Radiologists to Develop AI to Perform Routine Medical Tasks, Supplement Staff

A research team from the University of Iowa is developing a device that leverages artificial intelligence (AI) to perform simple radiology tasks and procedures in an effort to supplement staff and limit unnecessary risks for healthcare professionals.

According to the press release, the team plans to use the tool for interventional radiology (IR), which is focused on treating patients with minimally invasive techniques and technologies, such as guide wires, catheters, and other devices.

An ideal device to support IR staff and protect them from risks, such as unnecessary exposure to COVID-19, would utilize logic and reasoning similar to those that clinicians use to determine how to treat patients with simple ailments, the researchers posited. To create such a device, the researchers are developing complex AI algorithms.

“So, the idea is to have a device perform simpler medical tasks without involving me or another team member, that way we can focus on more complex tasks,” explained Sandeep Laroia, MD, the interventional radiologist leading the project, in the press release.

For example, the IR procedure used to treat ascites, a common condition where excess fluid builds up around the spaces of the abdomen, is relatively simple. Still, patients must visit the doctor often, up to once or twice a week in some cases, while the condition persists, he added.

According to Laroia, with a device that could perform the procedure on its own or only requires that a clinician start and finish the procedure, clinicians could focus on more pressing tasks.

In addition to supplementing an IR team during staffing challenges, the device also has the potential to protect care teams by reducing unnecessary risks.

“We need that human healing touch, but it’s becoming increasingly complicated to get providers to the patient for a variety of reasons,” Laroia said. “In a pandemic, for example, we want to limit exposure, so we can keep ourselves and the patient healthy, for the sake of protecting our staff, but also because we need to be there for our patients.”

The device could also benefit patients living in rural areas who may struggle with access to care, the press release states.

“Sometimes people travel several hours to get simple procedures done,” Laroia noted. “So, the question is how can we decentralize some of these treatments so that they’re more readily accessible and not just in more urban areas?”

One potential solution is leveraging an AI-based device, like the one his team is developing, for remote patient monitoring and care, he stated.

“One of the challenges we continue to face is bringing advanced treatments to the patients,” he said. “We’re looking for ways to still deliver our high-quality care, but in patients’ neighborhoods.”

This research highlights a growing interest in assistive technologies and support devices in radiology.

In 2019, interventional radiologists at the University of Washington leveraged virtual reality (VR) in combination with 3D images from CT angiography scans of patient blood vessels to speed catheter placement, which has the potential to make IR treatments more efficient, safe, and accessible.