An Intelligent Wearable ECG Sensor in Intra-medical Virtual Chain Network and Inter-medical Virtual Chain Network

Today, distributed medical network necessitates considering various dimensions of decentralized diagnosis environments to improve present wearable ECG sensors in terms of their versatility and security. This paper proposes five important mechanisms for the enhancement of ECG sensors. The proposed mechanisms are the partition of diagnosis, switch modes as per the communication status and severity, feedback channel, emergency alert system, and medical virtual chain (MVC). The proposed partition of diagnosis is based on the involvement of processing nodes according to the computational need. Two switch modes open-loop and closed-loop switch modes are put forward to deal with unpredictable communication scenarios and the severity of the patient. An MVC-based medical network is proposed where communication exists in and between medical organizations. The non-partitioned diagnosis used a CNN inference module with low computational complexity to give a diagnostic accuracy of 95.67% at the local ECG sensor, whereas the partitioned diagnosis was completed with an accuracy of 99.17% at the edge tier. In case of a poor communication scenario and low severity level of the patient, switch mode was activated to open loop to find that it was nearly six times more efficient than the regular transmission. The average delay observed for an intelligent sensor accessing a copy of MVC was only 0.3708 ms. Likewise, the average delay to record medical sessions initiated by the respective intelligent sensors in multiple medical institutions was 14.17 ms. These simulation results for the partition of diagnosis in multiple layers, switch mode initiation and MVC-based node interactions indicated that the proposed mechanisms for the intelligent ECG sensors are efficient and effective and hence can be implemented in a real-world scenario.