AI leads to better outcomes for patients

Imagine you just had surgery. You look down at your surgical wound, and you are unsure if it is infected or not or if you should call your doctor or go to the hospital. Soon, your smartphone might be able to help you with this decision. The MUSC Harvey and Marcia Schiller Surgical Innovation Center is working on a project that would allow patients to take a picture of a wound, then have artificial intelligence (AI) analyze the wound and recommend the best care.

This innovative center was established through a philanthropic gift to support the application of technological advances, like AI and machine learning (ML), in health care. ML is a type of AI that learns from data, thereby allowing it to continue to improve its predictive abilities.

The connection between AI and machine learning and health care may not be obvious at first. But to Arman Kilic, M.D., director of the Schiller Surgical Innovation Center, AI is the key to better outcomes for patients by, as he explained, helping doctors to make more optimal care decisions and by helping health systems to operate more efficiently.

“The center’s goal is to unite all the various innovation efforts that are happening within the department,” said Kilic. “A large focus of that is AI and ML. We have probably somewhere in the range of 20 to 25 different AI projects that are ongoing right now in the surgical department through the Innovation Center with a wide variety of domains like surgical oncology, trauma, chest wall surgery, transplantation, cardiac surgery and vascular surgery.”

One of Kilic’s key projects involves modeling the risk of cardiac surgical procedures. He said that currently, most doctors are working from clinical risk models based on limited data and drawing solely on prior experience without data-driven insights. “A lot of the clinical risk models done in health care have only modest performances,” Kilic said. “That's because most of them either use limited data or apply relatively primitive statistical approaches.”

Kilic says AI can be much quicker and more accurate when making data-driven recommendations. That’s because AI can analyze a large amount of data in seconds and provide physicians with a complex risk model that takes all available information into account.

"Even if you're using the same data set, machine learning algorithms can often improve predictive performance compared with traditional approaches.” Kilic believes that AI’s main role in health care will come in the form of augmented or assistive technology. Kilic explained that MUSC is among a small group of academic medical centers in the U.S. with dedicated spaces to advance surgical innovation. In addition to the center’s work in AI and ML, it serves as a hub to centralize the department’s Human-Centered Design program and investigator-initiated clinical trials, allowing for shared resources as well as greater collaboration and expertise to advance surgical innovation.

Today, with ongoing challenges in staffing and supply chain issues that are becoming increasingly more burdensome in health care systems, Kilic said AI can help by automating many processes in the hospital to improve efficiency and allowing physicians, nurses and other care providers to spend their time on other tasks. But Kilic is quick to point out that this will not replace critical health care professionals, more, he said, it should be seen as a resource for those working in the field – a way to help them to do their jobs better. Kilic said it’s crucial for medical professionals to embrace this type of technology, as it will have an expanding role in health care.

“We will ultimately be the gatekeepers to this technology and guide when and how it is used,” Kilic said. “I don't think we should shy away from it because in my mind, it's the same thing as a new device or a new pharmaceutical. Innovation is an inherent part of what we do in medicine.”