和计算机辅助医疗诊断（CAD）的发展，从医疗图像恢复患者的人体模型和分割每个器官成为可能。然后，通过将恢复的体素模型转换为四面体模型，不仅可以利用手术器械实现器官的变形和切除，而且可以实现操作者在手术中感觉到的触觉。本研究的目的是开发一个系统来模拟开颅手术夹闭术和弹簧圈栓塞术，预防脑动脉瘤破裂，这可能是蛛网膜下腔出血的情况成为最近的问题。真实的时间开发的必要技术，使用多核和GPU，此外，通过共享远程位置之间的系统，我们成功地使非熟练的医生经验的熟练的医生和触觉感觉的医生感觉这样做。

In the conventional medical operation simulation system, it is difficult to realize the operation according to a patient as the organ model is made by hand By advancing of high-resolution digital image and computer assisted medical diagnosis (CAD), restoration of human body model of a patient from medical image and segmentation of each organ become possible. Then by translating the restored voxel model into tetrahedral one, implementation of not only deformation and abscission of organ with surgical instruments but also the haptic feeling an operator feels in the operation becomes possible. This research aims at developing a system to simulate the craniotomy clipping art and the coil embolectomya art ; the preventive measure of cerebral aneurysm rupture which is possible case of subarachnoid hemorrhage becoming problem recently. Real time development of necessary techniques using multi-core and GPU, Furthermore, by sharing the system between remote location, we are successful in making non-skill physician experience manipulation of a skilled physician and haptic sensation the physician feels in so doing.