Improving Diagnostic Precision in Thyroid Pathology by Synergistic Use of AI and Molecular Markers
Authors' conclusion
Integrating AIBx V2 imaging model with ThyroSeq v3 preserved the high sensitivity of molecular testing while improving specificity and PPV.
Study context
- Population
- Indeterminate thyroid nodules42 with surgical pathology
- Participants
- 108 participants
- Comparison
- AIBx V2 alone vs ThyroSeq v3 alone vs combined AIBx V2+Mol
- Outcome
- Diagnostic test performance vs surgical pathology
Methodological detailsretrospective
- Retrospective
Performance
As reported in the abstract
Test performance in surgical pathology subset (n = 42)
| Test or model | AUC | Sensitivity | Specificity | PPV |
|---|---|---|---|---|
| ThyroSeq v3 | 0.70 | 0.95 | 0.45 | 0.65 |
| AIBx V2 | — not stated in the abstract | 0.77 | 0.60 | — not stated in the abstract |
| AIBx V2+Mol | 0.77 | 0.95 | 0.60 | 0.72 |
AUC
ThyroSeq v30.70
AIBx V2— not stated in the abstract
AIBx V2+Mol0.77
Sensitivity
ThyroSeq v30.95
AIBx V20.77
AIBx V2+Mol0.95
Specificity
ThyroSeq v30.45
AIBx V20.60
AIBx V2+Mol0.60
PPV
ThyroSeq v30.65
AIBx V2— not stated in the abstract
AIBx V2+Mol0.72
— not stated in the abstract
Further findings
| Outcome | ThyroSeq v3 | AIBx V2+Mol |
|---|---|---|
| Test performance in entire cohort (n = 108; nodules without surgical pathology deemed benign) | ||
| Sensitivity | — not stated in the abstract | 0.95 |
| Specificity | 0.87 | 0.90 |
| PPV | 0.65 | 0.72 |
| AUC | 0.91 | 0.93 |
— not stated in the abstract
Original publication
Improving Diagnostic Precision in Thyroid Pathology by Synergistic Use of AI and Molecular Markers. Endocr Pract. Published 3 February 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →