Computer Vision News - July 2020

2 Summary We Tried For You 14 The story of MONAI is a fascinating one: its creation and development are directly generated from the intention to bring together the best expertise on the market, in a full collaboration between industries and universities, and, more generally, everybody truly making a contribution in the field of deep learning for medical purposes. The topic of this month is an extremely recent open source framework for deep learning in healthcare imaging. It’s called MONAI and it is still under active development, first released on the 21st of April (v0.1), and soon to be followed by a v0.2 with new capabilities. M O N A I by Marica Muffoletto It was one of the main leaders of this project, Prerna Dogra , senior product manager at NVIDIA , to give us an insight into the first steps of MONAI: “It truly started with conversations between Nvidia & King's College London (KCL) - we like many others had invested in our respective toolkits Nvidia Clara & NiftyNet and overall for the Medical Imaging community therewere thesedisjointed initiatives capturing AI development researcher workflow, however there was no true point of convergence. This is what motivated Nvidia & KCL to join forces and the idea of MONAI was pitched at MICCAI 2019 in a lunch briefing. Our endeavour with MONAI is to converge community thought leaders and researchers around a common SW foundation and really come together to determine best practices for AI development in Healthcare Imaging.” Prerna Dogra