Detection of Covid-19 from Chest X-Rays using Improvised Convolutional Neural Network Model

Abstract

The COVID-19 pandemic, caused by the rapid global spread of the coronavirus disease 2019, has had a significant impact, affecting millions of individuals. In order to effectively combat the transmission of COVID-19, early identification is crucial. To address this, a proposed model called CovidDetector has been created, utilizing a Convolutional Neural Network (CNN) to automatically screen chest X-ray images for COVID-19. The CovidDetector model comprises three convolutional layers with increasing filter numbers, followed by max-pooling layers and two fully connected layers. This model has been trained and tested on a dataset containing three categories: Covid-19, Normal, and Viral. By analyzing chest X-ray images, it can determine whether Covid-19, Normal, or Viral patterns are present. The evaluation metrics, including F1 score: 96.7, precision, and recall, all yielded a high accuracy of 95% when compared to LSTM (Acc: 83%, F1-Score: 84.22), RNN (Acc: 87.19%, F1-Score: 89.05), MLP (Acc: 88.22%, F1-Score 92.37). Its potential application in a clinical setting can assist in early virus detection, ultimately leading to lives being saved. As AI technologies continue to advance and improve, the CovidDetector model exemplifies the effectiveness of deep learning techniques in the battle against COVID-19.