Creating an Exocerebellum

Abstract

As neuroprostheses have advanced through technological innovation, new surgical capabilities, and enhanced clinical care and training paradigms, their degree of integration with their human users has changed dramatically. In particular, tighter coupling of prostheses with users combined with rapid step-changes in computing capabilities available to neuroprosthesis designers have made it now possible for prosthetic systems to adapt during both training and deployment. Could we now imagine creating what might be considered an external, machine cerebellum to support and augment human action, perception, and cognition during sensorimotor decision making? And perhaps someday also imagine creating an exocortex? This talk will explore how modern trends in artificial intelligence and machine learning can support the creation of an exocerebellum, drawing on ideas within the Alberta Plan for AI Research and from concrete examples within the past 10 years of work by our lab on developing upper-limb prosthetic interfaces (specifically techniques and viewpoints from the field of reinforcement learning). Presented as ‘Creating an Exocerebellum’, CAPNet CSBBCS Satellite Symposium, Edmonton, Alberta, Jun. 25, 2024.