This white paper addresses the challenges of developing software products and services for healthcare organizations in the era of Artificial intelligence. Deep Learning technology and abundant medical data enable a wide spectrum of applications for the healthcare industry including some that are regulated as software medical devices. Creating real clinical value requires multidisciplinary collaboration and an openness to ongoing discovery. Medical AI is now emerging from academic conferences and into the clinics. This requires process maturity.
Effective AI applications must accurately solve scientific problems such as precise segmentation and classification of medical images but do so in such a way that is both useful and usable. Special considerations in medical AI include data management, blended open source and proprietary software, and integration with clinical workflows.
The paper is built upon CuraCloud’s extensive experience developing medical AI applications both in academic settings and in collaboration with clinical partners world-wide. It explains how to combine agile approaches within a collaborative, and multidisciplinary AI development process that implements design controls specified in FDA regulation. The paper describes a proven process for collaborative AI development so that the technology is human-centered and that it is fit for purpose.