1 paper
Mingjian Li, Mingyuan Meng, Michael Fulham +3
Medical image representations can be learned through medical vision-language contrastive learning (mVLCL) where medical imaging reports are used as weak supervision through image-t…