I agree, would be fantastic to have equivalents of ImageNet for different types of medical images - high quality, pref labeled anonymised data sets where the d

I agree, would be fantastic to have equivalents of ImageNet for different types of medical images - high quality, pref labeled anonymised data sets where the data set size is sufficiently large to train high quality networks.