Design and Evaluation of ECG Compression Techniques Using FPGA

Electrocardiogram (ECG) signals are crucial for diagnosing and monitoring heart health but can produce significant data files that create challenges for storage, transmission, and analysis in eHealth systems. This paper presents various ECG compression techniques implemented on Field-Programmable Gate Arrays (FPGAs) to mitigate these challenges. The study covers lossless methods like Run-Length Encoding (RLE) and Differential Pulse Code Modulation (DPCM) as well as lossy methods such as Wavelet Transform and Singular Value Decomposition (SVD). Performance metrics, including Compression Ratio (CR), Percent Root Mean Square Difference (PRD), Signal-to-Noise Ratio (SNR), Power Consumption, and Latency, are analyzed to assess their effectiveness. The findings indicate that while SVD on PYNQ-Z2 FPGA achieves the highest compression ratio, methods like SPIHT and hybrid approaches provide a balanced trade-off between compression efficiency and signal fidelity. Future work is suggested to focus on optimizing power consumption, improving compression ratios, and reducing latency. This paper provides insights into developing efficient, real-time ECG compression solutions suitable for resource-constrained eHealth applications.