Artificial Intelligence Models to Predict Recurrence in Patients with Differentiated Thyroid Cancer: A PROBAST + Artificial Intelligence (AI) Systematic Review of Clinical Applicability
Authors' conclusion
AI models for predicting DTC recurrence show promising discriminatory performance but have substantial methodological shortcomings in development, and a lack of calibration and external validation necessary for clinical use. Given the guidelines' emphasis on risk-adapted de-escalation of therapy, the development of robust, validated, and clinically calibrated AI tools is a priority. Current evidence remains insufficient to support routine clinical use.
Study context
- Population
- Original research developing or validating AI models for recurrence in differentiated thyroid cancer (DTC)patients encompassed by 34 included studies (21 unique datasets and 13 UCI-derived publications)
- Participants
- 21,267 participants
- Outcome
- Studies included
Methodological detailsPROBAST + AI framework
- PROBAST + AI framework
Pooled result
As reported in the abstract
- Records screened
- 673
- Studies included
- 34
- Recurrence rate
- 7.8%
- Developed in Asian cohorts
- 17 of 21 datasets
- Structured clinical/tabular variables
- 82.4%
- Reanalysis of a single dataset
- 13 of 34, 38%
Further pooled results
Discrimination
- Discrimination reported
- 22 of 34 studies (64.7%)
- AUC
- 0.67-1.00
- Studies assessing calibration, none with standardized metrics
- 10
PROBAST + AI assessment
- High overall quality concerns in development studies
- 26 of 34 (76.5%)
- Low overall quality concern
- 6 of 34 included studies (17.6%)
- Low applicability concern
- 17 of 34 (50.0%)
Original publication
Artificial Intelligence Models to Predict Recurrence in Patients with Differentiated Thyroid Cancer: A PROBAST + AI Systematic Review of Clinical Applicability. Thyroid. Published 12 November 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →