A comparison of three types of pulse signals: Physical meaning and diagnosis performance

Abstract

Pulse diagnosis has been extensively applied in China and Ayuredic for thousands of years. Recently more and more research interests have been given on computerized pulse diagnosis where sensor techniques are used to acquire the pulse signal and machine learning techniques are adopted to analyze the health condition based on the acquired pulse signals. By far, a number of sensors had been employed for pulse signal acquisition, which can be grouped into three categories, i.e., the pressure sensor, the photoelectric sensor, and the ultrasound sensor. To guide the sensor selection for computational pulse diagnosis, in this paper we analyze the physical meanings and sensitivities of signals sampled by these three types of sensors. The complementary information of different sensors is discussed from both cardiovascular fluid dynamics and comparative experiments by evaluating the disease classification performance. Signals acquired using different sensors are sensitive to different physiological and pathological factors. By combining signals from different sensor, improved diagnosis performance can be obtained.