Electrocardiogram Compression Using Adaptive Decomposition Techniques for Smart Healthcare System

Purpose: Electrocardiogram (ECG) compression is a critical aspect of smart healthcare systems, to efficiently store, transmit, and process ECG data. The research aims to find an efficient compression technique that reduces storage and transmission requirements without compromising diagnostic information, especially in the context of remote monitoring, telemedicine, and the increasing use of wearable devices. Method: The proposed approach integrates Adaptive Variational Mode Decomposition (AVMD) with Wavelet Packet Transform (WPT) for signal decomposition and transform coding. AVMD adaptively decomposes the signal into Intrinsic Mode Functions (IMFs), while WPT enhances time-frequency resolution and maintains a balanced representation of the signal. The framework is further optimized through arithmetic coding for efficient data compression. Experiments were conducted on ECG signals from the Physionet and Mendeley databases to validate the method's performance and reliability. Results: The proposed method achieved a compression ratio of 41.19 and a Percentage Root Mean Square Difference of 0.29 for the Massachusetts Institute of Technology-Beth Israel Hospital Supraventricular Tachyarrhythmia database record no. 209, demonstrating superior compression efficiency and minimal signal distortion compared to existing techniques. Conclusion: The integration of AVMD and WPT, combined with arithmetic coding, offers a highly effective solution for ECG data compression. The method ensures efficient storage and transmission of ECG data while preserving critical diagnostic information, making it a robust choice for smart healthcare.