Title: Analysis of pulse diagnosis data for the elderly by using two analytical methods

Authors: Siyu Zhou; Atsushi Ogihara; Shoji Nishimura; Zhiwei Leng; Qun Jin

Addresses: Advanced Research Center for Human Sciences, Waseda University, Tokorozawa, Japan ' Faculty of Human Sciences, Waseda University, Tokorozawa, Japan ' Faculty of Human Sciences, Waseda University, Tokorozawa, Japan ' College of Humanities and Management, Zhejiang Chinese Medical University, Zhejiang, China ' Faculty of Human Sciences, Waseda University, Tokorozawa, Japan

Abstract: The use of information and communication technology (ICT) to analyse pulse diagnosis data has become one of the pathways of modernising traditional Chinese medicine (TCM). In this study, we used two methods of medical statistics and machine learning to analyse diagnosis data. We used the Youden index to evaluate the authenticity of the diagnosis and the Kappa statistic to evaluate the consistency of the diagnosis made by the pulse diagnosis instrument and the TCM doctor. The accuracy of a single pulse was almost 80%. The authenticity and consistency were acceptable. The k-NN method was used to match the diagnosis results of the pulse diagnosis instrument and the TCM doctor. The overall accuracy rate was 62%, similar to that in previous studies. In practice, medical statistical methods (Youden index and Kappa statistic) are used to determine the accuracy of a single pulse, and machine learning methods (k-NN method) are used to classify pulse matching.

Keywords: pulse diagnosis; data analysis; traditional Chinese medicine; TCM; k-NN; Youden index.

DOI: 10.1504/IJSHC.2019.101599

International Journal of Social and Humanistic Computing, 2019 Vol.3 No.2, pp.158 - 175

Available online: 06 Aug 2019 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article