ResSwishNet-50 Model for Detection of COVID-19 using ECG-based Images
Tanees Riaz, Department of Software Engineering, University of Engineering and Technology, Taxila, Pakistan.
Ali Javed, Department of Software Engineering, University of Engineering and Technology, Taxila, Pakistan.
Corresponding Author:
Tanees Riaz (tanees.riaz@students.uettaxila.edu.pk)
Abstract:
This study proposes a computer-aided diagnostics method for detecting Coronavirus Diseases 2019 using electrocardiogram (ECG) images. The method utilizes numerical features extracted from ECG lead signals and used to generate a mapped image for the classification of COVID-19-positive and COVID-19-negative cases. A novel ResSwishNet-50 deep learning model with swish activation is introduced to address the dead neuron problem which improves the accuracy of the proposed method. The method achieves the accuracy of 99.0% in binary classification, demonstrating its effectiveness in detecting COVID-19. This approach is innovative in that it utilizes ECG images, a modality that has been largely overlooked in COVID-19 detection research. The method has the potential to assist in the early detection and diagnosis of COVID-19. However, further evaluation of the method on larger datasets and in real-world scenarios is necessary to validate its effectiveness.
Keywords:
Pixel-Indexing Method; Deep Learning; Hexaxial Reference System; ResSwishNet-50