A Hybrid AI Framework with Automatic Gain Control for Real-Time ECG Signal Monitoring

Monitoring of Electro Cardio Gram (ECG) signals is essential for the early identification and diagnosis of cardiovascular disorders. However, noise, motion distortions, and amplitude changes brought on by patient-specific circumstances and inconsistent sensors frequently make real-time ECG signal capture difficult. This study suggests a Hybrid Artificial Intelligence (AI) framework combined with an Automatic Gain Control (AGC) system for reliable and flexible ECG signal monitoring in real-time settings in order to overcome these problems. By normalizing dynamic amplitude fluctuations through adaptive signal preprocessing with AGC, the suggested system guarantees constant signal quality across a range of patient profiles. A dual-stage AI pipeline that combines a Long Short-Term Memory (LSTM) network for chronological pattern recognition then classification with a Convolutional Neural Network (CNN) for feature extraction comes next. The hybrid model improves accuracy by successfully differentiating between normal and pathological cardiac events, including bradycardia, tachycardia, and arrhythmia. The AGC mechanism maintains ideal gain levels in spite of input signal unpredictability by using feedback control based on real-time signal statistics to increase robustness. Using publically accessible datasets like the MIT-BIH Arrhythmia Database, experimental evaluations were carried out. The suggested system outperformed traditional fixed-gain systems by achieving a classification accuracy of 98.4%, sensitivity of 97.8%, and a 35% reduction in signal-to-noise distortion. Wearable health monitoring systems and ICU-grade diagnostic platforms can benefit from the real-time adaptability and enhanced detection performance that come with integrating AGC with AI. Future research will examine the integration of reinforcement learning for dynamic control optimization and on-chip implementation for edge devices.