Open jaanli opened 5 months ago
Thanks for your interest. I have added information on annotation format to the readme. The DICOM files would have to be extracted from their radiology department's system first and then converted into a format like png prior to doing any annotation with Prodigy. The extraction process would depend on their department's IT setup. Pydicom can be used to open DICOM files in python to facilitate image extraction into a format that Prodigy can annotate.
Hi @jrzech ! It's been an honor to collaborate with your team @onefact - we are in touch with residents in Malawi who are interested in being trained using your algorithm for a follow-up study (happy to collaborate with you and your team in case there is continued interest).
However, we couldn't find the code to annotate images from their local database in the radiology department I chatted with using Label Studio or Prodigy.
Would this be available please? It would save us a lot of time in replicating this incredible work (and the follow-up work from last month: https://link.springer.com/article/10.1007/s00256-024-04698-0)