Enhanced ECG-Based Cardiac Disorder Detection Using Recurrent Neural Network-Enhanced Temporal Transformer and Fire Hawk Optimization

S. Dinakar Raj; Kavitha Veerappan; Jeya Bright Pankiraj; K. Sangeetha; Sumesh Joe Cherian; S. Rajeswari

doi:10.5935/jetia.v12i60.3697

Autores

S. Dinakar Raj Assistant Professor, Department of Electronics and Communication Engineering, Saveetha School of Engineering, SIMATS Engineering, Saveetha University, Chennai, Tamilnadu, India.
Kavitha Veerappan Research Scholar, Department of Electronics and Communication Engineering, Indian Institute of Information Technology, Tamil Nadu, India.
Jeya Bright Pankiraj Assistant Professor, Department of Electronics and Communication Engineering, Loyola institute of Technology & Science, Loyola Nagar, Thovalai, Kanyakumari District, Tamilnadu, India.
K. Sangeetha Assistant Professor, Department of Electronics and Communication Engineering, SNS College of Engineering, Coimbatore.
Sumesh Joe Cherian Lecturer in Electronics Engineering, Mahakavi Vennikulam Gopalakurup Memorial Government Polytechnic College, Vennikulam, Pathanamthitta, Kerala, India.
S. Rajeswari Assistant Professor, Department of Computer Science and Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram, Andhra Pradesh, India.

DOI:

https://doi.org/10.5935/jetia.v12i60.3697

Resumo

Quick treatment and management of heart disease depend on an accurate and early diagnosis. An Electrocardiogram (ECG) is a vital indicator tool used in medical practice to evaluate cardiovascular illness. In order to tackle the challenges of cardiac classification, this study presents an innovative method based on Deep Learning (DL) approach, namely Recurrent Neural Network Enhanced Temporal Transformer (RETT). The implemented research is initiated with an Adaptive Wavelet Filter based pre-processing that enhances the quality of an ECG signal. Subsequently, the Hilbert Huang Transform (HHT) based feature extraction is exploited for extracting necessary features, which are crucial for detecting abnormalities and diagnosing heart disorders. Then, RETT classification approach is utilized for detecting the cardiac affliction accurately. Finally, the performance of a RETT approach is enhanced with the aid of a Fire Hawk Optimization (FHO) algorithm. The developed approach is executed in python software that shows the performance of DL that is useful for evaluating cardiovascular disease with an accuracy of 99.70%. This research aids in the generation of precise and effective diagnostic tools for dealing with pressing cardiac health issues.