Artificial intelligence is beginning to provide an early down payment on benefits to healthcare providers, but the industry is still early in understanding how to use advanced computing to improve care.

Still, the potential is great, because providers are accumulating significant patient data that can be used to deliver precise and effective care, said presenters this week at Solve: Healthcare, an event sponsored by Intel to discuss the role of AI in healthcare and medicine.

“We’re using less than 5 percent of (patient) data” as clinicians make decisions on providing care, says Rachel Callcut, MD, a trauma surgeon and director of data science and advanced analytics for UCSF Medical Center.

Also See: How artificial intelligence is already paying dividends in healthcare

“It’s hard for the bedside physician to synthesize all this information,” Callcut adds. “Even if you think about the most simplistic application of AI—and we’re at the very beginning of integrating AI into care delivery—it’s clear we’ll get to the point where AI will do things that we can’t even imagine right now.”

However, AI and related technologies are being overhyped at this point, says Michael Blum, MD, director of the Center for Digital Health Innovation at UCSF, where he is also the associate vice chancellor for informatics and a cardiologist.

“The reality is that AI is not how we’re going to fix problems,” Blum says. Answers won’t come by just implementing a new technology portfolio. “That’s not transformational. We have to use data and AI to keep you from getting desperately ill.”

Some early uses of AI and other advanced technologies have proved promising, presenters noted. For example, brain imaging has the capability of showing what parts of the brain are engaged in different emotional states, but the associated computations to derive benefits would take years to perform, until recently, says Jonathan Cohen, a professor of psychology and co-director of the Princeton Neuroscience Institute.

“Now, we can take two years of computing time down to real-time, and we’re able to provide a patient with information about their internal state,” he says. That offers the potential for giving patients insight into their emotions and the ability to recognize patterns and control them. “It’s a transformative change in what we can do, and we’re starting clinical trials on this.” AI has the potential to give neuroscientists insights into how the brain works and better treat patients, he says.

AI is enabling Intermountain Healthcare to provide personalized medicine, says Lonny Northrup, senior health informaticist at the integrated delivery system. “We’ve moved from reporting why things have happened to prescribing what should happen at a population level,” he says.

Intermountain is using analytics and AI with 70 care process models, and is working to use predictive analytics to determine what is likely to happen to individuals. Intermountain is looking to “connect the patient to his or her care team 24 hours a day through a personalized platform that’s learning the patient’s motivational profile, helping that patient achieve at least 95 percent engagement with the things that will help them improve their health.”

Making sense of the vast stores of patient medical information that’s been digitized is one of the most promising uses for AI, says David Holmes, collaborative scientist and biomedical engineer at The Mayo Clinic.

“Mayo has a long history of collecting medical data, and up until the last 20 years, it’s been a very manual process,” Holmes says. “The biggest place AI can help us now is in wrangling the data to make it easier to use. If we can curate data, we can better use it.”

Mayo hopes to have researchers cull through data to find patterns, but it also hopes to use computational power to look through medical information to find patterns that researchers wouldn’t typically expect.

AI will help providers improve care as it extends to settings outside of hospitals’ walls, Holmes adds. “As we do more home healthcare, we have to move fluidly between devices in the home and the inpatient setting,” he notes. “AI will be at the edge of care and will be more critical in home healthcare,” he predicts.

Register or login for access to this item and much more

All Health Data Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access