Peer Reviewed

D. K. Adrian Williams

Physician-scientist and MD-PhD in training. I find the most interesting science happening right now and use AI to make it make sense.

Episodes

  1. 26 May

    Beyond the Checklist: Foundation Models and the Evolution of Precision Psychiatry

    Psychiatry is entering a new era. For decades, the field has relied heavily on clinical observation, patient narratives, and symptom-based diagnostic categories. But emerging advances in AI, neuroimaging, EEG, and computational psychiatry are pushing us toward something more precise: a biologically grounded model of mental illness. In this episode, we explore how foundation models could reshape psychiatric diagnosis and treatment. We break down how large-scale AI systems, including transformer-based models, can be trained on brain and behavioral data to detect patterns that may not be visible to the human eye. Using techniques like self-supervised learning and masked autoencoding, these models can begin to learn the underlying “grammar” of neural signals from EEG, fMRI, and other clinical data streams. We discuss what this could mean for psychiatry residents and the future of clinical practice, including: * How AI may help predict antidepressant treatment response * How multimodal data could support biologically informed subtypes of depression * How foundation models may eventually assist with diagnosis, prognosis, and personalized treatment planning * How AI scribes and documentation tools could reduce administrative burden and resident burnout * Why bias, black-box models, and cognitive deskilling remain serious risks Most importantly, we ask what role the psychiatrist plays in an increasingly data-driven future. Rather than replacing clinical judgment, foundation models may serve as a kind of “perceptual prosthesis,” helping psychiatrists see patterns more clearly while preserving the human relationship at the center of care. This episode is for psychiatry residents, trainees, and clinicians interested in the future of precision psychiatry, AI in medicine, and the evolving relationship between biological data and clinical care.

About

Physician-scientist and MD-PhD in training. I find the most interesting science happening right now and use AI to make it make sense.