Sistema automático de detecção de anomalias em eletrocardiogramas baseado em lógica Fuzzy

Abstract

The interpretation of electrocardiogram (ECG) signals is challenging and prone to human errors, even with specialized knowledge. To minimize such errors, a Fuzzy system was developed using the Intel® Quartus® platform for FPGA development and the MATLAB software's toolbox, Fuzzy Logic Designer, as a reference for simulated values to verify the consistency of the system's values. Initially aimed at enhancing the ECG analysis, the system was constrained to detect heart diseases solely based on the duration and amplitude parameters of the ECG P-wave. With this constraint, the system successfully classified heart diseases as negative, suspicious, or positive. However, there were differences between the simulated part and the development platform, particularly in the defuzzified output, due to the nature of hardware description implementation. Despite these disparities, the approach demonstrated potential, highlighting the need for further research to expand the additional knowledge base for the Fuzzy system. By incorporating more membership functions and rule base containing specialized knowledge in the field of ECG interpretation, there is a potential for the system to improve its detection capabilities and provide significant relevance in future studies or interpretations of various heart conditions.