PRIMITIVE CODING OF STRUCTURAL ECG FEATURES

Authors
KOSKI A
Citation
A. Koski, PRIMITIVE CODING OF STRUCTURAL ECG FEATURES, Pattern recognition letters, 17(11), 1996, pp. 1215-1222
Citations number
26
Categorie Soggetti
Computer Sciences, Special Topics","Computer Science Artificial Intelligence
Journal title
Pattern recognition letters → ACNP
ISSN journal
01678655
Volume
17
Issue
11
Year of publication
1996
Pages
1215 - 1222
Database
ISI
SICI code
0167-8655(1996)17:11<1215:PCOSEF>2.0.ZU;2-6
Abstract
In this paper, a hybrid of self-organizing feature map and learning ve ctor quantization and a rule base classification for the extraction of structural primitives of ECG signals is presented. Structural primiti ves are the elementary building blocks of an ECG signal and they are u sed in the syntactic recognition and analysis of the signal. The analy sis based on the coding by the codebook vectors and slope rules was as successful as the analysis of the manually labelled test set.