A Guide to Conducting a Successful Computer Vision Project

Four Duke AI Health Fellows, including myself, began development of the guide in late 2021. It was originally meant to be a ‘lessons learned repository’ for other AI Health fellows working with imaging projects, but over the course of our fellowship, it has grown into a 50+ page document that outlines the process of a computer vision project from start to finish, with dedicated sections for Duke-specific resources.

The seminar, hosted by Duke AI Health, highlights a few key points that the fellows wish we would have known before starting our projects. You can find a link to the guide and the seminar slides below. We want to emphasize that the guide is meant to be a living document, and we are continuing to add to and improve this resource.