Last week GE Healthcare launched the Edison Developer Program to encourage innovation and better deploy artificial intelligence applications for providers. The program will give smaller companies more opportunities to integrate their AI into GE products. Ultimately, GE hopes to make its offerings more efficient and user-friendly.
“As we looked across our portfolio of diagnostics, therapeutics and monitoring, we really felt compelled that the only way we could derive the outcome changes needed for our customers, providers and patients, at a global scale, was to drive a digital thread through the GE Healthcare product line that leverages artificial intelligence among many other digital capabilities,” said Karley Yoder, GE Healthcare’s VP and GM of AI, in a phone interview.
The developer program builds on the existing Edison platform, which leverages partnerships with corporate and academic researchers to funnel AI into GE products. Early efforts are focused on imaging. Yoder notes there are more than 140 venture capital-backed medical imaging startups, with more than $600 million in funding.
GE is particularly interested in technologies like SonoCNS, an AI tool that helps providers assess fetal brain health during sonograms. SonoCNS also simplifies the ultrasound workflow, which is no small thing. GE wants to elevate technology that integrates into existing systems and makes life easier for providers.
“You can create brilliant innovation, brilliant insights but if you break the workflow, if you add more time for a doctor who already has too much on their plate or a radiologist who can’t get through their workload, you’re not going to see adoption of those technologies,” Yoder said.
To help simplify the integration process, GE developed the Edison Open AI Orchestrator, a digital overlay that helps consolidate the company’s AI offerings to make them more user-friendly. To some degree, it’s AI organizing AI.
Beyond imaging, Yoder sees tremendous potential to automate complex processes and generally make healthcare less siloed. She notes there’s a lot of data being produced but very little is actually making it into patient care.
“We’ve never been better at creating data in healthcare, but we’re horrible at using it,” she said. “You see the stats, anywhere from 1% to 3% of data is actually used efficiently within healthcare. We’re starting with use cases that leverage imaging and EMR data…but we’re building the platform in a way that you could layer on genomics, pathology, lab data, etc.”
GE is actively recruiting developers to join the program. They will be working closely with potential partners to ensure each technology is a good fit for specific GE products and the market as a whole.
“We don’t want to just be a shopping cart, where customers can find innovation but then have to figure out how to integrate it on their own,” said Yoder. “We want it to be a one stop shop. So, we actually work with the developers to integrate their work directly through GE healthcare products.”
The company is well aware that technology can, but does not always, generate solutions. Ultimately, GE wants its AI offerings to be helpful additions and not barriers clinicians must learn to overcome.
“I often say that, with the EMR, we took a really painful paper process and made a really painful digital process,” Yoder said. “How do we not do that with AI? We have this incredible technology, but if we just reapply the same processes we have now…we’re not harnessing the technology with the power it deserves.”