Panel: Healthcare must engage with, understand fast-developing AI field 

Healthcare organizations and clinicians must learn to use artificial intelligence, experts said at a Tuesday panel in Milwaukee.

The Northwestern Mutual Data Science Institute, a partnership between Northwestern Mutual, Marquette University and the University of Wisconsin-Milwaukee, hosted a summit on AI, including its use in healthcare.

Alex Baenen, senior director of surgical and procedural digital ecosystem at GE HealthCare, said medical device makers without a digital or AI strategy are not “future-proofing” their work. Those looking to buy expensive equipment want to know how it’ll improve clinical outcomes, make them more efficient and keep them current.

“AI is moving so quickly,” he said. “All of us have to be able to speak the language.”

Alla Woodson, general manager of anesthesia and global services for patient care solutions at GE HealthCare, said engineers and product managers need to develop their products with the help of clinicians to ensure they solve real problems.

That process also builds trust in the technology, especially when workers fear AI may displace them.

“Many people ask the question, ‘Will we be replaced by AI?’” she said. “The answer is no. People who do not use AI will be replaced by people who do.”

Dr. Anai Kothari, assistant professor at the Medical College of Wisconsin’s surgical oncology division, said a few years ago he maybe wouldn’t have thought AI competency would be needed to function well in healthcare.

“Today, I think it’s an absolute necessity,” he said.

AI tools are widely available. The key difference makers are the data sets used for algorithms and having the skills to analyze them.

“That’s a huge competency that’s needed,” said Dr. Taly Gilat Schmidt, director of advanced development at Onc.AI, a digital health company focused on oncology decision-making.

Dr. Joel Weiner, vice president and medical director at Northwestern Mutual, said AI, “in its simplest definition,” is a computer doing something that a human does.

“AI, in my opinion, will definitely not replace humans in healthcare,” Weiner said. “I think it will assist humans, and that level of assistance will become greater over time.”

But Dr. Keith Dookeran, an associate visiting professor of epidemiology at the University of Wisconsin-Milwaukee, said that situation “depends.”

“What about situations where you don’t have physicians?” he asked. “All you might have is AI.”

Dookeran said AI could “democratize” care. But he said algorithms depend on access to data, and many people cannot get access to hospital-based data. That could lead to problems with bias, as data could be bad or not generalizable.

“What about quality and fairness with the use of AI?” he said. “Does your data set lead you to make the wrong conclusion?”

This article first appeared in the Wisconsin Health News daily email newsletter. Sign up for your free trial here.