Roadmap for Representative Artificial Intelligence Models for Thyroid Cancer
Artificial intelligence (AI) models in thyroid cancer are often not generalizable due to biased or non-representative training and validation datasets. We propose a four-stage roadmap emphasizing representative data collection, inclusion of contextual variables, robust training/validation design, and post-deployment monitoring to improve fairness and clinical utility.
Authors
- Lisa Caulley (ORCID: https://orcid.org/0000-0002-9661-4634)
- Rashi Ramchandani (ORCID: https://orcid.org/0000-0002-6144-6423)
- Evan J. Propst (ORCID: https://orcid.org/0000-0002-8013-8530)
- Risa Shorr (ORCID: https://orcid.org/0000-0003-0388-5812)
- Jonathan D. Wasserman (ORCID: https://orcid.org/0000-0001-7088-8146)
- Eddie Guo (ORCID: https://orcid.org/0000-0002-7223-0505)
- Antoine Eskander (ORCID: https://orcid.org/0000-0003-4326-1393)
- Nikolaus E. Wolter (ORCID: https://orcid.org/0000-0002-5196-6150)
- Sami G. Sabbah
- Jennifer M. Siu (ORCID: https://orcid.org/0009-0007-1040-725X)
- Sanaz G. Biglou
- Michael Mostowy
- Donya Mahiny (ORCID: https://orcid.org/0000-0002-7111-2478)
- Christian Hurtubise
- Gift Anicho‐Okereke
Institutions
- Sunnybrook Health Science Centre (CA)
- University of Ottawa (CA)
- University of Calgary (CA)
- University of Toronto (CA)
- Aarhus University (DK)
- Ottawa Hospital (CA)
- Hospital for Sick Children (CA)
- Toronto East General Hospital (CA)
- Aarhus University Hospital (DK)
- SickKids Foundation (CA)
- Jacobs Institute (US)
- Sunnybrook Hospital (CA)
- Ottawa Hospital Research Institute (CA)
- University at Buffalo, State University of New York (US)
Publication Details
- Journal
- The Laryngoscope
- Published
- 2025-10-03
- DOI
- https://doi.org/10.1002/lary.70169
- Primary Topic
- Radiomics and Machine Learning in Medical Imaging
- Type
- article
- Field-Weighted Citation Impact
- 0.00