Open AccessCOVID-19 Data Analysis using Chest X-Ray
Author(s) -
Ishtiaque Ahmed,
AUTHOR_ID,
Manan Darda,
Neha Tikyani,
Rachit Agrawal,
Manjusha Joshi,
AUTHOR_ID,
AUTHOR_ID,
AUTHOR_ID,
AUTHOR_ID
Publication year - 2021
Publication title -
international journal of advanced medical sciences and technology (ijamst)
Language(s) - English
Resource type - Journals
ISSN - 2582-7596
DOI - 10.54105/ijamst.c3018.081421
Subject(s) - covid-19 , radiography , stride , pandemic , outbreak , predictive value , medicine , radiology , livelihood , computer science , disease , pathology , physical medicine and rehabilitation , geography , infectious disease (medical specialty) , archaeology , agriculture
The COVID-19 pandemic has caused large-scale outbreaks in more than 150 countries worldwide, causing massive damage to the livelihood of many people. The capacity to identify contaminated patients early and get unique treatment is quite possibly the primary stride in the battle against COVID-19. One of the quickest ways to diagnose patients is to use radiography and radiology images to detect the disease. Early studies have shown that chest X-rays of patients infected with COVID-19 have unique abnormalities. To identify COVID-19 patients from chest X-ray images, we used various deep learning models based on previous studies. We first compiled a data set of 2,815 chest radiographs from public sources. The model produces reliable and stable results with an accuracy of 91.6%, a Positive Predictive Value of 80%, a Negative Predictive Value of 100%, specificity of 87.50%, and Sensitivity of 100%. It is observed that the CNN-based architecture can diagnose COVID-19 disease. The parameters’ outcomes can be further improved by increasing the dataset size and by developing the CNN-based architecture for training the model.