Questionnaire-based machine learning predicted congenital heart disease in a retrospective case-control study
Source check: Source checked. Not independent scientific confirmation.
Finding: A retrospective multicenter case-control study used questionnaire data and explainable machine-learning models to identify factors associated with congenital heart disease and predict case status. The best reported model had modest discrimination, including on an external hospital sample; this is not causal evidence.
Technology: LASSO feature selection and comparison of support-vector machine, XGBoost, random forest, logistic regression and LightGBM models, interpreted with SHAP.
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- LASSO feature selection and comparison of support-vector machine, XGBoost, random forest, logistic regression and LightGBM models, interpreted with SHAP.
- Test and result
- XGBoost achieved AUC 0.724 internally and 0.706 in an external sample of 285; echocardiography supplied the reference diagnosis.
- Limitation / still unknown
- Retrospective case-control prediction with modest discrimination; associations do not establish causal risk factors or clinical benefit.
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/PMC13162403/
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 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.