#RSNA17

As Google makes its RSNA debut, companies and providers talk up big data partnerships

December 05, 2017
by Lisa Chamoff, Contributing Reporter
CHICAGO — The question on a lot of physicians’ minds has been whether Silicon Valley tech giants are going to revolutionize health care just as they changed how people use the web.

That question was answered with Google’s debut at RSNA 2017, and the company made no secret of how they’re already partnering with other health imaging businesses and facilities to make the buzzwords of cloud computing and machine learning a reality.

“Our mission in health care is much like our mission as a company,” said Gregory Moore, the vice president of Google Cloud Health and a practicing neuroradiologist. “To organize health and life sciences information, to make it accessible, secure — which also means compliant — and useful. As an AI-first company, useful for us is using machine learning and artificial intelligence.”

Moore acknowledged that diagnostic radiology has human limits. As a neuroradiologist, Moore said he typically takes care of 50 patients per day, and with an average of 435 images per study, that means spending 1.52 seconds per image. While radiologists have methods and tools that help them sift through so much data, more help is needed.

Google’s Cloud service is also helping facilities achieve the holy grail of interoperability. Ambra Health, a medical data and image management company, announced during RSNA that it is partnering with Google to offer its Ambra Suite of imaging solutions through Google Cloud.

Morris Panner, the chief executive officer of Ambra Health, said Google allows for greater accessibility, avoiding data silos in imaging, allowing radiologists to more easily share images with referring physicians and patients.

“What this holds out is an honest broker approach to enable the kind of interoperability that everybody is seeking in health care,” Panner said. “What Google lets us do is build those hooks and build that kind of interoperability that really makes a big difference.”

AI development is also impossible with data silos, Panner said.

“Without sufficient data sets, it gets pretty frustrating to run the kind of algorithmic research that we’d like to do,” Panner said.

Boris Zavalkovskiy, director of diagnostic services for ‎Cancer Treatment Centers of America, said the network of five hospitals is using Google’s cloud platform as a duplicate archive so other providers can see what kind treatment patients underwent at their facilities. They also hope to use the data to analyze specific types of cancer.

This will “enable physicians a better opportunity to assess where the patient is in the progression of the disease and what are the better treatments that are out there,” Zavalkovskiy said.

Janak Joshi, senior vice president and chief technology officer at lifeIMAGE, an image exchange platform that also recently announced a partnership with Google Cloud, said the platform helps its clients judge risk.

The company recently worked with Google to identify breast cancer screening rates for high-risk women, finding that about 18 percent of women faced complications because providers were not able to characterize the right information at the right time. They collected information from various data sources, including lifeIMAGE and its partners, as well as claims systems, and put it into Google’s BigQuery‎, which can analyze massive amounts of data quickly.

“Aggregating data on that scale for millions of patients is something we could not have done simply, or without a huge cost burden behind it,” Joshi said. “Ultimately, you’re able to … collaborate across a single patient or cohort of patients in a matter of seconds.”

The company is already looking into several other initiatives with Google.

“We are looking at Google to bring us together in ways that are unprecedented globally,” Joshi said.