Premature ventricular contraction detection system from the 12 leads ECG by artificial intelligence

Background Electrocardiographs (ECG) are obtained by a digital signal. However, they are still printed out of paper to read by physicians. Digitizing the analog ECG from the paper to digital signal make us much easier to access of bid data pool form the daily clinical practice and previous resources. Objective The goal of this study is to digitize paper ECG to detect premature ventricular contraction (PVC). Methods This system consists of 2 steps; digitization and PVC detection. Results First, for digitization, ECG are filtered by the specific cut-off value of red, green and blue, then the filtered ECG image is changed into gray scale. In order to extract ECG signal, the algorithm fine the only one of the biggest white body throughout the X-axis. The X and Y axis is matched with distance and amplitude, depending on dots per inch (DPI). Second, to detect PVC, ECG signal is filtered to eliminate baseline wandering. The characteristics of PVC is higher amplitude and longer duration than normal sinus rhythm, we set two criteria to detect the PVC: 1.5 times the duration, 1.2 points out of the amplitude. For the synchronization of timing, lead II rhythm strip was used by PVC detection and then the rest of 12-lead ECG is matched based on lead II synchronization (Figure 1). We applied this algorithm to the 300 real patient's ECG. 290 of 300 (96.7%) ECG are successfully digitized signal and PVC detection. Conclusion We successfully developed the algorithm analog ECG signal into digital signal to detect PVC. In the future, this method helps to gather big data from ECG papers to develop a new algorithm to localization of PVC. Funding Acknowledgement Type of funding source: Foundation. Main funding source(s): Korean Heart Rhythm Society 2019 Research Fund

The baseline wandering noise removal from ECG signal using forward–backward Riemann Liouville fractional integral-based empirical wavelet transform approach

The electrocardiogram (ECG) non-invasively monitors the electrical activities of the heart to diagnose the heart-related diseases. The baseline wandering noise affects the diagnosis of the heart diseases. In this paper, the baseline wandering noise removal is done using forward–backward Riemann Liouville (RL) fractional integral-based empirical wavelet transform (EWT) approach. In the designed methodology, firstly, the noisy ECG signal is decomposed into various modes from low to high frequencies. Then, the first mode is processed to remove the baseline wandering noise. The processed EWT mode is filtered by the fractional RL filter used in the forward direction and then in the backward direction for removing the baseline wandering noise from the ECG signal. After that, the processed and the unprocessed modes are used to reconstruct the denoised ECG signal. The clean ECG signal record is taken from MIT-BIH ECG-ID database, and the baseline wandering noise record is taken from the MIT-BIH noise stress test database. The performance of the proposed approach is validated in terms of the output signal-to-noise ratio (SNR[Formula: see text]). The comparative study has also been done between the proposed denoising approach and the existing state-of-the-art denoising algorithms. The experimental result proves the supremacy of our proposed denoising approach.