Bringing the Power of Artificial Intelligence to Healthcare Research

Artificial intelligence (AI) has the power to help society in many ways, including improving the speed and proficiency of completing routine tasks, while also solving some of the world’s most complex problems.

Anai Kothari, MD, MS, an assistant professor at the Medical College of Wisconsin (MCW) who specializes in surgical oncology, is examining the power of AI to advance research. Newer AI tools can help complete some of the more laborious aspects of research more quickly, he says, so researchers can focus on ideas that generate new knowledge.

Dr. Kothari wrote a paper on the topic that was published in the Journal of the American Medical Association (JAMA). Titled, “Practical Guide to the Use of AI-Enabled Analytics in Research,” it discussed the vast possibilities of using AI to assist researchers in tasks such as writing code, conducting data analysis, and improving efficiency in general.

“The reason we wanted to tackle AI-enabled analytics is that I don't think a lot of people understand or even recognize that you can use newer AI tools, particularly those that are in the generative AI space, to actually augment and assist the research process,” Dr. Kothari says.

He says that large language model programs have pushed the capabilities of AI further because they also allow researchers to use natural language to interact with the data and report findings.

Using large language models and other AI tools can also help generate ideas that can move research past the hypothesis phase.

“Basically, AI can help researchers take their idea and get it off the ground,” Dr. Kothari says.

Using Predictive Analysis to Improve Care

Dr. Kothari first focused on AI during his surgical residency, when he worked with a mentor who specialized in predictive analysis. His training involved using AI to look at data using an advanced analytical approach. His AI work continued during a fellowship in Houston, where he helped to build data platforms during the pandemic.

“We had all this important health information available, and we thought we might be able to gain insight quickly to actually improve how we're responding to patients who are coming into the health system with this new and unknown condition,” Dr. Kothari says.

He continues to build upon that knowledge through his work as the AI Lead at Inception Health, an affiliate of Froedtert and MCW. Inception Health focuses on improving care through a digitally focused strategy. The group also partners with other innovative companies to advance health research and technology.

One of those partnerships, in collaboration with the MCW Cancer Center, is with a company called Triomics, which specializes in AI use for cancer care providers. Dr. Kothari says they utilized AI to help match eligible patients with clinical trials, a previously laborious task.

Controlling for Bias in AI

While AI has the power to transform, it also has the power to increase inequities and be used in harmful ways. Dr. Kothari cautions anyone using AI in research to take careful considerations anytime they plan to use it.

“Data will inevitably have within it the same biases and challenges that a person would have because of what they have been exposed to,” he says. “If we forget about that and apply it, we run the risk of continuing to propagate those same biases.”

He says anytime AI is used in research, authors should outline models and codes, protocols used to acquire data, address privacy concerns, share algorithm data and the methods used to test and validate the model, and that a human should be involved in every step of the process, including testing results.

“You need to keep track of it so there's no misalignment between what it's doing and the original goal,” he says.

Creating Value for Patients

In addition to his other projects, Dr. Kothari recently became the inaugural director of the Bud and Sue Selig Hub for Surgical Data Science, a program created to tackle real-world problems across Wisconsin using data science and AI.

The Selig Hub has already brought together innovators from across the city, including local colleges and technology accelerators, to solve time-sensitive problems like predicting cancer treatment response, identifying patients at high risk for surgical complications, and improving efficiency in the operating room. Unlike other AI efforts, the Selig Hub really is centered on bringing the value of these new technologies to the patients treated close to home.

“I envision a lot of really cool collaborations that come from the Selig Hub because it's citywide,” Dr. Kothari says.

Overall, Dr. Kothari hopes that the power of AI continues to be used responsibly to help improve research and make the world a better place.

“AI has become kind of a catch-all solution to problems, so I feel like there's receptivity to using this technology across a wide array of cases,” he says. “The big hesitancy–and I totally agree with this–is how do you ensure safe and responsible deployment?"