GE’s Edison partnership with ACR AI-Lab to aid algorithm development

09-Apr-2019 - Last updated on 09-Apr-2019 at 14:05 GMT

(Image: Getty/peshkova)

GE Healthcare’s Edison platform collaborates with ACR AI-Lab to extend its AI capabilities and allows ACR members access to algorithm creation in an aim to turn research into product development.

Through the partnership, the American College of Radiology’s ACR AI-LAB, a software platform for radiologists to develop and validate artificial intelligence (AI) algorithms, will use GE Healthcare’s Edison platform for the development and adoption of AI.

Launched in November 2018, the platform’s data aggregation services enable radiologists to combine data sets and clinical information from electronic health records (EHRs) to create algorithms. According to the company, the platform acts as a “digital thread” for its existing partnerships.

“Edison is GE Healthcare’s intelligence platform, designed and optimized to create new algorithms from third parties. The ACR just unleashed a massive number of institutions to write algorithms, and GE Healthcare and Edison help these institutions get their algorithms into production use in Edison applications and devices,” Kieth Bigelow, general manager and SVP of AI and analytics at GE Healthcare told us.

Through collaboration with ACR AI-LAB, ACR members will have access to Edison, making it easier for them to develop and deploy algorithms across hospitals and national research centers.

Kieran Murphy, president and CEO of GE Healthcare said in a statement that through the partnership Edison will provide ACR with diverse data sets from across modalities, vendors, health care networks, and life sciences settings.

AI to convert research to product development

Bigelow told us that the idea behind integrating Edison with ACR AI-Lab was simply, “scale.”

Discussing the potential power of AI in health care, he said “the opportunity before us requires a thriving ecosystem of AI in order for us to bring about the health care transformation we seek.”

Giving the example of chest x-rays, which are able to detect more than 740 medical conditions, the ability to multiply the anatomy by the various modalities like CT scans and MRI’s will require thousands of algorithms.

“That’s where the ACR AI-LAB consortium comes into play. With its 38,000 members, ACR has the volume, scale, and clinical expertise to develop applications specific to health care and address the long tail opportunity before us,” Bigelow explained.

Supporting ACR and the greater ecosystem of tech and industry academics is what, according to GE Healthcare, will develop AI “that will truly transform health care globally.”

Bigelow stated that this partnership demonstrates why Edison was created in the first place “to easily build safe and effective AI [and to] dramatically accelerate the number of useful algorithms from providers across the world and turn the corn to convert AI from research to true product development.”