Host Ben Everett sits down with Tufts University physician-scientist Dr. Michael Halassa to discuss algorithmic circuit psychiatry. This framework aims to modernize mental health care by mapping subjective experiences onto objective neural computations. By shifting focus to brain circuit mechanics, they explore a new paradigm for treating complex psychotic disorders. This conversation redefines psychiatry as a data-driven, precision-oriented field of medicine.

The episode moves beyond the "chemical imbalance" theory to examine the dynamics of excitation and inhibition. Dr. Halassa explains how large language models and machine learning provide new test beds for analyzing reasoning and belief updating, and that, by using "behavioral clamps" and task-based biomarkers, researchers can now operationalize delusions through the study of counterfactual decision-making. He also notes that causal circuit validation in animal models remains essential for identifying precise drug targets and improving clinical outcomes. The discussion finishes up by touching on emerging muscarinic therapies and the future of psychiatric training.

Episode Highlights:

00:00 – Why Algorithmic Circuit Psychiatry Could Modernize Mental Health Care

02:36 – From Physics to Psychiatry: Building a Scientist-Clinician Lens

05:52 – Decoding Brain Circuits With Computational Models and Modern Tools

10:37 – Returning to Inpatient Psychosis Care and Reframing Clinical Reality

14:47 – Moving Beyond “Chemical Imbalance” Thinking in Schizophrenia Treatment

20:43 – Fixing Computational Psychiatry Limits With Mechanistic, Circuit-Based Models

25:08 – Creating Task-Based Biomarkers to Measure Belief Updating and Reasoning

29:10 – Operationalizing Delusions Through Counterfactual and Decision-Making Tasks

33:59 – Translating Algorithms Into Drug Targets and Better Animal Research

37:54 – Using LLMs and Machine Learning to Test Psychiatric Mechanisms In Silico

44:57 – Redesigning Animal Models to Validate Causal Brain Circuit Algorithms

53:03 – Training the Next Generation for Precision Psychiatry

56:44 – Defining Clinical and Scientific Milestones for the Future of Mental Health Care

Key Takeaways:

"Psychiatry feels different from other fields. We don't have biomarkers to guide decision making."

"The brain functions in packets of information sent between areas. It's more complex than a single synapse."

"In psychiatry, you absolutely need a behavioral clamp. It's not just about resting state measurements."

"Machine learning was inspired by neuroscience. Now, it helps us understand altered thinking in machines."

"The burden is on us to train the next generation to tackle psychiatry's complexity."

"Talking to patients like equals is my default. We're all vulnerable to mental illness."

"Mental health is sidetracked by societal issues. We must agree we're all human beings."

Links:

Full transcript and show notes: https://www.psychiatrist.com/jcp/ep7-algorithmic-psychiatry-michael-halassa/

Journal of Clinical Psychiatry: psychiatrist.com/jcp/

“Developing algorithmic psychiatry via multi-level spanning computational models”:

https://pubmed.ncbi.nlm.nih.gov/40300598/

Dr. Halassa’s Substack: michaelhalassa.substack.com

The Halassa Lab: https://halassalab.tufts.edu/