Context

Pediatric chest X-rays are harder to properly acquire and standardize when compared to adults, as for a children in a dark room, with people watching them from a glass window with strange machinery doesn't make for a comfortable experience.
At the same time, children present a different physiology that is important to be captured in X-ray classification algorithms, as most datasets tend to focus on adults only.

Content

5,856 Chest X-rays labelled as either pneumonia or normal.

Acknowledgements

All credits are due to the authors of the dataset:

Kermany D, Goldbaum M, Cai W et al. Identifying Medical Diagnoses and Treatable Diseases by Image-Based Deep Learning. Cell. 2018; 172(5):1122-1131. doi:10.1016/j.cell.2018.02.010

How to cite this dataset:

Kermany, Daniel; Zhang, Kang; Goldbaum, Michael (2018), “Labeled Optical Coherence Tomography (OCT) and Chest X-Ray Images for Classification”, Mendeley Data, v2
http://dx.doi.org/10.17632/rscbjbr9sj.2