Boundary estimation of cardiac events S1 and S2 based on Hilbert transform and adaptive thresholding approach

A computerized cardiac disorders classification system needs a proper boundary estimated cardiac cycle of recorded heart sound signals. In the proposed algorithm the boundaries of two primary heart sounds, S1 and S2 events are estimated using Hilbert transform method and a statistical approach. This method uses an adaptive threshold value which is calculated from the first- and second-order moments of the heart sound or Phonocardiogram signal (PCG) envelope. Here, Hilbert transform is used for extracting the envelope of the PCG signal and a threshold is applied to detect the boundary regions of the primary events, S1 and S2, of the same signal. The performance of the algorithm is evaluated for normal and five commonly occurring pathological cases. The proposed method is computationally fast and obtained an accuracy of 97.24 percent.