Automated real-time electrocardiogram diagnosis based on the modified Pan-Tompkins algorithm for long-term monitoring systems

The object of this study is the process of automated analysis of the electrocardiographic signal (ECS) during long-term monitoring in real time, carried out by mobile wireless systems. The study considers the problem related to the insufficient accuracy of automated diagnostics during long-term monitoring of the electrocardiogram (ECG) under conditions of limited computing resources and the presence of noise. A modified Pan-Tompkins algorithm for determining the boundaries of the QRS system has been developed. Based on this algorithm, the PCard software module for the hardware and software system was implemented, enabling high-quality automated diagnostics both under the standard mode and during long-term ECG monitoring in 12 leads in real time. The PCard software module allows for ECG registration, digital filtering, measurement and calculation of electrocardiographic parameters, automatic determination of diagnostic criteria and diagnostic conclusions, formation of a general diagnostic conclusion of ECG, as well as medical processing of ECG. The high quality of the diagnostic analysis was confirmed by the obtained accuracy rates of the algorithm for determining normal complexes – 99.99%, for determining ventricular complexes – 99.90%, for determining various pathologies – 98.43%. The ECG processing time was about 4.7 seconds for a 40-minute record. The proposed method for determining the boundaries of QRS complexes is based on the finite difference method, which distinguishes it from common methodologies using spectral analysis, wavelet transforms, or Fourier transforms. This methodology simplifies determining the parameters of the basic ECG elements and significantly reduces the amount of calculations, which generally increases the processing time and reduces the required volume of system resources

Text 125182.pdf - Published Version Available under License Creative Commons Attribution. Download (1MB) Official URL or Download Paper: https://journals.uran.ua/eejet/article/view/336172

Actions (login required)

View Item