Learned CT perfusion patterns correlate with pulmonary hypertension
Source check: Source checked. Not independent scientific confirmation.
Finding: A Bayesian Gaussian-mixture model separated lung attenuation patterns after neural segmentation in CT scans from 52 patients with chronic thromboembolic pulmonary hypertension. A larger normal-attenuation fraction was associated with lower pulmonary vascular resistance and other measured severity indicators. These single-centre correlations do not establish causality, prognosis or replacement of catheter measurements.
Technology: Variational Bayesian Gaussian-mixture model; TotalSegmentator
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- A learned mixture model fits patient-specific attenuation distributions while a pretrained neural network delineates lungs and veins for spatial quantification.
- Test and result
- Automated CT measures compared with invasive hemodynamics, clinical biomarkers and radiologist annotations, with multiple-comparison correction.
- Limitation / still unknown
- External clinical utility and outcome prediction remain unestablished; a fixed three-component assumption and scanner/artifact effects can alter inferred tissue classes.
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/PMC12559064/
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: Primary relevant model Methods, Results and limitations; exact primary identity and current 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.