The Effect Of Covid-19 On Cardiac Function: A Machine Learning Approach.

Authors: Dr.S.Vallinayagi, Mrs.M. Gandhimathi

Abstract: The global virus outbreak in December 2019 led to the COVID-19 pandemic, which is arguably the biggest public health emergency in history. COVID-19, which was initially thought to be only a respiratory disease, is actually a blood-related illness that affects the respiratory system. In this study, we sought to investigate the impact of COVID-19 on cardiac function using a machine learning method to analyze electrocardiography (ECG) signals. Given its effects on haematological factors, how does COVID-19 affect cardiac function? Can the clinical diagnosis of COVID-19 be supported by automatically analyzing electrocardiography? We made use of a publicly accessible database of ECG signals captured in emergency care settings and displayed as pictures of printed recordings. Signals linked to myocardial infarction, irregular heartbeats, a history of myocardial infarction, COVID-19, and healthy heartbeats are all included in this database. We suggested a system to help with COVID-19 diagnosis based on hybrid deep learning architectures that use Random Forests for classification and pre-trained convolutional neural networks for feature extraction. Looked at the architectures of LeNet, ResNet, and VGG16. The VGG16 and Random Forest model with 100 trees, which used attribute selection via particle swarm optimization, produced the best results. There are now 773 attributes instead of 4096.