AI Science Tracker
EN

Justus Liebig University Giessen

1 discovery from 2025

Discoveries

A small vascular pilot compares simulated endoleak risk with observed outcomes

Source check: Source checked. Not independent scientific confirmation.

Finding: A single-center pilot assessed a learned endoleak risk index in 10 historical and 10 planned vascular-repair cases. In the historical group, four of six high-risk cases developed the target endoleak, while four low-risk cases did not. The small mixed study does not establish that model-guided planning reduces complications.

Technology: Finite-element digital twins and machine-learned Endoleak Risk Index

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

AI contribution
Learned aggregation of geometric/biomechanical simulation indicators produced a patient-specific risk category; clinicians retained decision responsibility.
Test and result
The commercial model combines about 50 engineered simulation indicators using machine learning. In 10 retrospective cases with at least 3 years of follow-up,6 had elevated risk and 4 developed type 1a endoleak;2 elevated-risk cases remained event-free. The additional 10 planned cases assessed clinician decisions; not all underwent the originally planned intervention, and follow-up was short.
Limitation / still unknown
Only 10 historical cases support the represented outcome comparison, with no controlled assessment of planning benefit. The learned algorithm architecture and training overlap are not detailed here. Prospective decisions changed treatment or prevented the planned intervention in some cases, so absence of an event cannot validate an untreated counterfactual. The planned cohort had only median 3 months follow-up and no target events.
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/PMC12704054/

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 full-text bibliographic dates, computational classification Methods, held-out Results and explicit cohort/reference limitations..

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 the interactive profile for the timeline, or browse all organizations and model families.