Enhanced Data Security and Arrhythmia Detection With AE-3S AND DC2SNN Using ECG Signals

Arrhythmia is characterized by irregular heartbeats, often caused by abnormalities in heart rate or rhythm. ECG signals are essential for accurately detecting and classifying arrhythmias, aiding in identifying specific types for appropriate treatment. ECG signals contain sensitive health information, but none of the existing works, concentrated on securing the ECG signal processing in VLSI design. Therefore, this paper proposes AE-3S and DC2SNN for both data security and Arrhythmia detection. In the suggested process, EICFMD significantly boosts the signal quality through dynamic artifact suppression, whereas NGQSO fine-tunes the choice of features with the highest discrimination power, thus making AE-3S encryption and DC2SNN-based irregular heartbeat classification more efficient together. Initially, signal is converted from analog to digital and then undergoes preprocessing through various steps. Next, PTA is utilized for peak detection and time series is constructed on the preprocessed signal. Subtle characteristics of the signal are also preserved. Features are then extracted from the detected peaks, constructed time series, and preserved characteristics. From these features, NGQSO and AE-3S are utilized for feature selection and data security. Finally, Arrhythmia detection is performed using DC2SNN. Thus, the proposed model efficiently secures ECG signals and detects arrhythmias with high accuracy of 99.15%.