Related Experiment Videos
Artificial Intelligence in Obstetrics: Current Trends and Future Directions
1Gray Faculty of Medical and Health Sciences, Tel Aviv University, Tel Aviv 6997801, Israel.
Journal of Clinical Medicine
|August 13, 2026
Summary
Artificial intelligence (AI) shows promise in obstetric care, particularly for image-based tasks like fetal biometry. However, AI tools require more external validation and clinical-utility assessment before widespread adoption in healthcare.
Area of Science:
- Obstetric care applications of AI, machine learning (ML), deep learning (DL), computer vision, and natural language processing (NLP).
Background:
- AI is increasingly utilized in various obstetric domains, including ultrasound interpretation, fetal monitoring, maternal risk stratification, and telehealth.
- Applications span from biometry and anomaly detection to preeclampsia prediction and labor support.
Purpose of the Study:
- To review the literature on AI applications in obstetric care published between 2016 and 2026.
- To assess the level of evidence supporting each AI application in obstetrics.
Main Methods:
- A narrative (non-systematic) review of published literature from 2016-2026.
- Distinguishing the level of evidence for various AI applications in obstetrics.
Main Results:
- High performance reported for image-based tasks like fetal biometry and anomaly detection (accuracy, AUC > 0.85).
- Modest performance for intrapartum cardiotocography (CTG) analysis (AUROC ~0.60-0.70), similar to clinician variability.
- Most evidence is retrospective and internally validated; limited external/prospective validation and regulatory clearance.
Conclusions:
- AI holds significant potential for improving prenatal diagnosis and personalized obstetric care.
- Current claims of clinical readiness are often premature.
- Prerequisites for safe AI adoption include prospective/external validation, calibration, clinical-utility assessment, transparent reporting, bias evaluation, and clinician oversight.
