ファジィモデリングを用いた加速度脈波分類(一般講演 A1)

抄録

The signals of acceleration plethysmogram is obtained as a fingertip content pulse wave detected by an optical sensor. It is observed nonivasively and is used mainly for estimation of arteriosclerosis and blood vessel age. In the estimation and classification of acceleration plethysmogram, the traditional method has a problem that its method cannot classify the plethysmogram exactly because of imperfect sampling of its waveform. In this paper, the authors propose a method of applying fuzzy neural network to taking the precise waveform information in the classification of plethysmogram data. The authors add a technique of genetic algorithm to the optimization of fuzzy reasoning using the steepest descent method. In a technique using the genetic algorithm, the gene can select some kinds of MSFs. This method models acceleration plethysmogram. In addition, the authors take the values of the waveform out of the modeling data as a vector, and classify the data of plethysmogram using self-organizing map. The advantages of this new method have been proved by comparing to the traditional method.