Gestational diabetes & diabetes in pregnancy
Machine Learning–Based Prediction of Large-for-Gestational-Age Infants in Mothers With Gestational Diabetes Mellitus
Authors' conclusion
We demonstrated better performance in predicting LGA infants among women with GDM using the AI-based fusion model. The characteristics of the CGM profiles allowed us to determine the appropriate window for intervention.
Study context
- Population
- Women with gestational diabetes mellitus (GDM) from a prospective cohort at a university hospital; Continuous glucose monitoring (CGM) during 20 to 34 gestational weeks
- Participants
- 371 women
- Comparison
- Artificial intelligence (AI)-based convolutional neural network (CNN) fusion model vs 3 conventional models
- Outcome
- Prediction of LGA infants (AUCROC and accuracy)
Methodological detailsprospective cohort, single university hospital, convolutional neural network (CNN)-based fusion model
- Prospective cohort
- Single university hospital
- Convolutional neural network (CNN)-based fusion model
Performance
As reported in the abstract
LGA prediction
| Test or model | AUCROC | Accuracy |
|---|---|---|
| Fusion model | 0.852 (95% CI, 0.680-0.966) | 84.4% |
Fusion model
AUCROC0.852 (95% CI, 0.680-0.966)
Accuracy84.4%
Further findings
Large-for-gestational-age (LGA) incidence
- GDM women with LGA neonates
- 76 (20.5%) of 371
24-hour CGM glucose profiles, LGA vs non-LGA
- Profiles differed at midmorning; consistent across subgroups by pregestational BMI, therapeutic protocol and CGM administration period
Original publication
Machine Learning–Based Prediction of Large-for-Gestational-Age Infants in Mothers With Gestational Diabetes Mellitus. J Clin Endocrinol Metab. Published 16 July 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →