Identification using thoracic electromapping of patients with sustained postinfarct ventricular tachycardia. Preliminary results of a new method of analysis of ST-T isoarea maps

Recent studies have demonstrated that body surface maps (BSM) can be employed as non-invasive diagnostic tool for recognizing cardiac states at risk for repetitive ventricular arrhythmias in patients (pts) with old infarction. Our study reports preliminary results of a new method of statistical anal...

Full description

Saved in:
Bibliographic Details
Published inGiornale italiano di cardiologia Vol. 17; no. 9; p. 754
Main Authors Rolli, A, Bonatti, V, Maffei, M L, Maurizio, A R, Finardi, A, Francescon, P, Lazzeroni, E, Botti, G
Format Journal Article
LanguageItalian
Published Italy 01.09.1987
Subjects
Online AccessGet more information

Cover

Loading…
More Information
Summary:Recent studies have demonstrated that body surface maps (BSM) can be employed as non-invasive diagnostic tool for recognizing cardiac states at risk for repetitive ventricular arrhythmias in patients (pts) with old infarction. Our study reports preliminary results of a new method of statistical analysis of ST-T isoarea maps for identifying patients with post-infarction sustained ventricular tachycardia. 38 pts with previous myocardial infarction have been studied, 25 without and 13 with sustained ventricular tachycardia (VT). The two groups of pts did not differ significantly for age, site of infarction and ejection fraction. BSM have been recorded by means of an automated 35-channels instrument from 140 thoracic leads. For each lead ST-T deflection area has been calculated in microV.sec and taken as input variables for stepwise discriminant analysis which allowed identification of the integral values significantly discriminant (for F less than 0.15) between the two groups. Canonical analysis has been applied to identified values to obtain, by canonical coefficients, linear combination of the values for the highest correlation with the two groups of pts. To test the power of the method, the two groups of pts have been divided randomly in a learning set (17 pts without and 9 pts with VT) and a test set (8 pts without and 4 pts with VT).
ISSN:0046-5968