Application of Graph Fourier Transform in the Diagnosis of Left Bundle Branch Block from Electrocardiographic Signals

In recent years, there has been a growing interest in left bundle branch block (LBBB) due to its role as an indicator for evaluating the benefits of cardiac resynchronization therapy (CRT). Considering that LBBB arises from spatial electrical heterogeneity, we hypothesized that graph theory could effectively represent this variability. In our study, we used Fourier graph transform to analyze and filter frequencies in the transformed space, aiming to classify associated signals. Our results showed an accuracy of 97.87%, indicating a promising application of graphs in diagnosing LBBB. These findings are encouraging and point towards further exploration of the potential of graph theory to enhance LBBB diagnosis.