Whenever Dr. Ralf Menkhaus prepares to administer future parents their first fetal ultrasound, he knows the pressure is on: Equal parts thrilled and anxious, expectant parents are desperate to catch a glimpse of their unborn child. Yet, as a fetal medicine specialist, Menkhaus’ top priority is capturing crucial information about the health of the fetus. He’s looking for evidence, for instance, of conditions like spina bifida, a neural tube defect that affects the spinal cord. Following a trail of clues, such as changes in the fetal brain, doctors may make this diagnosis during an ultrasound at the 20th week of pregnancy.
But measuring fetal brains is an imprecise science. To get it right, Menkhaus must make multiple measurements with the ultrasound machine, relying on his 20 years of experience to swiftly and accurately interpret what he sees. Even though he takes the utmost care, his assessment could differ from that of another sonographer, possibly leading to a missed or delayed diagnosis. All of this goes on while nervous parents-to-be gaze at Menkhaus expectantly.
Menkhaus was delighted, then, to discover a way to simplify the process. For the last few months, he and his colleagues in Minden, Germany, have been using GE Healthcare’s SonoCNS, an artificial intelligence tool that automates fetal brain assessment and measurements. A longtime devotee of Voluson E10, the ultrasound system SonoCNS runs on, Menkhaus was asked by the company for his feedback on how the new technology would change his daily workload.
As someone who trains medical professionals to use ultrasound systems, the first thing he noticed was how easy the AI technology was to learn. Rather than memorizing a long list of procedural steps to take fetal brain measurements, sonographers just obtain a 3D volume of the fetal brain with the system. SonoCNS will identify the standard planes for anatomy assessment and perform measurements and calculations.
SonoCNS was trained using Edison, GE Healthcare’s cloud-based app development and data storage and analysis platform, to determine how the fetal brain is developing. “It’s particularly beneficial for less experienced gynecologists,” Menkhaus says. “Now, with the click of a button, they can have accurate and reliable measurements.” By standardizing fetal brain measurements, anyone can receive timely and trustworthy diagnostic care regardless of where they’re treated.
Taking as much guesswork as possible out of these measurements can be lifesaving. For instance, agenesis of the corpus callosum is a rare fetal condition in which the band of white matter connecting the two hemispheres of the brain fails to develop properly. It’s challenging to diagnose — but, if they detect it early enough, doctors can initiate further diagnostic steps like karyotyping or fetal MRI to counsel future parents more precisely.
Better yet, Menkhaus no longer feels so rushed during his ultrasound appointments. SonoCNS reduces the number of required keystrokes by three-quarters. That “saves at least half a minute per scan,” he estimates, “which adds up quickly when you’re seeing several patients.”
Those spared minutes can make all the difference. They’re the moments when Menkhaus gently delivers news of an upsetting diagnosis, taking time to help expectant parents seek more treatment or prepare for a child with special needs. In most cases, however, he gets to allay patients’ fears and watch the looks on their faces when he reveals if the nursery will be pink or blue. In moments like that, Menkhaus can revel in the joy of his profession.
A version of this article originally appeared on GE Healthcare’s newsroom, The Pulse.