System Quality of Human Physiological Signal Based Medical Auxiliary Diagnosis

In order to improve the quality of mobile medical assistant diagnosis, this paper proposes the quality research of medical assistant diagnosis system based on human physiological signal. Firstly, the human physiological signal is preprocessed by wavelet transform, and the signal is decomposed and reconstructed. Then, the feature extraction method based on windowing is used to extract the peak valley feature of the hyperbolic pulse of the denoised signals. Iterative self-organizing data analysis method is used to classify the extracted signal features. In addition, by changing the parameters between and within clusters, updating the cluster center and cluster category, and splitting or merging them, the quality of medical auxiliary diagnosis is improved. In the experimental part, 80 patients in Department of cardiovascular medicine and Department of cardiovascular surgery of a hospital were selected as the experimental objects, and 80 physiological signals were collected. Compared with the maximum likelihood method and semantic knowledge base method, the experimental results show that this method improves the quality of traditional medical auxiliary diagnosis. The diagnostic accuracy rate is higher than 98.5%, and the diagnostic sensitivity is higher than 96.5%.