Compressed Sensing Framework for Multi-channel ECG Signals

Abstract

Electrocardiogram (ECG) signals are the manifestation of underlying electrical phenomena of heart, which are responsible for its various functionalities. ECG is used as an important non-invasive tool by the cardiologists to diagnose and assess a wide range of cardiac ailments. With advancements in wireless body area network (WBAN) technologies, significant research has been done in recent decades to develop low-cost personalized remote health monitoring systems for next-generation of e-healthcare solutions. With ever increasing number of cardiovascular patients, WBAN-enabled ECG telemonitoring has generated significant interest among the biomedical community. Ambulatory ECG enables remote monitoring of vital heart parameters and allows early medical interventions in case of life-threatening heart diseases. However, existing ECG monitoring systems still suffer from various challenges, such as limited autonomy, bulkiness, limited functionalities, etc. In recent years, compressed sensing (CS) has emerged as a promising framework to address these challenges. Low-complex and highly energy-efficient data reduction procedure of CS makes it an attractive choice over traditional wavelet-based techniques for embedded on-node ECG data compression in resource-constrained telemonitoring applications.