Learned brain-imaging measurements agree with expert references across separate cohorts
Source check: Source checked. Not independent scientific confirmation.
Finding: A CNN and YOLO-based pipeline automated region selection for a brain diffusion-imaging index. In a separate 1000-person UK Biobank test, agreement with the average of two neurologists’ measurements gave an intraclass correlation of 0.90; smaller OASIS3 and EPAD tests gave 0.80 and 0.78. These validate measurement agreement, not direct brain waste clearance or clinical benefit.
Technology: VGG-style convolutional network; YOLOv5s; dALPS
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- CNN slice selection and learned object detection automate the original index region placement; the index formula itself is preserved. Their outputs enable the cross-cohort reference-agreement evaluation.
- Test and result
- Development used 1000 subjects from four other cohorts, with subject-level training/validation splits. The finalized pipeline was tested against averaged expert measurements in 1000 UKB subjects and 200 each from OASIS3 and EPAD.
- Limitation / still unknown
- Manual measurements are an imperfect expert reference, not a physiological gold standard. Cohort/scanner distributions differed. Abstract and discussion quote ICC 0.95, while specific independent-test Results give0.90/0.80/0.78; this entry follows those explicit test results. No causal genetic, medication or disease mechanism, treatment guidance or outcome benefit is represented. Supplementary methods were not independently inspected. The linked corrigendum adds OASIS acknowledgments without a stated scientific result change. External archived cohorts do not establish prospective new-data testing or separate-team replication.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://orca.cardiff.ac.uk/id/eprint/184570/1/1-s2.0-S2352396426000149-main.pdf · https://pubmed.ncbi.nlm.nih.gov/41619353/ · https://pmc.ncbi.nlm.nih.gov/articles/PMC12933589/
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: Primary institution-deposited article front, computational imaging development, independent expert-reference testing Methods/Results and measurement caveats; primary linked correction body.
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 2/5 · 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.