The primary diagnostic tool for monitoring problems with the heart or related diseases is the electrocardiogram (ECG). The rising mortality rate caused by cardiac abnormalities leads to the development of more advanced techniques to detect heart anomalies are is of paramount importance. An electrocardiogram (ECG) is used for multiple diagnostic purposes since it produces continuous tracings of the electrophysiological activity generated from the heart. This paper developed a Bidirectional recurrent chimp search (Bi-RCS) method for diagnosing abnormalities by collecting ECG signals via various phases such as the data acquisition phase, signal processing phase, feature extraction phase, abnormality detection phase, real-time FPGA integration phase, monitoring and reporting phase as well as feedback phase. Signal pre-processing removes contaminants like noise using the Adaptive filtering methodology which severely limits the utility of the recorded ECG for better clinical evaluation. Bi-RNN is utilized for extracting related features and detecting abnormalities and the Chimp search optimization is employed to tune the parameters of Bi-RNN. After detecting the abnormalities, the reports of patients are monitored and recorded to provide an alert signal to healthcare providers. Finally, the potency of the methodology is analyzed with the ECG signal dataset and metrics like precision, F1-score accuracy, sensitivity, and specificity values. The obtained outputs equalized with the existing methods like KNN, EBT, and SVM. The analysis showed convincing performance with an accuracy of 99% and less error rate of 0.05.
Read full abstract