AI Estimation of Gestational Age

  • Research type

    Research Study

  • Full title

    Evaluation of AI algorithm for Gestational Age Estimation

  • IRAS ID

    356332

  • Contact name

    Kypros Nicolaides

  • Contact email

    k.nicolaides@nhs.net

  • Sponsor organisation

    King's College Hospital NHS Foundation Trust

  • Duration of Study in the UK

    0 years, 4 months, 0 days

  • Research summary

    Accurate estimation of gestational age (GA) is essential for optimal antenatal care, influencing screening, diagnosis, and clinical decision-making throughout pregnancy. The standard method relies on ultrasound measurements of fetal biometry, such as crown-rump length (CRL) in the first trimester or head circumference (HC) in later stages, performed by trained sonographers. However, these measurements require expertise and can be time-consuming.

    This observational study aims to evaluate an artificial intelligence (AI)-based GA estimation algorithm developed by General Electric Healthcare (GEHC). The AI system processes real-time ultrasound video streams to estimate GA without requiring specific biometric measurements. This approach has been tested in low- and middle-income countries, demonstrating promising accuracy comparable to conventional methods. Further validation in routine clinical settings is necessary before considering broader implementation.

    The study will be conducted at the Fetal Medicine Research Institute, King’s College Hospital, London, and will involve 800 participants across a range of gestational ages (8 to 34 weeks). Participants will undergo standard fetal biometry assessments, and the AI system will generate GA estimates. These estimates will remain blinded, will not be displayed to clinicians or patients, and will not influence clinical care.

    The primary objective is to assess whether the AI-based GA estimation is non-inferior to standard biometric methods. Secondary objectives include evaluating the time required for AI-based GA estimation and determining the proportion of scans for which the AI system returns a GA estimate.

  • REC name

    London - Surrey Research Ethics Committee

  • REC reference

    25/LO/0426

  • Date of REC Opinion

    6 Jun 2025

  • REC opinion

    Favourable Opinion