Clinical findings in the dataset

At the NeurIPS workshop a few examples were given of incorrectly reconstructed anomalies (subchondral osteophytes or meniscal tears). Is this information available somewhere? Were this image acquired clinically or are these of volunteers? Could we expect clinical findings in the Brain MRIs?

As the goal should be to obtain images of diagnostic quality, it would be useful to have this information to be able to assess their diagnostic properties.

This information was shown after correspondence with the NYU hospital who were collaborators on the study. I believe the hospital has elected not to share clinical information of this nature on a large scale publicly for a variety of reasons.

The paper here describes the data set in some further detail, including the fact that it came from clinical studies:

1 Like