Abstract: This paper investigates how to effectively mine contextual information among sequential images and jointly model them in medical imaging tasks. Different from state-of-the-art methods that ...