2014-10-30

VISCERALdetection training set expanded

The training set for the VISCERAL lesion detection benchmark has been expanded. There are now 25 whole body CT scans and 25 whole body MRT2 scans with their corresponding lesion annotations in the training set. The test set for this benchmark will contain only these two imaging modalities. Participants can choose in which modality to participate or submit in both modalities.

The 7 volumes discarded from the previous set should no longer be considered in the official training set for this benchmark but they can still be accessed from the cloud.