AI Science Tracker
EN

AI angiography detects more reference coronary lesions than visual estimation

Asan Medical Center; University of Ulsan College of Medicine; Ulsan University Hospital; Keimyung University Dongsan Medical Center · 29 Jan 2025 · medicine

Discovery

AI angiography detects more reference coronary lesions than visual estimation

Source check: Source checked. Not independent scientific confirmation.

Finding: An ensemble-based angiography system detected 1,705 of 1,828 manually defined coronary lesions in 420 patients, compared with 69% detection by visual estimation. Models trained on separate 2016 data were tested retrospectively on selected 2022 images from three hospitals. The result establishes reference-based lesion detection, not improved treatment outcomes.

Technology: MPXA-2000; DeepLabV3+, U-Net++ and U²-Net segmentation ensemble

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Learned vessel segmenters generated lumen boundaries; an ensemble and downstream lesion analysis located and quantified lesions for comparison with manual QCA and cardiologist visual estimates.
Test and result
Retrospective three-center evaluation of 1,076 selected angiography images from 420 patients. Nine view-specific segmentation models combined DeepLabV3+/EfficientNet-B4, U-Net++/EfficientNet-B4 and U²-Net; training images from 2016 were separate from 2022 evaluation data. Against blinded manual QCA, the system detected 1,705/1,828 lesions (93%) versus 69% for visual estimation; on multiple-lesion images it identified all reference lesions in 86% versus 33%.
Limitation / still unknown
Manual QCA is the anatomical reference, not physiological significance or clinical outcomes. Selected frames excluded occlusions and poor-quality images; only reference lesions meeting size/stenosis criteria were counted. Lesion observations repeat within patients, and reported significance does not establish cluster-adjusted uncertainty. No prospective procedural or patient benefit was tested; training details in the supplementary material were not independently inspected.
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/PMC11880186/

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: 21 Nov 2026. Review scope: Read actual primary cohort, separate model-training data, named segmentation ensemble, blinded manual/AI/visual-reference Methods, detection definition, statistical Methods, Results and limitations. Supplementary training tables were not independently inspected. Exact primary identity/date and fresh primary PubMed correction/retraction relationships checked..

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.

Open in the full catalog for filters, charts and related discoveries. Browse medicine discoveries.

Related catalog entries

Links use documented relationships where available, plus shared catalog context. A shared topic or model family does not establish a scientific dependency.