Author_Institution :
Sch. of Basic Med., Shanghai Univ. of Traditional Chinese Med., Shanghai, China
Abstract :
Objectives: Coronary Heart Disease (CHD) is the leading cause of death for adults worldwide. Pulse diagnosis is a specific diagnostic method in the Traditional Chinese Medicine (TCM). The purpose of this study is to identify the differences of the pulse among CHD patients, cardiovascular disease patients without CHD (Non-CHD) and healthy individuals (Normal). Additionally, to calculate the classification recognition rates with the significant pulse parameters. Methods: 251 subjects were recruited (102 in the CHD group, 94 in the Non-CHD group and 55 in the Normal group). Pulse signals were collected by a pressure sensor (ZBOX-I Digital Pulse Analysis System, designed by Shanghai University of TCM, China) from left-Guan. The time-domain parameters were extracted from the ZBOX-I system. The original data were transformed to frequency-domain and then the relative parameters were calculated. The differences of the parameters among these three groups were compared and contrasted. Moreover, a Fisher Linear Discriminant (FLD) was used to verify the accuracy of the significant parameters. Results: This research showed that the parameters such as: h1, h3, h4, t1, t4, t5, h3/h1, h4/h1, w/t, pulse rate and maximum power value were significantly different. About 81.5% accuracy was attained between CHD group and Normal group in the classification, while 68.5% accuracy between CHD group and Non-CHD group. Conclusions: These results demonstrate that some parameters in time-domain and frequency-domain are good indicators to distinguish CHD patients from healthy individuals as well as cardiovascular disease patients without CHD. Furthermore, they prove that pulse diagnosis, which is a non-invasive and inexpensive method, is a valuable auxiliary tool for the CHD diagnosis and its treatment evaluation.
Keywords :
cardiovascular system; diseases; frequency-domain analysis; medical signal detection; patient diagnosis; CHD patients; Fisher Linear Discriminant; ZBOX-I Digital Pulse Analysis System; classification recognition rates; coronary heart disease; death; frequency domain transform; maximum power value; pressure sensor; pulse diagnosis; pulse signal analysis; time domain parameters; traditional chinese medicine; Cardiovascular diseases; Educational institutions; Heart; Medical diagnostic imaging; Time-domain analysis;