The Consortium has developed new international standards describing optimal fetal and newborn growth, which are being incorporated into clinical care around the world to improve health outcomes.
The INTERGROWTH-21st Project involved collecting data throughout pregnancy, including over 500,000 fetal ultrasound images, from healthy women in eight geographically diverse regions of the world.
To classify the anatomical structures in those images automatically, the Consortium applied MedaPhor’s ScanNav® intelligent ultrasound technology, which uses AI algorithms and deep learning techniques to achieve over 99% accuracy.
Professor Stephen Kennedy, Head of the Nuffield Department of Women’s & Reproductive Health at the University of Oxford and Co-Principal Investigator of the INTERGROWTH-21st Project, commented: “ScanNav has categorised the Project’s massive ultrasound database in a way that would have been impracticable with human curation. Deep mining of the data has been made much easier.”
Dr Mohammad Yaqub, VP Engineering at MedaPhor, commented: “We have not only catalogued the INTERGROWTH-21st ultrasound database, but also demonstrated that ScanNav can classify fetal scans accurately in both the second and third trimesters of pregnancy. We hope the curation we have produced will help researchers in the future.”