Wearable ECG sensors heavily rely on cloud-based or other powerful devices for remote diagnosis and computations due to their limited resources. Prolonged dependence on the cloud may impede the advancement of wearable ECG sensors. Addressing major concerns, such as specialized diagnosis algorithms, efficient communication strategies, and security protocols for resource-constrained devices like wearable sensors, is crucial. In this paper, we propose an innovative approach to partitioning the ECG diagnosis algorithm among multiple processing layers, starting from the wearable sensor. Additionally, we introduce a novel communication strategy: open-loop and closed-loop switch mode algorithms for wearable and edge devices. The open-loop mode conserves battery power and is initiated in less severe conditions, while the closed-loop mode, with a feedback channel, is activated in more critical situations. Our paper also introduces the concept of the Medical Virtual Chain (MVC) to secure access and document ECG diagnosis results and other medical data. The MVC plays a vital role in authorizing and authenticating the incoming processing layer within the network through an entrance application and layer signature. We present satisfactory results from multiple diagnosis experiments using various ECG datasets, with 2D-CNN classification accuracy ranging from 91.11 % to 99.3%.