AI Science Tracker
EN

Handan Central Hospital

2 discoveries from 2026

Discoveries

A hybrid radiomics–3D CNN model using non-contrast CT for esophagogastric variceal bleeding risk prediction in cirrhosis

Source check: Source checked. Not independent scientific confirmation.

Finding / tested result: A fusion model combining radiomics and 3D CNN features from noncontrast CT predicted esophagogastric variceal bleeding in 137 cirrhosis patients; five-fold results report mean AUC 0.884, with prospective multicenter validation still needed.

Technology: A fusion model combines CT radiomics and 3D CNN features to predict variceal-bleeding risk.

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

AI contribution
A fusion model combines CT radiomics and 3D CNN features to predict variceal-bleeding risk.
Limitation / still unknown
noncontrast CT predicted esophagogastric variceal bleeding in 137 cirrhosis patients; five-fold results report mean AUC 0.
Evidence sources and annotation context

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

Evidence sources: https://doi.org/10.21203/rs.3.rs-10427383/v1 · https://www.researchsquare.com/article/rs-10427383/v1 · https://pubmed.ncbi.nlm.nih.gov/36284789/ · https://www.researchsquare.com/article/rs-10427383/v1.pdf

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: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

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.

Survival-forest-prioritized EPC1 suppresses measured A549 cell phenotypes

Source check: Source checked. Not independent scientific confirmation.

Finding: A lung-cancer study selected EPC1 after it ranked highest in a learned survival model. Overexpression experiments in A549 cells then reduced proliferation, migration, invasion and colony formation. The finding is limited to one cell line; neither a clinical therapy nor direct control of autophagy was established.

Technology: Random survival forest gene-importance ranking

Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.

AI contribution
Learned survival-forest importance guided selection of the experimentally tested gene, following conventional coexpression/differential-expression screening.
Test and result
A random survival forest selected from 117 model combinations ranked EPC1 highest among 69 candidate genes. Original Results explicitly link that ranking to selection for gain-of-function testing. The source reports three independent experimental repeats in A549 cells, reduced proliferation at later measured time points and reduced migration, invasion and colony formation.
Limitation / still unknown
EPC1 had already appeared in an earlier 2020LUAD prognostic signature; this record does not claim first discovery of EPC1 or its association with lung cancer. The prognostic model was selected by average performance across training and two nominal validation cohorts, so those cohorts do not independently confirm a frozen selected model; no prognostic performance number is promoted here. One cell line and overexpression do not establish physiological dose effects, patient benefit or a general tumor-suppressor role. Phenotypic assays do not demonstrate that EPC1 directly regulates chaperone-mediated autophagy. Search absence does not prove experimental priority.
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/PMC13452251/

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 computational gene selection, ensemble model selection/importance Results and biological outcome Results actually read; no experimental recipe needed..

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 3/5 · 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 the interactive profile for the timeline, or browse all organizations and model families.