Speakers take part in a panel titled AI in Healthcare: The Big Questions That Still Need to Be Answered, at Cleveland Clinic's 2019 Medical innovation Summit.Photo by Omar Ford
CLEVELAND – Two words dominated much of the program agenda for Cleveland Clinic’s 2019 Medical Innovation Summit (MIS) on Tuesday – Artificial Intelligence.
From a keynote address from Craig Mundie, Senior Advisor to the CEO, and former Chief Research and Strategy Officer, Microsoft to a session discussing companies with non-traditional approaches to healthcare, the topic of AI was discussed.
But is AI just the latest fad in healthcare? Is it fully-defined? And are there real-world examples to look to, to see AI in action? Richard Zane, MD CIO of UCHealth, who was part of an MIS panel titled AI in Healthcare: The Big Questions that Still Need to Be Answered sat down with MD+DI to discuss all things AI.
A few hours removed from his appearance on the MIS panel, Zane said that AI isn’t reinventing the wheel, but rather enhancing what is already there so it can be more efficient.
“People get a little bit freaked out when you talk about artificial intelligence,” Zane told MD+DI. “At the core, AI is just pattern recognition. I really like to describe it as clinical decision support. It’s supporting providers to make better decisions. Whether that’s procedurally, but the needle here or knife there … or actual decisions on what medicines to give, we’re helping to make a diagnosis.”
He added, “I think that the practice of medicine is pattern recognition. A surgeon who’s done aortic valve replacement 600 times theoretically is better than someone who has done it 12 times. The same is true about decision-making. The more decisions you make the better you get at decision-making. It’s the same thing with AI. It’s not differentiated and it shouldn’t be scary …”
Zane said UC Health was already using AI – but he said to understand how people need to first settle on it is defined.
“It depends on how we’re going to agree to define it,” he said. “If we’re simply going to agree to define AI as some type of electronic interface that either supports decisions or interfaces with a patient, we’re using it all over the place.”
UC Health has a solution aimed at patients that can provide directions; schedule appointments and do a variety of tasks that don’t involve medical decision-making.
“At the other end of the spectrum is actual at the point-of-care, helping doctors to make better decisions. Where we have started is embedding accepted guidelines into workflow and decisions. We bring commonly agreed-upon guidelines into the workflow based on very specific patient characteristics so it serves it up to the provider. It doesn’t make decisions for the provider, but it serves up information and serves up different variables from which the provider can choose.”
AI is playing a big role in the medical device and diagnostics industry. Companies are either making acquisitions; forming collaborations, or developing technologies to get into the space.
For example, late last week Medtronic and its partner Cosmo Pharmaceuticals launched a product both called the first AI-based system used to detect colorectal polyps. The duo debuted the GI Genius Module during United European Gastroenterology Week in Barcelona, Spain.
In June, the Dublin-based company signed another distribution agreement – this time with Viz.Ai to improve stroke outcomes.
Medtronic took a deeper dive into the space last year when it acquired its frequent collaborator Nutrino for an undisclosed amount. Tel Aviv-based Nutrino uses AI for nutrition-related decisions.
Nearly a year ago – Edwards Lifesciences collaborated with Bay Labs to improve the detection of heart disease. The agreement called for the integration of Bay Labs’ EchoMD algorithms into Edwards Lifesciences’ CardioCare quality care navigation platform.
Beta Bionics, a company that won MD+DI Reader’s Choice for Medtech Company of the Year in 2018, is combining AI and continuous glucose monitoring for a better way to manage diabetes. The firm has received a great deal of spotlight and even won an IDE from FDA for its solution. Such a feat shows that the possibility of an AI-driven device for diabetes could quickly become a reality.
A Look to the Future
One of the biggest questions surrounding the space is what will the healthcare landscape look like in the future. Zane answered that question for MD+DI noting that these solutions were only going to grow in scale. But he noted people shouldn’t think that it will happen overnight.
“I firmly believe people are way over-estimating the importance of AI in the short-term and way underestimating the importance of it in the long-term,” Zane said. “In 20 years 85% of medical decision-making is going to be made by machines.”