Intel and Samsung detail AI-powered fetal ultrasound tools
Intel and Samsung Medison, a division of Samsung developing ultrasonic devices, today detailed AI-powered, FDA-cleared ultrasound solutions — BiometryAssist and LaborAssist — that automate fetal health measurements. The companies claim BiometryAssist and LaborAssist can deliver a better understanding of a patient’s birthing progress by automatically taking metrics like fetal angle of progression during labor.
According to the World Health Organization, about 295,000 women died during pregnancy and following childbirth in 2017. Research from the Perinatal Institute shows that tracking fetal growth can help prevent stillbirths, as it allows physicians to recognize growth restrictions. But there’s little standardization around fetal measurement because doctors often disagree on where to start.
Intel and Samsung Medison say BiometryAssist and LaborAssist perform measurements in as little as 85 milliseconds, courtesy of Intel’s machine learning toolkit OpenVino. BiometryAssist and LaborAssist estimate the baby’s head direction and fetal position, suggesting the right mode of delivery and potentially reducing the number of unnecessary cesarean sections. Using machine vision algorithms to analyze ultrasounds, the apps are capable of measuring head angle to within 8 degrees with 95% confidence, according to an internal Samsung test on Intel hardware.
Algorithms that ship with BiometryAssist and LaborAssist also guide caliper placement around the fetus. In addition to the biparietal diameter — one of the basic biometric parameters used to assess fetal size — they can measure femur length, head circumference, and abdominal circumference. Moreover, they’re able to segment outlines for the head, taking measurements in an average of 1.5 seconds.
Clinicians can play a demo animation for patients based on the measurements to show labor progress. Other visualization tools reveal a full labor history, including images and measurements in a 3×2 layout.
Intel and Samsung Medison say BiometryAssist and LaborAssist are already in use in 80 countries, including the U.S., Italy, France, Brazil, Russia, and South Korea. Going forward, the companies plan to collaborate on emerging ultrasound technologies, like nerve tracking.
AI is increasingly being applied to ultrasound scans, in part because the high-frequency waves are able to penetrate skin and soft tissue structures, making them ideal for noninvasive imaging of blood clots and the gallbladder, kidney, pancreas, and liver. In August 2018, Dia Imaging Analysis raised $5 million for machine learning tools that analyze ultrasound scans. In September 2018, Butterfly raised $250 million to further develop its semiautomated full-body ultrasound technology. And last month, Exo raised $40 million to bring an AI-powered cloud-based workflow to ultrasound scanners.