AI Science Tracker
EN

ScreenDx identifies previously underdetected interstitial lung disease in COPDGene scans

Stanford University; IMVARIA; University of North Carolina; University of Utah; University of Alabama at Birmingham · 26 Nov 2025 · medicine

Discovery

ScreenDx identifies previously underdetected interstitial lung disease in COPDGene scans

Source check: Source checked. Not independent scientific confirmation.

Finding: A retrospective COPDGene analysis used a learned CT model to detect cases subsequently relabeled as clinically significant interstitial lung disease. The original abstract reports 84.8% sensitivity and 98.0% specificity in a deliberately sampled cohort. This does not establish prospective screening benefit.

Technology: ScreenDx deep learning CT classifier

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

AI contribution
A learned ScreenDx model evaluates CT features compatible with ILD; this is a retrospective test against subsequently assigned clinical reference labels.
Test and result
The source selected patients subsequently relabeled by COPDGene investigators as ILD-positive and sampled COPD/control negatives to target approximately 1–2% ILD prevalence. At a previously selected threshold, abstract-reported sensitivity was 84.8%(95% CI 68.1–94.9%) and specificity 98.0%(97.3–98.5%).
Limitation / still unknown
This review had the complete original abstract and primary bibliographic metadata, not full Methods. Exact sample counts, details of clinical adjudication, model-development overlap and threshold-lock timing were not authenticated. Positives had already been relabeled by investigators; these are not newly confirmed diagnoses made in a prospective AI screening program. Deliberate negative sampling limits prevalence-dependent interpretation. No improved patient outcomes or causal clinical benefit is established. The original journal history gives online publication November 26,2025, while PubMed records November 27; the represented date follows the original publisher.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://pubmed.ncbi.nlm.nih.gov/41314436/ · https://www.resmedjournal.com/article/S0954-6111%2825%2900608-0/abstract

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: Complete original PubMed/EuropePMC abstract plus indexed original journal abstract and original publisher publication history actually read; full Methods unavailable (journal403)..

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.

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.