I know that the leaderboards are not up yet, but I'd like be grateful if you could clarify the position about the metadata. I still need to complete one run of all images on the training data. So far, before I put in the parallelization, it took too long. I am hoping it will still be possible to do this with the leaderboards - as well as to take the challenge. I'm concerned because the metadata on the training system has, in the file /metadata/images_crosswalk.tsv: subjectId examIndex imageIndex view laterality filename cancer On the express lane system /metadata/images_crosswalk.tsv only has: subjectId examIndex imageIndex view laterality filename So I'm not able to do training there, only test that the training code is working. Will /metadata/images_crosswalk.tsv on the leaderboard system have the field 'cancer' for the images in the /trainingData directory? It's really crucial for me that it does! Or, equivalently, that /metadata/exams_metadata.tsv has the fields 'cancerL' and 'cancerR', which is equivalent. I do understand that the test images, for the challenges, will, of course, not have this information! That's the point, to find it out. Is there any information yet, as to how we need to present the data for the challenges? Presumably, we'll have images, with names such as abd_1234.dcm, and we'll need to produce our prediction as: Sub-Challenge 1: 0 = no cancer anticipated 1 = cancer anticipated Sub-Challenge 2: 0.00 - 1.00 = probability cancer will develop Will we need to produce a file with lines like: abd_1234.dcm 1 or abd_1234.dcm 0.60 For the two challenges? If so, what will it be called, where must it be put, and what is the format of it?