michalkoziarski/DiagSet

Diagset_B and Diagset_C

Closed this issue · 3 comments

Hi, may i know if these datasets are available since i can't seems to find it in the website.

Yes, they should be available, due to the large size you can't download them as an archive (as .tgz files), but you can use the script provided at the website (https://ai-econsilio.diag.pl/download-script/; linked on the main page) to do it, which uses DiagSet-B.csv/DiagSet-C.csv files (also linked on the main page). Alternatively, you can manually browse the scans from https://ai-econsilio.diag.pl/annotated and download only the ones you are interested in (scan IDs match the content of .csv files).

ok can i know what does each of this label stands for is it:
BG: Background or Non-tissue region
T: Tumor
N: Normal or Non-tumor
A: Adipose tissue
R1, R2, R3, R4, R5: The Gleason graded

It's described in the paper: "scan background (BG), tissue background (T), normal, healthy tissue (N), acquisition artifact (A), or one of the 1-5 Gleason grades (R1-R5)".