Abstract
Detection of cardiac abnormalities through annotation and modeling based on 3-Lead ECG data has been reported in literature very extensively. However, we realize that analysis based on 12-Lead, dimension resolved, ECG data is vital for accurate detection of critical cardiac events such as Myocardial Infarction (MI), Ischemia, Bundle Branch Blocks, Pericarditis etc. and understanding of underlying conditions leading to it. In this work we present an approach to first annotate 12-Lead ECG data and further analyze them with the objective of classifying them broadly into normal or abnormal (MI) conditions. The algorithm that we have developed enables the representation of all 12-Lead ECG data in form of a critical feature set which are then subjected to clinically established rule set to detect cardiac abnormality. With a preliminary implementation of this methodology, the classification results seem encouraging leaving immense scope for further enhancements towards developing a robust Cardiac Decision Support System.