A coronary CT model retains diagnostic discrimination in sex subgroups
Source check: Source checked. Not independent scientific confirmation.
Finding: A post-hoc analysis tested an existing coronary CT ischemia model against invasive pressure measurements in two cohorts. Vessel-level AUCs in women and men were 0.85 and 0.87 in CREDENCE and 0.81 and 0.86 in PACIFIC-1. These subgroup results do not establish equivalent performance, prospective clinical benefit or a newly developed model.
Technology: Convolutional coronary CT analysis and random-forest ischemia classifier
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Existing neural segmentation/plaque quantification supplied features to a separately trained random forest for vessel-level ischemia prediction.
- Test and result
- The previously developed classifier used quantitative plaque and stenosis features from neural image analysis; its random forest was trained in a separate 307-patient CREDENCE derivation cohort. This post-hoc study evaluated the 305-patient/868-vessel CREDENCE validation cohort and 208-patient/612-vessel PACIFIC-1 cohort. Sex-specific AUCs exceeded SPECT/MPI comparators; differences from PET in PACIFIC-1 were not significant.
- Limitation / still unknown
- Vessels are clustered within patients; subgroup tests were post hoc and do not establish equivalence or subgroup fairness. These are reused observational trial cohorts, not four independent samples or a new prospective deployment. Nonsignificant PET differences do not prove equivalence. Clinical outcomes and cost-effectiveness were not tested.
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/PMC12923349/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 21 Nov 2026. Review scope: Original full-text bibliographic date, relevant computational/statistical Methods and Results, bounded cohort and selection caveats; no operational clinical instructions..
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 1/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.