Comparison of Imaging and Data Prediction Compression Methods for Implanted Real-Time Peripheral Nervous System

This study focuses on the problem of reducing the transmission rate of electroneurographic (ENG) signals for implantable medical devices. Such devices represent a significant innovation in the healthcare sector. We examine seven compression algorithms by implementing a variety of techniques that includes, among the others, image compression and predictors. Our results show that Fractal Compression (FC) and Vector Quantization (VQ) algorithms are the most effective, with a low value of percentage root difference (PRD), acceptable Compression Ratio (CR), and temporal redundancy. The analysis suggests that the optimal time window for compression should be between 10 ms and 50 ms. Our findings indicate that the FC and VQ algorithms could be suitable for real-time application, opening promising avenues for future research in the field of neural interfaces.