COVID19, Pneumonia and Normal Chest X-ray PA Dataset

Published: 23-03-2021| Version 2 | DOI: 10.17632/mxc6vb7svm.2
Contributors:
Zabirul Islam

Description

The dataset is organized into 3 folders (covid, pneumonia , normal) which contain chest X-ray posteroanterior (PA) images. X-ray samples of COVID-19 were retrieved from different sources for the unavailability of a large specific dataset. Firstly, a total 1401 samples of COVID-19 were collected using GitHub repository [1] , [2] , the Radiopaedia [3] , Italian Society of Radiology (SIRM) [4] , Figshare data repository websites [5] , [6] . Then, 912 augmented images were also collected from Mendeley instead of using data augmentation techniques explicitly [7] . Finally, 2313 samples of normal and pneumonia cases were obtained from Kaggle [8] , [9] . A total of 6939 samples were used in the experiment, where 2313 samples were used for each case. References: [1]https://github.com/agchung [2]https://arxiv.org/abs/2003.11597 [3]https://radiopaedia.org/?lang=gb [4]https://www.sirm.org/en/category/articles/covid-19-database/ [5]https://figshare.com/articles/dataset/COVID-19ChestX-RayImageRepository/12580328/2 [6]https://figshare.com/articles/COVID-19ImageRepository/12275009/1 [7]https://data.mendeley.com/datasets/2fxz4px6d8/4 [8]https://www.kaggle.com/paultimothymooney/chest-xray-pneumonia [9]https://www.kaggle.com/nih-chest-xrays/data

Files

Steps to reproduce

Steps to reproduce is given in Kaggle notebook (https://www.kaggle.com/amanullahasraf/covid-19-pretrained18).