Data Compression of Ambulatory ECG by Using Multi-Template Matching and Residual Coding

Takanori UCHIYAMA; Kenzo AKAZAWA; Akira SASAMORI

Data Compression of Ambulatory ECG by Using Multi-Template Matching and Residual Coding

Takanori UCHIYAMA, Kenzo AKAZAWA, Akira SASAMORI

Full Text Views

0

Cite this

Summary :

This paper proposed a new algorithm of data compression for ambulatory ECG, where no distortion was included in the reconstructed signal, templates were constructed selectively from detected beats, and categorized ECG morphologies (templates) could be displayed in decoding the compressed data. This algorithm consisted of subtracting a best-fit template from the detected beat with an aid of multi-template matching, first differencing of the resulting residuals and modified Huffman coding. This algorithm was evaluated by applying it to ECG signals of the American Heart Association (AHA) data base in terms of bit rates. Following features were indicated. (1) Decompressed signal coincided completely with the original sampled ECG data. (2) Bit rate was approximately 800 bps at the appropriate threshold 50-60 units (1 unit2.4µVolt) for the template matching. This bit rate was almost the same as that of the direct compression (encoding the first differenced signal of original signal). (3) The decompressed templates could make it easy to classify the templates into the normal and abnormal beats; this could be executed without fully decompressing the ECG signal.

Publication: IEICE TRANSACTIONS on Information Vol.E76-D No.12 pp.1419-1424

Publication Date: 1993/12/25

Publicized

Online ISSN

DOI

Type of Manuscript: Special Section PAPER (Special Section on ECG Data Compression)

Category

Cite this

Copy

Takanori UCHIYAMA, Kenzo AKAZAWA, Akira SASAMORI, "Data Compression of Ambulatory ECG by Using Multi-Template Matching and Residual Coding" in IEICE TRANSACTIONS on Information, vol. E76-D, no. 12, pp. 1419-1424, December 1993, doi: .
Abstract: This paper proposed a new algorithm of data compression for ambulatory ECG, where no distortion was included in the reconstructed signal, templates were constructed selectively from detected beats, and categorized ECG morphologies (templates) could be displayed in decoding the compressed data. This algorithm consisted of subtracting a best-fit template from the detected beat with an aid of multi-template matching, first differencing of the resulting residuals and modified Huffman coding. This algorithm was evaluated by applying it to ECG signals of the American Heart Association (AHA) data base in terms of bit rates. Following features were indicated. (1) Decompressed signal coincided completely with the original sampled ECG data. (2) Bit rate was approximately 800 bps at the appropriate threshold 50-60 units (1 unit2.4µVolt) for the template matching. This bit rate was almost the same as that of the direct compression (encoding the first differenced signal of original signal). (3) The decompressed templates could make it easy to classify the templates into the normal and abnormal beats; this could be executed without fully decompressing the ECG signal.
URL: https://global.ieice.org/en_transactions/information/10.1587/e76-d_12_1419/_p

Copy

@ARTICLE{e76-d_12_1419,
author={Takanori UCHIYAMA, Kenzo AKAZAWA, Akira SASAMORI, },
journal={IEICE TRANSACTIONS on Information},
title={Data Compression of Ambulatory ECG by Using Multi-Template Matching and Residual Coding},
year={1993},
volume={E76-D},
number={12},
pages={1419-1424},
abstract={This paper proposed a new algorithm of data compression for ambulatory ECG, where no distortion was included in the reconstructed signal, templates were constructed selectively from detected beats, and categorized ECG morphologies (templates) could be displayed in decoding the compressed data. This algorithm consisted of subtracting a best-fit template from the detected beat with an aid of multi-template matching, first differencing of the resulting residuals and modified Huffman coding. This algorithm was evaluated by applying it to ECG signals of the American Heart Association (AHA) data base in terms of bit rates. Following features were indicated. (1) Decompressed signal coincided completely with the original sampled ECG data. (2) Bit rate was approximately 800 bps at the appropriate threshold 50-60 units (1 unit2.4µVolt) for the template matching. This bit rate was almost the same as that of the direct compression (encoding the first differenced signal of original signal). (3) The decompressed templates could make it easy to classify the templates into the normal and abnormal beats; this could be executed without fully decompressing the ECG signal.},
keywords={},
doi={},
ISSN={},
month={December},}

Copy

TY - JOUR
TI - Data Compression of Ambulatory ECG by Using Multi-Template Matching and Residual Coding
T2 - IEICE TRANSACTIONS on Information
SP - 1419
EP - 1424
AU - Takanori UCHIYAMA
AU - Kenzo AKAZAWA
AU - Akira SASAMORI
PY - 1993
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E76-D
IS - 12
JA - IEICE TRANSACTIONS on Information
Y1 - December 1993
AB - This paper proposed a new algorithm of data compression for ambulatory ECG, where no distortion was included in the reconstructed signal, templates were constructed selectively from detected beats, and categorized ECG morphologies (templates) could be displayed in decoding the compressed data. This algorithm consisted of subtracting a best-fit template from the detected beat with an aid of multi-template matching, first differencing of the resulting residuals and modified Huffman coding. This algorithm was evaluated by applying it to ECG signals of the American Heart Association (AHA) data base in terms of bit rates. Following features were indicated. (1) Decompressed signal coincided completely with the original sampled ECG data. (2) Bit rate was approximately 800 bps at the appropriate threshold 50-60 units (1 unit2.4µVolt) for the template matching. This bit rate was almost the same as that of the direct compression (encoding the first differenced signal of original signal). (3) The decompressed templates could make it easy to classify the templates into the normal and abnormal beats; this could be executed without fully decompressing the ECG signal.
ER -