The 15 full papers included in this volume were carefully reviewed and selected from 22 submissions. The iMIMIC papers focus on introducing the challenges and opportunities related to the topic of interpretability of machine learning systems in the context of medical imaging and computer assisted intervention.