Ensuring generalizability and clinical utility in mental health care applications: Robust artificial intelligence‐based treatment predictions in diverse psychosis populations
We present a robust framework for training and assessing the clinical utility of prediction models in psychiatry. Our models generalize across different psychosis populations and show promising calibration and net benefit. However, performance disparities across demographic and treatment subgroups highlight the need for more diverse clinical samples to ensure equitable prediction.
Authors
- Oliver Howes (ORCID: https://orcid.org/0000-0002-2928-1972)
- Paris Alexandros Lalousis (ORCID: https://orcid.org/0000-0001-9642-3080)
- Fiona Coutts (ORCID: https://orcid.org/0009-0009-8951-1467)
- Jeffrey A. Lieberman (ORCID: https://orcid.org/0000-0002-1899-1776)
- Sergio Mena (ORCID: https://orcid.org/0000-0002-5102-2336)
- Rene Kahn
- Christoph U. Correll (ORCID: https://orcid.org/0000-0001-5720-7203)
- Alkomiet Hasan (ORCID: https://orcid.org/0000-0002-2662-9742)
- Nikolaos Koutsouleris
- W. Wolfgang Fleischhacker
- Esin Ucur
Institutions
- Feinstein Institute for Medical Research (US)
- Hofstra University (US)
- Innsbruck Medical University (AT)
- University of Augsburg (DE)
- King's College London (GB)
- Universität Innsbruck (AT)
- New York Psychoanalytic Society and Institute (US)
- German Centre for Cardiovascular Research (DE)
- KUKA (Germany) (DE)
- Max Planck Institute of Psychiatry (DE)
- Bezirkskrankenhaus Augsburg (DE)
- Columbia University (US)
- Charité - Universitätsmedizin Berlin (DE)
- Ludwig-Maximilians-Universität München (DE)
- New York College of Health Professions (US)
- Icahn School of Medicine at Mount Sinai (US)
Publication Details
- Journal
- Psychiatry and Clinical Neurosciences
- Published
- 2025-11-06
- DOI
- https://doi.org/10.1111/pcn.13914
- Primary Topic
- Machine Learning in Healthcare
- Type
- article
- Field-Weighted Citation Impact
- 0.00