Session S24.4

Diagnosis of Bundle Branch Block by Analyzing Body Surface Potential Maps

V Donis*, MS Guillem, AM Climent, F Castells,
FJ Chorro, J Millet

Universitat Politècnica de València
Valencia, Spain

Bundle branch block (BBB) is a heart disease produced when one of the branches or fascicles of the His bundle cannot transmit the electrical impulses that cause ventricular contraction. BBB is diagnosed by the analysis of the ECG morphology and the duration of the QRS complex. As Body Surface Potential Mapping provides more information than 12-lead ECG, this study aims to determine whether studying mapping information diagnosis of BBB can be improved. A commercial BSPM system of 64 leads was used to obtain the recordings. Electrodes were distributed non-uniformly upon the chest and back, specifically 48 were placed on the chest and 16 on the back.
Templates of the 64 leads of each subject were obtained by means of a template-matching averaging technique. Representative templates of each group of more than 5 subjects with the same diagnosis were obtained. Specifically two groups were considered: Left BBB and healthy subjects. Templates of each lead of all patients belonging to the same group were averaged.
A semi-automatic method was used to detect the onset and the offset of the QRS complex of each lead template. Then, the averaged onset and offset of the 64 leads of each subject and each group were computed. Afterwards, fourteen body surface potential maps were calculated equally spaced along the QRS complex. They were elaborated by using the templates of each individual and the averaged templates of each group, and interpolating by cubic splines the potential on the rest of the surface.
Map comparison was performed by means of a 2D correlation. For each group and individual, the two maps computed at the beginning of the QRS complex and the two computed at the end were not considered for comparison because of their low signal level. The 2D correlation was computed between the representative maps of each group and the maps of subjects from the group, and from the other groups.
Comparison results showed that only when comparing the representative maps of one group with the maps of subjects belonging to the group, the correlation index was high (>0.7). Specificity was 100% for both the LBBB and healthy group and sensitivity was 100% for LBBB patients and 89% for the healthy group.
Our study concluded that BSPM systems are useful to have more information about the propagation pattern in BBB patients. Moreover, obtaining representative maps of a concrete type of blockade or of healthy subjects makes it possible to classify subjects in their appropriate group. As a consequence, it provides an improvement in the differential diagnosis of BBB patients and this technique may be helpful for diagnosing other diseases.

(Abstract Control Number: 78)