ECG Compression Based on Successive Differences.

This paper aims to reduce the amount of physical space the electrocardiogram data occupies and needs to be transferred, designing an efficient algorithm that can run on hardware commonly found in a medical device setting. The algorithm needs to be compatible with real-time data transfer and be fast and lossless. We introduce an algorithm inspired by pulse code modulation in telecommunication theory based on successive differences between neighboring electrocardiogram samples. The research hypothesis addressed in this paper finds if the new algorithm can reduce the data size by a meaningful amount and efficiently. Further on we tackle the research questions to find the level of distortion, compression, and processing speed of compression and decompression, along with the analysis of benefits and disadvantages. We reduced the storage space or bits to be transferred up to 46% reaching a compression ratio of 2.17 for an electrocardiogram benchmark dataset of up to 100 Mbits.