A Computer-Aided Recognition Method of Heart Rate Deflection Point

Abstract

Lactate threshold or gas exchange threshold is commonly used to guide exercise intensity, but direct measurement of these two are never easy for general population. Among all physiological indicators, heart rate is very easy to obtain. And the heart rate deflection point is consistent with the lactate threshold during incremental exercise. However, previous studies suffer from expertise or a priori information requirement, computation inefficiency, lack of cohort diversity, etc. Based on prior knowledge, this contribution proposes a computer-aided methods to automatically identity heart rate intersection points by sections, and further optimization. As result, among 200 healthy college student volunteers, only 8 subjects fall beyond the 95% confidence interval in residual analysis. Therefore, a self-consistent, economic, noninvasive method to estimate the lactate threshold with heart rate data only is demonstrated.