Discovering acute myocardial infarction markers in body surface mapping

Author

Lopez, J.A. ; Nugent, C.D. ; Black, N D

fYear

2003

fDate

24-26 April 2003

Firstpage

184

Lastpage

186

Abstract

Since acute myocardial infarction (AMI) is localised on the torso, and due to the high number of electrocardiographic leads involved in body surface mapping BSM, it is desirable to find a reduced set of BSM electrocardiographic leads and features which could be potential markers of MI. We applied Fisher criterion for the discrimination, in order to suggest the most discriminative areas of the torso along with their associated ECG features. We derive these results from a BSM data base consisting of 170 non AMI individuals and 67 AMI individuals. We then indicate with statistical continuous map representations of lead relevance and feature relevance scores that, for AMI the most discriminative information is not localised on the precordial area but on the upper left chest and lower mid-axillary right area.

Keywords

electrocardiography; information services; medical information systems; medical signal processing; statistical analysis; statistical databases; Fisher criterion; acute myocardial infarction individuals; acute myocardial infarction markers; body surface mapping data base; body surface mapping electrocardiographic leads; discriminative information; electrocardiographic features; feature relevance scores; lead relevance; lower mid-axillary right area; precordial area; statistical continuous map representations; torso; upper left chest; Ambient intelligence; Cardiography; Data mining; Electrocardiography; Feature extraction; Informatics; Myocardium; Simultaneous localization and mapping; Time measurement; Torso;

fLanguage

English

Publisher

ieee

Conference_Titel

Information Technology Applications in Biomedicine, 2003. 4th International IEEE EMBS Special Topic Conference on

Print_ISBN

0-7803-7667-6

Type

conf

DOI

10.1109/ITAB.2003.1222506

Filename

1222506