Diagnostics

GE brings in AI-powered software to cardiac ultrasound device

GE Healthcare is incorporating an AI-powered software imaging analysis tool in its Vscan ultrasound device for cardiac ultrasound developed by an Israeli company

The VScan handheld cardiac ultrasound device will now use AI-powered imaging analysis developed by DiA Imaging Analysis

Israeli diagnostic software company DiA Imaging Analysis recently announced that its year-old partnership with GE Healthcare has produced fruit. The companies have made DiA’s AI-powered LVivo EF cardiac decision-support software available for GE’s Vscan Extend handheld ultrasound tool. They believe the combination will make it easier for physicians to more accurately measure cardiac ejection fraction.

“Physicians mainly evaluate ultrasound images by visual estimation,” said DiA CEO and co-founder Hila Goldman-Aslan in a phone interview. “They look at images and say what they think is going on. It’s very subjective, and it’s very dependent on the user’s experience. What we are providing is fully automated measurement and calculation.”

Based in Beersheba, Israel, with a new office in Hartford, Connecticut, seven-year-old DiA stuck with their original goal to make it easier to assess cardiac ultrasound. With support from the Israeli Innovation Authority, they developed automated software to take the guesswork out of image analysis.

“We invented a new image processing technology, based on advanced pattern recognition and machine learning algorithms, that actually imitates the way the human eye identifies borders and motion,” Goldman-Aslan said.

Given the knee jerk reaction among some to AI, she was quick to clarify that the software is meant to augment the work of humans.

“This will not replace the physicians,” he said. “It will empower them to use their training and skills for things they are good at and are less subjective.”

The FDA-cleared software is designed to calculate ejection fraction, the amount of blood exiting the heart each time the muscle contracts. These measurements are particularly important in the emergency room and intensive care unit. In addition to automating analysis, DiA is touting the software’s ability to achieve accurate results in a low-energy, low-memory devices. GE was intrigued by the possibilities.

“Our algorithm can work with low process and power demands of mobile phones and tablets, portable devices that have limitations with that kind of analysis,” Goldman-Aslan said. “Other companies usually use the cloud or very high-end devices that usually go to the ultrasound unit. We were able to create tools that work on the devices themselves.”

DiA recently completed a $5 million funding round, led by Connecticut Innovations. They are also collaborating with Google and others to further expand their technology’s reach.

“Our mission is to become the leader in providing AI solutions for ultrasound,” said Goldman-Aslan. “People want to use ultrasound, and we want them to feel comfortable with the device and the interpretation.”

Photo: GE Healthcare and DiA Imaging Analysis