Supervised and unsupervised learning for diagnostic ECG classification

Author

Silipo, Rosaria ; Bortolan, Giovanni ; Marchesi, Carlo

Author_Institution

Dept. of Syst. & Inf., Florence Univ., Italy

Volume

3

fYear

1996

fDate

31 Oct-3 Nov 1996

Firstpage

931

Abstract

A hybrid system with RBF pre-processing, a system with supervised learning, is compared with some Kohonen self-organizing maps in a subtle ECG classification task. Based on ECG measures, they are supposed to detect normal condition, presence of infarction and of hypertrophy, and at the same time to sub-classify those pathologies. During the evaluation process the hybrid system produces better results. In terms of average sensitivity and specificity (83% vs. 62% of sensitivity and 84% vs. 92% of specificity), but Kohonen maps allow a detailed description of the similarities among input data. An integration of the two techniques should improve the final results

Keywords

electrocardiography; learning (artificial intelligence); medical signal processing; neural nets; self-organising feature maps; unsupervised learning; ECG measures; Kohonen self-organizing maps; RBF pre-processing; diagnostic ECG classification; electrodiagnostics; hybrid system; hypertrophy; infarction; input data similarities; pathologies subclassification; sensitivity; specificity; subtle ECG classification task; Electrocardiography; Engineering in Medicine and Biology Society; Neural networks; Pathology; Self organizing feature maps; Sensitivity and specificity; Signal analysis; Supervised learning; Time measurement; Unsupervised learning;

fLanguage

English

Publisher

ieee

Conference_Titel

Engineering in Medicine and Biology Society, 1996. Bridging Disciplines for Biomedicine. Proceedings of the 18th Annual International Conference of the IEEE

Conference_Location

Amsterdam

Print_ISBN

0-7803-3811-1

Type

conf

DOI

10.1109/IEMBS.1996.652647

Filename

652647