Artifacts mitigation in ambulatory ECG telemetry

Abstract

In remote monitoring applications of vital signs including ECG, it is extremely important to ensure that the diagnostic integrity of the signals is not compromised due to the presence of sensing artifacts and channel errors. It is also important for the platform to be extremely power efficient in order to facilitate wearable sensors with user friendly form factors. We present a novel, low power application layer solution that is agnostic to wireless protocols and mitigates artifacts due to packet losses in Body Area Networks (BANs). In our previous work, we presented initial results based on this approach and demonstrated that greater than 99% beat detection accuracy can be achieved even at a packet loss rate as high as 20%. Our contributions in this work include validation of the above on a platform with an ultra low power wearable single lead ECG pendant. We present details of implementation and then extend the platform to mitigate ECG sensing artifacts including power line interference and baseline wandering. The proposed approach enables us to offload most of the complex processing from sensor nodes to the receiver node with better a battery budget, for improved sensor life. Finally, present a qualitative and quantitative assessment of the system.