Department of Brain Repair and Rehabilitation Seminar: Professor Parashkev Nachev

It is in the fundamental nature of problems in medicine--and neurology in particular--that complex models--AI--offer the only tolerable solutions. Yet despite major technological advances, real-world applications are yet to make any meaningful impact. Progress is impeded by unclarities about what AI is, what developing it involves, and what it could conceivably deliver, that on the one hand undermine sensible attempts to apply it, and on the other fail to inhibit misguided attempts doomed to failure. Here I sketch out a practical approach to the critical evaluation of translational ideas in healthcare AI, with special attention to fidelity, generalisability, interpretability, and equity.

Please confirm your attendance via Eventbrite

Please send enquiries to: a.gent@ucl.ac.uk