Classification of normal and His bundle branch block subjects in Chagas myocarditis using P-R segment features analysis

It has been already reported that a large number of patients suffering Chagasic myocarditis show left and/or right His bundle blocks (HBB) due to a deterioration of A-V conduction during the chronic stage of the disease. His-Purkinje activity can be observed in the P-R segment as high frequency activity at the end of the P wave (P), in the the averaged and filtered high resolution ECG signal. In this work, in the P-R segment, temporal and energetic features were measured as a way to determine the conditions of the His-Purkinje system. P activity was observed and delimited for normal and chagasic subjects on averaged and high pass filtered XYZ recordings. Time intervals, means and RMS values were calculated and used as input features of a classification and neural-net based decision map generating algorithm. Mean and RMS value of the signal in most of the subintervals where greater in normals. Energetic parameters where the most significant ones in terms of differentiation between normals and chagasic. The proposed analysis could lead to the development of a non-invasive and low cost methodology for the early detection of A-V conduction problems in Chagas disease.