A novel method for diagnosing hypertension based on wrist pulse signals and constitution

Pulse diagnosis is of great research significance and practical value in clinical diagnosis. However, current research has focused on objectifying the pulse signal and has neglected the influence of the TCM constitution on the pulse signal and disease. The effective utilization of constitution factors is the urgent need for pulse analysis and the incorporation of constitution factors into pulse diagnosis to diagnose hypertension will help to improve the TCM connotation of pulse diagnosis. To solve the problems, we first developed a simple, convenient, and noninvasive multichannel pulse sampler, and constructed a public pulse dataset. Then we propose a pulse preprocessing framework, which integrates four pulse preprocessing operations. This method makes the pulse preprocessing operation more efficient. Then we used statistical analysis to determine the hypertension susceptibility constitution factors. The results showed that Yang-deficiency (YaD), Yin-deficiency (YiD), and Phlegm-dampness (PD) were significantly correlated with whether the subjects suffered from hypertension. The pulse feature samples that included hypertension susceptibility constitution factors had the highest classification accuracy in distinguishing between the two types of healthy samples and hypertensive samples. This shows that the inclusion of constitution factors in the pulse features is important for disease classification.