Detecting Tardive Dyskinesia Using Video-Based Artificial Intelligence
<b>Conclusion:</b> Our algorithm reliably detected suspected TD, reaching higher sensitivity and specificity than trained raters using the standard assessment. The algorithm can be used to monitor patients taking antipsychotic medications, allowing scarce resources to assess identified patients for a conclusive diagnosis by psychiatrists.
Authors
- Owen S. Muir (ORCID: https://orcid.org/0000-0002-4003-338X)
- Bradley Grimm (ORCID: https://orcid.org/0000-0003-2896-5974)
- Joseph H. Friedman (ORCID: https://orcid.org/0000-0003-2534-4160)
- Joel W. Hughes (ORCID: https://orcid.org/0000-0002-6476-6588)
- Rakesh Ranjan (ORCID: https://orcid.org/0000-0003-0963-4512)
- Loren Larsen (ORCID: https://orcid.org/0000-0001-5818-6651)
- Anthony A. Sterns (ORCID: https://orcid.org/0000-0001-8609-9619)
- Fred Ma
- Carlene MacMillan
- Bretton H. Talbot
Institutions
- Oracle (United States) (US)
- Kent State University (US)
- Scripps Institution of Oceanography (US)
- Providence College (US)
- Butler Hospital (US)
- Scripps (United States) (US)
- Miami University (US)
Publication Details
- Journal
- The Journal of Clinical Psychiatry
- Published
- 2025-05-28
- DOI
- https://doi.org/10.4088/jcp.25m15792
- Citations
- 2
- Primary Topic
- Machine Learning in Healthcare
- Type
- article
- Field-Weighted Citation Impact
- 4.96