You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I think this was not reported before; I noticed that cases 42 and 114 have a lot of holes in the annotations of the kidney tumor. Note that both cases have large kidney tumors, I did not see this problem on small kidney tumors. See image below of case 42:
The text was updated successfully, but these errors were encountered:
Thanks for catching this! Our annotation procedure made use of a threshold, and it looks like it was set slightly too high for these hypodense tumors. I'll re-check the remainder and make sure this isn't the case in the final training set release.
I think this was not reported before; I noticed that cases 42 and 114 have a lot of holes in the annotations of the kidney tumor. Note that both cases have large kidney tumors, I did not see this problem on small kidney tumors. See image below of case 42:
The text was updated successfully, but these errors were encountered: