Wavelet Based Analysis of the Heart Rate during Recovery in a Guided Rehabilitation Exercise Program

Exercise testing as a part of the Cardiac Rehabilitation (CR) has been reported to be beneficial for patients with Coronary Artery Disease. Abnormal heart rate during the recovery phase of the exercise has been proven to contain information concerning patient's clinical status. Many studies also denoted the significance of the first two minutes of the recovery as a predictor of mortality. In this study, employing signals recorded from patients during a CR exercise program at home, instead of a stress test in a clinical environment, we analyzed the HR signal toward the understanding of the balance of sympathetic and parasympathetic nervous system at the start of the program and after its completion. Features from time and frequency domain, already used for the evaluation of the exercise, were implemented. As an amelioration of the patient's status is detected based on these features, while the CR program proceeds, we used wavelet domain features in order to understand the modification of the sympathovagal activity with regards to the evolution of the recovery phase. Unlike the already used features, the wavelet domain ones are able to reveal information concerning the modification of the Autonomous Nervous System balance during the recovery which is achieved as the CR program proceeds, as well as the importance of the 2 first minutes of the recovery, even in a CR program performed at home.