Header menu link for other important links
Lossless electrocardiogram compression using segmented beats by dynamic bit allocation and Huffman encoders: A study on performances
Published in CRC Press/Balkema
Pages: 87 - 92
Lossless compression of Electrocardiogram (ECG) can ensure preservation of clinical features in the dataset. In this work, we applied two such encoders, Huffman coder and Dynamic Bit Allocation (DBA) to compare their performance for single lead ECG compression. Each ECG beat was segmented into three zones P, QRS and T. The encoders were applied to compress biased first order difference for P and T and second order difference for QRS with ptbdb and mitdb data at 8 bit quantization and resampled at 1 kHz. An average compression ratio for of 5.69 and 5.14 were obtained with mitdb and ptbdb data respectively using Huffman coder, where as the same using DBA were 3.99 and 3.97 respectively. The Huffman encoder showed better performance in terms of bits per sample efficiency (1.46 vs. 1.99) and nearly same for data overhead per packet (1.06% vs. 1.04%). However, the DBA encoder showed lower time complexity. © 2017 Taylor & Francis Group.
About the journal
JournalComputational Science and Engineering - Proceedings of the International Conference on Computational Science and Engineering, ICCSE2016
PublisherCRC Press/Balkema
Open AccessNo