Development of a new analytical method for the electrocardiogram using short-time First Fourier Transforms

Abstract

To detect the minute electric potential inside the QRS complex, a new frequency domain method was designed using short-time First Fourier Transforms (SFFT) and high-frequency sampling (oversampling). SFFT improved the frequency resolution by oversampling that was applied to this analysis. The electric potential data of 15,000 points received weighted, running average processing and was subtracted from the original waveform to reduce the low-frequency component. The data in a segment of 160ms, including QRS, was processed by frequency analysis with the SFFT computation routine. The ECG of healthy individuals was analyzed by this method and its usefulness evaluated. The processing waves of the X-axis, Y-axis, and Z-axis of a representative normal subject were formed into 3 groups of peak electric potential. SFFT enabled the detection of the structure inside the QRS complex without signal averaging, and is considered capable of evaluating the process of excitement inside the QRS complex in the various heart diseases.