The first pilot of ScanNav, MedaPhor Group’s real-time image analysis software, will take place at the Fetal Medicine Department of St George’s University Hospitals NHS Trust, in London.
ScanNav is believed to be the first CE marked artificial intelligence (AI) system to carry out an automated, real-time “peer review” of obstetric ultrasound images as the patient is scanned, according to MedaPhor, a provider of intelligent ultrasound software and simulation.
Monitoring performance by manually auditing images retrospectively is very time consuming, bu ScanNav supports clinical staff by instantly confirming the images they save conform to protocol, meaning additional images can be taken straight away if required.
Initially targeted at the U.K. NHS Fetal Anomaly Screening Programme (offered to all women at 20 weeks of pregnancy), ScanNav evaluates more than 50 individual criteria to verify the six views required by the pregnancy screening programme are complete and fit for purpose.
ScanNav uses deep learning technology to assess the same features that sonographers look for in ultrasound images. The system has "learnt" this using over 350,000 images that were assessed by a panel of senior sonographers. Initial validation studies show the AI system is as good as an expert colleague in providing peer review.
“Newly qualified sonographers, or those in training, may find this particularly helpful, giving confidence and enhancing skills to attain excellent imaging techniques,” says Katy Cook, Lead Sonographer in St George’s Hospital Fetal Medicine Department. “This AI software could also potentially automate the required auditing for obstetric scanning and demonstrate quality and competency for every sonographer in a busy clinical setting. This new way of assessing images looks very interesting and could have great potential.”