Dr AI, cure my low back pain!
summary
A prototype uses a case-based reasoning case base of patient histories and exercise plans, then applies genetic-algorithm adaptation to recombine two retrieved plans into a new seven-exercise plan. Evaluation is leave-one-out with three medical experts rating quality and recommending plans; adapted plans appear among the top two almost as often as expert-created plans and score comparably on average, while unadapted reuse performs worse. This is a proof of concept, not a deployed system.
AI-written summary · about
chapters (4)
related talks
39mHuman Values in Clinical Algorithms, Old and New
17mAI-enhanced Systems Engineering for the Development of complex Medical Devices
25mAI for healthcare: Scaling Access and Quality of Care for Everyone
23mAI in the Life Science Sector
16mMultimodal AI: Transforming Patient-Centered Healthcare
28mAI in healthcare - Deep learning for precision psychiatry
17mAI Innovation in Health
07mHow AI is Enabling Healthcare's Biggest Blindspot