Pulse diagnosis using signal processing and machine learning techniques
Date
2021-05
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
For many centuries, pulse diagnosis has been a technique studied and applied in determining the state of health of the human body. However, its subjective nature puts it at a disadvantage in presenting an accurate diagnosis of the human body. In this project, practical research is done to standardize pulse diagnosis by acquiring body signals and applying signal processing and machine learning techniques for analysis. A prototype acquisition system is designed to obtain the body signals such as the pressure pulse waves from arteries and ECG signals. The system's efficiency was verified using cross-correlation analysis between the data acquired from the system and the standard data from Lei Zhang’s database and MIT physio net database. For the diagnostic system for the signal analysis, the time domain, frequency domain, and time-frequency domain of the signal processing techniques are adopted to extract features such as the power spectral density to be further used for classification and distinction between different signal groups. Applying the different techniques showed a distinction between the different groups of signals that aided the Support Vector Machine and K-Nearest Neighbour classification models to achieve above 90% and 80% accuracy. The experimental process and results provided insight into ways pulse diagnosis could be standardized for healthcare services.
Description
Capstone Project submitted to the Department of Engineering, Ashesi University in partial fulfillment of the requirements for the award of Bachelor of Science degree in Electrical and Electronic Engineering, May 2021
item.page.type
Capstone project
item.page.format
Keywords
health diagnosis, medical technique