体調急変検知に向けた心臓血管系モデルの構築

抄録

These days, the number of traffic accidents due to drivers' sudden diseases is increasing. Therefore, it is desired to detect sudden diseases by making use of drivers' biological signals measured while driving. To detect such diseases, we need to know how biological signals change with the disorders. However, it is difficult to measure physiological signals at sudden diseases and propose the method for detecting such conditions. In this study, we construct a cardiovascular model which can reproduce mechanism of a cardiovascular system, and calculate biological signals by assuming sudden diseases in the model. In the cardiovascular model, stroke volume, the blood volume pumped by a heart per beat, is determined based on the operating conditions of the system. Besides, we considered positions of baroreceptors and pressure sensors of a cardiac system in blood vessels. Then, we succeed to construct detection method for sudden diseases based on the calculated biological signals. We use K-nearest method and LOF method for a detection algorithm and found that combination of systolic blood pressure and heartbeat interval is adequate for detection, and LOF method is useful for the discrimination of disease.