ARTIFICIAL NEURAL NETWORKS FOR AUTOMATIC ECG ANALYSIS

The analysis of the ECG can benefit from the wide availability of comp uting technology as far as features and performances as well. This pap er presents some results achieved by carrying out the classification t asks of a possible equipment integrating the most common features of t he ECG analysis: arrhythmia, myocardial ischemia, chronic alterations. Several ANN architectures are implemented, tested, and compared with competing alternatives. Approach, structure, and learning algorithm of ANN were designed according to the features of each particular classi fication task, The trade-off between the time consuming training of AN N's and their performances is also explored. Data pre-and post-process ing efforts on the system performance were critically tested. These ef forts' crucial role on the reduction of the input space dimensions, on a more significant description of the input features, and on improvin g new or ambiguous event processing has been also documented. Finally the algorithm assessment was done on data coming from all the currentl y available ECG databases.