Developing a Simple Non–Laboratory-Based Machine Learning Tool for Prediabetes Screening in a Target Population: A Proof-of-Concept Study
Authors' conclusion
A machine learning screening tool using age and waist circumference was developed with promising results. Its simplicity, by only requiring two non-laboratory features, allows for easy implementation. However, to verify the model's generalizability and external validity, it needs to be evaluated using additional data.
Study context
- Population
- Adults from a prediabetes screening project; prediabetes HbA1c ≥39 and <48 mmol/mol vs normoglycemia HbA1c <39 mmol/mol88 with prediabetes
- Participants
- 501 adults
- Comparison
- Training set vs Validation set
- Outcome
- Detection of prediabetes (ROC AUC) by a non-laboratory-based ML tool
Methodological detailsproof-of-concept, decision tree machine learning model, 70% training / 30% validation split, forward feature selection, Twelve potential non-laboratory-based features extracted
- Proof-of-concept
- Decision tree machine learning model
- 70% training / 30% validation split
- Forward feature selection
- Twelve potential non-laboratory-based features extracted
Performance
As reported in the abstract
| Test or model | AUC |
|---|---|
| Training set | 0.8297 |
| Validation set | 0.7961 |
AUC
Training set0.8297
Validation set0.7961
Original publication
Developing a Simple Non–Laboratory-Based Machine Learning Tool for Prediabetes Screening in a Target Population: A Proof-of-Concept Study. J Diabetes Sci Technol. Published 22 September 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →