Is it possible to develop an "AI Pathologist" to pass the board-certified examination of the American Board of Pathology

( )?  To achieve this goal, we launch this medical visual question answering (VQA) challenge where deep learning models are to be developed for interpreting pathology images and answering questions about the image contents. 

The task is VQA on pathology images: given a pathology image and a question, the model needs to give the correct answer.  We have collected a dataset containing 4,998 images and 32,799 question-answer pairs. Half of these questions are open-ended (why, what, how, where, etc.) and the other half are “yes/no” questions.