Background The 12-lead ECG, together with patient history and clinical
findings, remains the most important method for early diagnosis of ac
ute myocardial infarction. Automated interpretation of ECG is widely u
sed as decision support for less experienced physicians. Recent report
s have demonstrated that artificial neural networks can be used to imp
rove selected aspects of conventional rule-based interpretation progra
ms. The purpose of this study was to detect acute myocardial infarctio
n in the 12-lead ECG with artificial neural networks. Methods and Resu
lts A total of 1120 ECGs from patients with acute myocardial infarctio
n and 10 452 control ECGs, recorded at an emergency department with co
mputerized ECGs, were studied. Artificial neural networks were trained
to detect acute myocardial infarction by use of measurements from the
12 ST-T segments of each EGG, together with the correct diagnosis. Af
ter this training process, the performance of the neural networks was
compared with that of a widely used ECG interpretation program and the
classification of an experienced cardiologist. The neural networks sh
owed higher sensitivities and discriminant power than both the interpr
etation program and cardiologist. The sensitivity of the neural networ
ks was 15.5% (95% confidence interval [CI], 12.4 to 18.6) higher than
that of the interpretation program compared at a specificity of 95.4%
(P<.00001) and 10.5% (95% CI, 7.2 to 13.6) higher than the cardiologis
t at a specificity of 86.3% (P<.00001). Conclusions Artificial neural
networks can be used to improve automated ECG interpretation for acute
myocardial infarction. The networks may be useful as decision support
even for the experienced ECG readers.