An Overview of Deep Learning Techniques on Chest X-Ray and CT Scan Identification of COVID-19

Table 4

Available data sources about COVID-19 radiology images for both chest X-ray and CT images.

No.	Sources	Data type	No. of images	Image type	Types of images	Links
1	J. P. Cohen’s GitHub	Viral pneumonia (SARS, varicella, influenza) and COVID-19, bacterial pneumonia (Streptococcus spp., Klebsiella spp., Escherichia coli, Mycoplasma spp., Legionella spp., unknown, Chlamydophilla spp.) and COVID-19, fungal (Pneumocystis spp., lipoid) and COVID-19	Raw images: 910, annotated images: 210	jpg and png	CXR	https://github.com/ieee8023/covid-chestxray-dataset
2	European Society of Radiology	Total cases or images unknown	N/A	pdf	CXR and CT	https://www.eurorad.org/advanced-search?search=COVID
4	Kaggle	Posterior-anterior (PA), anterior-posterior (AP) lateral for X-rays and axial or coronal for CT scans	Normal images: 1,576, pneumonia ARDS images: 2, viral pneumonia images: 1,493, COVID-19 images: 58, SARS images: 4, bacterial pneumonia images: 2,772, bacterial Streptococcus images: 5	png, jpg, jpeg, and others	CXR and CT	https://www.kaggle.com/bachrr/covid-chest-xray
5	UCSD-AI4H	Total: 349 images from 216 patients	COVID-19 images: 349, non-COVID-19 images: 397	jpg and png	CT	https://github.com/UCSD-AI4H/COVID-CT
6	MedSeg	Images were segmented by a radiologist using 3 labels: ground-glass (), consolidation (=2), and pleural effusion (=3).	Image volumes—9 volumes, a total of >800 slices, COVID-19 masks 350 annotated slices. Lung annotated slices	jpg	CT	http://medicalsegmentation.com/covid19/
7	COVID-19 Radiography Database		COVID-19 images: 219, normal images: 1,341, viral pneumonia images: 1,345	png	CXR	https://www.kaggle.com/tawsifurrahman/covid19-radiographydatabase