LUNA: AI Ultrasound for Birthweight Estimation
Research type
Research Study
Full title
LUNA: Evaluating the Accuracy of AI-Enabled Point-of-Care Ultrasound for Fetal Weight Estimation
IRAS ID
368110
Contact name
Argyro Syngelaki
Contact email
Sponsor organisation
King's College Hospital NHS Foundation Trust
Duration of Study in the UK
1 years, 3 months, 6 days
Research summary
We will recruit around 1,000 pregnant women at King’s College Hospital and Medway Maritime Hospital.
In all women, we will compare three ways of assessing baby growth: the standard tape measure (symphysis–fundal height), routine ultrasound scans, and an AI-supported ultrasound using a handheld device.
In one group of 500 women, we will assess how accurately each method estimates the baby’s weight close to birth (ideally within 48 hours).
In another group of 500 women, we will follow growth over time during pregnancy. At routine antenatal visits (around 24, 28, 32, 36 and 40 weeks), midwives will measure growth using the tape measure, perform a standard ultrasound scan, and carry out a brief AI-supported ultrasound sweep. This will allow us to see how well each method tracks baby growth across pregnancy.
The AI uses simple movements across the abdomen to estimate the baby’s weight automatically. The AI results will be recorded but not shown to the midwife or the woman, so that normal care is not affected.
After birth, we will compare all methods with the baby’s actual birthweight. This will help us understand which approach is most accurate, how consistent the measurements are, and whether AI could improve current care.
REC name
East Midlands - Leicester South Research Ethics Committee
REC reference
26/EM/0143
Date of REC Opinion
1 Jul 2026
REC opinion
Unfavourable Opinion