Analytical Validation of Multimodal AI Test Predicting Breast Cancer Recurrence Risk (Ataraxis Breast RISK).
Source check: Source checked. Not independent scientific confirmation.
Finding / tested result: A locked H&E-image and clinical-data AI test passed prespecified repeatability and reproducibility checks in 160 breast-cancer cases, including 94.7% agreement across two laboratories/scanners. This is analytical readiness evidence, not independent clinical utility evidence.
Technology: Kestrel vision-transformer patch embeddings are attention-aggregated, combined with clinical variables, and scored by an ensemble of Cox neural networks with calibration.
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Kestrel vision-transformer patch embeddings are attention-aggregated, combined with clinical variables, and scored by an ensemble of Cox neural networks with calibration.
- Limitation / still unknown
- Prespecified analytical acceptance criteria were assessed using repeat runs, two operators, two labs/scanners, controls, perturbed covariates and a clinical-validation bridge; cohort was retrospective and largely ER-positive/HER2-negative, limiting generalization.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13073172/ · https://europepmc.org/article/MED/41975736
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.