Real Time Cardiac Health Monitoring with DSNN on Low Power VLSI Architecture

Abstract

The early detection of cardiac abnormalities, such as arrhythmias, is crucial for preventing life threatening conditions and enabling timely medical intervention. Traditional electrocardiogram (ECG) monitoring systems face significant limitations, including high power consumption, bulky hardware, and limited real-time capabilities, making them unsuitable for continuous and portable health monitoring. To address these challenges, this paper presents a real-time cardiac health monitoring system leveraging a Data Shifting Neural Network (DSNN) implemented on a low-power Very Large Scale Integration (VLSI) architecture. The proposed DSNN model enhances detection accuracy by processing original and shifted ECG sequences in parallel, reducing false detections with minimal computational overhead. Integrated with wearable ECG/PPG sensors, neuromorphic hardware, and edge-based processing, the system achieves ultralow power consumption, high classification accuracy (≥95%), and low-latency performance (<30 ms), making it ideal for continuous monitoring in real-world environments. The architecture supports wireless connectivity through IoT, BLE, and 5G technologies, enabling seamless remote monitoring and telemedicine integration. Experimental results demonstrate the system’s robustness in dynamic scenarios, offering reliable, real-time arrhythmia detection in portable devices such as smart watches, ECG patches, and implantable monitors. With advancements in neuromorphic computing and adaptive learning, this work lays the foundation for next generation intelligent healthcare systems, empowering personalized, preventive, and accessible cardiac care.