Scalable coding and link adaptation for machine-to-machine telecardiology

Abstract

Medical telemetry is one of the most demanding applications in recent wearable computing era. Telecardiology, which uses the power of telecommunications for the remote diagnosis and treatment of heart diseases, is one of the key telemetry applications that leverages IoT-based technologies to improve patient care. Based on recent advances in wearable sensors and telecommunication technologies, this thesis proposes a universal platform for wearable daily cardiac monitoring service.

First, we propose an adaptive framework for layered representation and transmission of ECG (electrocardiography) data that can accommodate a time-varying wireless channel on cellular networks. The representation, combined with the layer-based earliest deadline first (LB-EDF) scheduler, ensures that the perceptual quality of the reconstructed ECG signal does not degrade abruptly under severe channel conditions and that the available bandwidth is utilized efficiently. Simulation shows that the proposed approach significantly improves the perceptual quality of the ECG signal reconstructed at the remote monitoring station.

Then we extend the proposed adaptive framework to support time-critical medical applications. In fact, the use of wireless technologies has been avoided for medical situations that demand instantaneous cardiac monitoring because of their considerable and nondeterministic end-to-end latency. This thesis introduces a universal platform for machine-to-machine (M2M) telecardiology over cellular networks, along with a novel conservative modulation and coding scheme to minimize and stabilize the delay down to 10 ms of ultra-low latency level, incurred during the process of ECG transmission over a wireless medium while maintaining the desired level of ECG pattern quality required for improving the chance of its interpretation. Machine-type communication (MTC) system is adopted for the delivery of patient ECG data to benefit from its inherent reliability, pervasiveness, security, and performance of 4G long term evolution (LTE) technologies with reduced cost and enhanced coverage. Extensive evaluations indicate that the proposed system provides a sufficient level of service for medical-grade instantaneous ECG monitoring in significantly deteriorated channel conditions.