An Automated Deep Learning System for Accurate Identification of Pneumonia and Covid-19 from Chest X-Ray Images

Abstract

Deep learning (DL), which has been developing rapidly over the past few years and is now very helpful in an array of sectors thanks to the increasing data readily accessible, The primary purpose of DL technology is to make decisions in a more timely, dependable, and accurate manner. As a result of this capacity, DL has found applications in the medical field, especially with the intention of concentrating on different kinds of medical imagery or visuals that are relevant to the patient's health. The diagnostic procedures used in these areas are reliant on the gathering and analyzing of a substantial quantity of medical images. This research presents a DL algorithm for identifying Pneumonia and Covid-19 utilizing Chest X-Ray images. The model is built on something called a "Convolutional Neural Network (CNN)". The outcome of this analysis enables the radiologist to extract insights and make decisions that assist them decide the patient's accurate diagnosis. This model is useful in two different respects. In the initial place, it is necessary to determine if a chest x-ray displays any alterations in relation to COVID 19 and infection. The second stage is to categorise the results using images of a computed tomography x-ray. The VGG16 model performed better than Inception V3 (IV3) model using multiple performance metrics. A web application to help radiologists gain perspectives and make judgements to perform accurate medical prognosis was also developed.