 | | | | 1) 東北大学 大学院工学研究科 バイオロボティクス専攻 |
| | Abstract: Cardiovascular diseases, particularly ischemic heart disease such as myocardial infarction, are the leading cause of death in the industrialized world. Cerebrovascular diseases, another group of vascular disorders, are also a major cause of death. These vascular diseases share a common background, atherosclerosis, and a common final event, the breakage or destruction of vascular structure. Subarachnoid hemorrhage, which is another very acute and fatal disease, is the direct result of the rupture of a cerebral aneurysm. Thus, both the onset and final outcome of fatal vascular diseases are related to mechanical events that occur on the vascular wall, probably due to alterations in blood flow. We need to use computational studies to elucidate the mechanism of such disease, to refine the diagnostic measures, and to develop therapeutic modalities, either invasive or non-invasive. This lecture discusses why computational study is necessary, how a computational model is built, the pre-requisites for computation, and the pitfalls of interpreting computational studies. Despite the rapid advances in invasive and non-invasive imaging technology in clinical medicine, the resolution and reproducibility of non-invasive methods, such as magnetic resonance imaging (MRI), are still insufficient for automatic modeling. Therefore, we believe that any clinical application should fully utilize human pattern recognition and image reconstruction abilities. Consequently, we developed a comprehensive computational analysis support system that includes the preprocessing of medical images, segmentation, structure registration, database and human interfaces, mesh generation, and the final evaluation of the computed results. To make the database more usable, a novel method of searching and finding patient datasets using morphological and physiological queries was developed. Although many unsolved problems remain, we now expect that a prototype was developed for clinical application, and we are now able to accumulate the necessary data and experience from real patients once this system becomes available to medical practitioners. |
| | | |
