AI for Doctors / Practical / Obstetrics and Gynaecology

AI in Obstetrics and Gynaecology

AI in obs and gynae is strongest in ultrasound and cervical screening. Tools that guide image capture, measure the fetus and flag anomalies are being adopted. Prediction of preeclampsia or preterm birth is still mostly research.

What AI does in Obstetrics and Gynaecology

Fetal ultrasound plane detection and biometryearly adoption

Guides the scan and measures standard parameters to cut examination time.

Examples: Sonio

Structural anomaly detectionearly adoption

Flags possible fetal anomalies on screening scans for a sonologist to confirm.

Examples: Sonio

AI cervical cytology screeningearly adoption

Scans Pap slides and highlights abnormal cells for review.

Examples: Hologic Genius Digital Diagnostics

Embryo assessment in IVFearly adoption

Ranks embryos from time-lapse images. Evidence on live birth benefit is limited.

Examples: Embryology AI platforms

Preeclampsia and preterm birth predictionresearch

Models combine history, blood pressure and markers to estimate risk.

Examples: Research models

Intrapartum monitoringresearch

Computer analysis of cardiotocography and remote maternal-fetal monitoring.

Examples: Research systems

routine = in everyday use in many centres · early adoption = in use at some sites, evidence still building · research = studies only

Key ideas for this specialty

Standard plane detection: The software recognises the correct ultrasound slice, such as the four-chamber heart view, and tells you when you have it.
Automated biometry: The machine measures head, abdomen and femur for you and estimates fetal weight.
Digital cytology: Pap slides are scanned into images and AI marks the cells a cytologist should review.
Dataset shift: A model trained on one population or machine does worse on another. This matters for Indian patients and low-cost scanners.
Automation bias: Trusting the machine even when it is wrong.

New to the basics? Start with the theory lessons.

Tools

Sonioinstitutional

AI ultrasound reporting software for women's health.

Regulatory status: Not listed

Digital cytology with an AI algorithm for cervical samples.

Regulatory status: FDA cleared (January 2024)

Handheld probe with AI-guided imaging aids.

Regulatory status: Not listed

OpenEvidencefreemium

Clinical question-answering tool for doctors.

Regulatory status: Not listed

MedSAMfree

Open model for segmenting medical images; try it on ultrasound stills.

Regulatory status: Not listed

A listing here is not an endorsement. Ask any vendor for its CDSCO licence and class before clinical use.

Where to learn

ISUOGguidelinefreebeginner

Society risk framework for AI tools in women's ultrasound (Nov 2025).

DeepLearning.AI (Coursera)coursefree auditintermediate

Three courses on image diagnosis, prognosis and treatment models. Needs basic Python.

University of Helsinkicoursefreebeginner

No-maths introduction to what AI can and cannot do. Good first step.

ICMRguidelinefreebeginner

India's own ethics framework. Useful before you plan a thesis using AI.

Stanford AIMIvideo seriesfreeintermediate

Talks and symposium recordings on medical AI.

ISUOGcoursepaidintermediate

9-hour course with CME that includes a unit on AI tools. About 85 GBP for non-members.

Do it yourself

Audit an AI ultrasound claim

Pick one vendor claim, for example fewer missed anomalies. Find the paper behind it. Note the population, machine and reference standard, and whether it fits your clinic.

Open the dataset or tool →
Compare chatbot answers with a guideline

Ask a chatbot a question from a recent FOGSI or ACOG guideline. Check each statement against the document. Log the errors.

Open the dataset or tool →

Use public or properly de-identified data only, and get ethics approval before any project on your own patients.

Papers worth reading

Societies and guidance

AI positioning statement and AI-related ultrasound education.

US body; check its clinical guidance pages for AI statements.

Reproductive medicine; relevant to embryo-selection AI.

Links checked 3 October 2026. “Not listed” means we could not confirm the detail, not that it is absent. Educational material, not clinical advice.