Giulia Zamagni: Critical Gaps in Machine Learning Models for Fetal Growth Restriction
Giulia Zamagni, PhD Candidate in Applied data Science and Artificial Intelligence at Università degli Studi di Trieste, shared a post on LinkedIn about a paper published in American Journal of Obstetrics and Gynecology MFM:
“I’m happy to share our latest work, published in American Journal of Obstetrics and Gynecology MFM: ‘Assessing Adherence to TRIPOD+AI Guidelines in Machine Learning Models for Predicting Small for Gestational Age and Fetal Growth Restriction: A Systematic Review’.
What did we find?
Despite the growing enthusiasm for machine learning in perinatal medicine, our systematic review reveals several critical gaps that limit clinical applicability:
Not a single model evaluated performance across clinically relevant subgroups. This means we have no idea whether these models behave differently – and potentially inequitably – for different populations.
Only 15% of studies evaluated calibration.
This is perhaps the most alarming finding.
In clinical prediction models, calibration is essential: predicted risks directly inform decisions such as surveillance intensity, timing of delivery, or need for intervention.
A model that is well-discriminating but poorly calibrated can overestimate or underestimate risks, leading to unnecessary interventions or missed opportunities for timely care. Yet, calibration was almost entirely overlooked.
Why this matters
Developing ML for healthcare must be more than a technical exercise: it requires robust methodology and thoughtful evaluation to truly support patient care.”
Title: Assessing adherence to TRIPOD+AI guidelines in machine learning models for predicting small for gestational age and fetal growth restriction: a systematic review
Authors: Giulia Zamagni, Camilla Fregona, Moira Barbieri, Maria Sole Scalia, Lorenzo Monasta, Christoph Lees, Tamara Stampalija, Giulia Barbati
Read the full article.
Stay updated on all scientific advances in the field of fertility with Fertility News.
-
Apr 30, 2026, 18:45AI in Ultrasound Follicular Monitoring Moves Toward Clinical Decision Support – RBMO
-
Apr 30, 2026, 11:42ASRM Encourages Registration for Conference on Practice Growth and Innovation
-
Apr 30, 2026, 11:26ESHRE and ALPHA Launch Global Survey to Update ART Laboratory KPIs
-
Apr 30, 2026, 11:10Susanne Johnson: Advanced Gynaecology Ultrasound Course Day 2
-
Apr 30, 2026, 11:02Platelets Reduce Oxidative Stress Damage in Human Endometrial Stromal Cells in Vitro – RBMO
-
Apr 30, 2026, 10:57GREEG Day 2026: Elevate Your Ultrasound Expertise – GREEG
-
Apr 30, 2026, 10:49Up to Two-Thirds of Elite Female Athletes Experience Irregular Periods – Fertility Plus
-
Apr 30, 2026, 10:47Applications for the ESHRE5 Young Ambassadors 2026 are Still OPEN, But Not for Long! – ESHRE
-
Apr 30, 2026, 10:42The Link Between Lifestyle and Anti-Müllerian Hormone – RBMO