Telemetric ECG evaluation using einthoven-leads

A test data set consisting of 1264 ECGs is compared and evaluated by means of cross correlation with 24584 validated ECGs. The calculations are first carried out for all 12 medical leads and subsequently only for the 3 leads according to Einthoven. The aim of the investigation is to estimate the changes in the sensitivity and specificity between the 12-channel- and the 3-channel evaluation. Basis of the evaluations are the computed probability distributions for the classes of main diagnoses "ventricular hypertrophy", "myocardial infarction", "conduction disturbance" and "ischaemia". The results are checked in an blind test using the CSE test data set