Research on Cardiovascular Disease Recognition Method Based on Pulse Signal

Nowadays, the morbidity and mortality of cardiovascular disease are increasing year by year, seriously endangering human health. During the onset of cardiovascular disease, some physiological signals in the human body will mutate. It is of great significance to use these physiological signals to identify and predict cardiovascular diseases. In this paper, the fingertip pulse wave data of patients with cardiovascular disease and normal people are collected, signal preprocessing is performed, time-frequency features are extracted by wavelet transform, and ResNet network and MobileNetV3 network are built for training and effect comparison. The experimental results show that the two Each model can classify the pulse wave signals of patients with cardiovascular disease and normal people with high accuracy, so as to realize the identification and prediction of cardiovascular diseases. Among them, the accuracy rate of the ResNet network reaches 75%, and the MobileNetV3 network consumes less resources. The accuracy rate reaches 76% in this case.