2 files


posted on 29.06.2021, 08:41 by Sijing Feng, Damian Azzollini, Ji Soo Kim, Cheng Kai Jin, Eve Kim, Simon Gordon, Jason Yeoh, Min A Han, Andrew Lee, Aakash Patel, Martin Urschler, Amy Fong, Cameron Simmers, Gregory Tarr, Stuart Barnard, Ben Wilson
19,237 anonymized adult chest x-ray datasets in 1024 x 1024 pixel DICOM format with corresponding anonymized free-text reports from Dunedin Hospital, New Zealand between 2010 - 2020. Images were manually annotated by RANZCR radiology trainee and radiologists with respect to pneumothorax, acute rib fracture, and chest tubes. Segmentation annotations were converted to run-length-coded (RLE) format in csv files. In the provided metadata, image filenames contain patient index (enabling analysis requiring patient grouping of images), as well as anonymized date of acquisition information where the temporal relationship between images is preserved.

To access the data, an ethics training process is required and is divided into 2 steps:
1. An online ethics course at https://globalhealthtrainingcentre.tghn.org/ethics-and-best-practices-sharing-individual-level-data-clinical-and-public-health-research/. You will need to register an account to be able to take the free online ethics course. Once you finished the course quiz, please send the course certificate to Sijing.Feng@southerndhb.govt.nz

2. Signing the Data Use Agreement. It can be accessed at Data_Use_Agreement_-_Hospital_Signed.pdf (trello-attachments.s3.amazonaws.com). Once you signed the Data Use Agreement, please also send the signed copy to Sijing.Feng@southerndhb.govt.nz

After you have successfully completed both of above steps, you will receive a link to download the dataset.


RANZCR Research Grant 2020



University of Auckland

Usage metrics

Computer Science