Continuous blood pressure prediction based on hierarchical adaptive algorithm

Continuous hierarchical adaptive blood pressure prediction method based on pulse condition is proposed in this paper, the linear blood prediction models and the network training parameters of different population and different pulse conditions can be obtained through a large number of experimental data. For one measurement, the first level blood prediction based on different pulse conditions is carried on to get the corresponding linear blood prediction model in order to get the primary predictive blood pressure. Then the trained BSO-BP neural network is used to adjust the blood pressure of the second stage by the pulse characteristic values, and the final blood pressure can be obtained. The experimental results showed that the blood pressure of patients with hypertension and normal people are improved by the proposed method. The results of continuous measurement also show great follow-up ability, strong ability to capture abnormal blood pressure, which means the method can predict hypertension effectively.