Unsupervised machine learning reveals prognostic value of dynamic carbohydrate antigen 125 trajectory in gastric cancer
Source check: Source checked. Not independent scientific confirmation.
Finding: Four trajectories among 1,015 patients showed differing survival associations after stage adjustment. The trajectory-only model had a C-index of 0.74 versus 0.59–0.66 for one-time measurements in the same data. Retrospective single-center data with 7,347 measurements. The comparisons are within the cohort and do not establish external validity or a causal effect.
Technology: Unsupervised K-means grouped serial CA125 measurements into trajectories without survival or stage labels; subsequent survival models estimated associations.
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Unsupervised K-means grouped serial CA125 measurements into trajectories without survival or stage labels; subsequent survival models estimated associations.
- Test and result
- Four trajectories among 1,015 patients showed differing survival associations after stage adjustment. The trajectory-only model had a C-index of 0.74 versus 0.59–0.66 for one-time measurements in the same data.
- Limitation / still unknown
- Retrospective single-center data with 7,347 measurements. The comparisons are within the cohort and do not establish external validity or a causal effect.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://link.springer.com/article/10.1186/s12885-025-15268-5
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 source full text Methods and Results; version/date checked against bibliographic record..
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 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.