A Deep Learning Based Approach for Cardiovascular Monitoring Using Seismocardiogram Signals

The electrocardiogram (ECG) is a frequently utilised electrical signal for cardiac activity, and the seismocardiogram (SCG) signal captures the mechanical vibrations and movements of the chest wall induced by the heartbeat. In this study, we present an additional framework which works as an extension converion SCG data to ECG signals and vice versa using an attentive network of cycle-generating adversaries [1]. Our proposed framework includes an identification module that can recognise the difference between ECG and SCG signals using a novel real-time algorithm that detects the complexes in the respective signals. Our approach combines subjective visual analysis and objective performance measurement using a publicly available combined ECG, breathing and seismocardiogram (CEBS) database. For the standard 24-hour MIT/BIH arrhythmia database, this algorithm correctly detects 93.01% of the QRS complexes. This integration greatly improves the performance and accuracy of generator and discriminator networks are used to synthesise the ECG and SCG signals [1]. The findings of our investigation demonstrate the consistency and dependability of our methods.