Estimation of Human Heart Activity Using Ensemble Kalman Filter

• Main Authors: Pradhnya Arun Priyadarshi, Surender Kannaiyan
• Format: Article
• Language: English
• Published: IFSA Publishing, S.L. 2017-02-01
• Series: Sensors & Transducers
• Subjects: Heart model; Ensemble Kalman filter; Electrocardiogram; Non-linear systems; State estimation techniques.
• Online Access: http://www.sensorsportal.com/HTML/DIGEST/february_2017/Vol_209/P_RP_0220.pdf
• ISSN: 2306-8515; 1726-5479

Heart beat measurement techniques come across various challenges. Electrocardiogram (ECG) obtained sometimes does not reveal complete information about electrochemical activity of human heart, because of which functioning of heart cannot be studied properly. In this paper Ensemble Kalman Filter (EnKF) is used to generate ECG signal efficiently with better accuracy such that the drawbacks of current techniques are eliminated. Here EnKF is applied to second order mathematical model of human heart, input applied to this mathematical model is a pacemaker signal. The initial values of heart muscle movements and electrochemical activity as a discrete data set are used and prediction steps are commenced. EnKF uses ensemble integration technique to model error statistics which helps obtaining more precise output. The results are obtained with negligible sum squared error, therefore the ECG obtained using EnKF can diagnose the disease related to heart with better accuracy.