Methodist Mansfield Hospital uses AI tools for brain mapping

The recent addition of artificial intelligence to a North Texas hospital has already saved the lives of potential stroke patients. But there are some concerns for healthcare professionals, as well as patients.

By Emmanuel Rivas Valenzuela, KERA NewsAugust 16, 2024 10:15 am, , ,

From KERA News:

Methodist Mansfield Hospital is currently at the forefront of medical technology with its recent integration of Artificial Intelligence (AI) in its new neuro critical care unit.

Methodist Mansfield’s AI technology was developed by the brain imaging analysis software company RAPID-AI. The technology was rolled out at the hospital late last year. It was then carried over into the neuro care unit which opened in February.

In adapting this tool, staff can save a significant amount of time, protecting brain tissue in the process.

“That speed gives us an extra five to 10 minutes shaved off of the arrival time,” Methodist Mansfield President Juan Fresquez said at an Arlington City Council meeting. “Just like a heart attack, when the blood flow is blocked, you start to lose heart muscle. It’s similar with the brain, every minute, every second counts.”

Emmanuel Rivas Valenzuela / KERA News

Methodist Mansfield's AI also is accessible to nurses via a phone application.

Methodist Mansfield utilizes the RAPID-AI technology by running CT scans through the software programmed to flag potential abnormalities. Once the AI finds an abnormality, the medical team is then notified. A neuro-radiologist then determines if there is a need for further medical treatment.

Fresquez would then recount the story of Kristin Booth, 45, who was taken to the hospital after waking up unable to speak or move her right arm.

According to the hospital’s in-house magazine Shine, the mother of five then underwent medical treatment to remove a large vessel blockage. Just two days after detection and treatment, she was cheering at her son’s basketball game.

Clots like the one Booth suffered could result in permanent trouble walking and speaking, as well as paralysis or numbness of the face, arm or leg if left untreated. But thanks to the AI software’s swift detection, staff were able to prevent these lasting effects.

Juan Fresquez is Methodist Mansfield’s president and serves on the Mansfield City Council. Methodist Health System

But, as Fresquez later told KERA, AI technology is not entirely novel in healthcare.

“Artificial intelligence has been around for quite some time,” said Fresquez. “I’ve been in healthcare since the mid-80s and have had an opportunity to witness the early rise of artificial intelligence, which really began with Computer Aided Detection (CAD) and imaging. It’s really an infant version of what we have as far as artificial intelligence today.”

The use of AI has recently become a controversial talking point in recent years. A world where AI takes over traditional human roles has always been portrayed in countless pieces of science fiction media like the Will Smith-led film “I, Robot” or Disney-Pixar’s “Wall-E.” But with the advancements of generative AI tools like ChatGPT and Dall-E, the concern has never been louder.

SAG-AFTRA and WGA, the actors and writers unions, even took measures to prevent studios from using AI to create content during last year’s historic strikes.

But “fear” is not the word Neuro ICU Manager Rachel Cochran would use. She asserts that AI cannot do what trained doctors and radiologists can do, as the technology is not always perfect.

Rachel Cochran is the neuro ICU manager and has been at Methodist-Mansfield since 2016. Emmanuel Rivas Valenzuela / KERA News

“The human physician is always going to be the one who is diagnosing the patient, (AI) is not replacing that person,” said Cochran. “Healthcare is very holistic; we have to consider a lot of different things when we’re considering treating people.”

Jamboor Vishwanatha is the vice president for Minority Health and Health Disparities at the UNT Health Science Center. He believes that, like Cochran, AI will never be able to fully replace humans in the healthcare industry.

“At best, the AI tools will be assistive rather than replacing the professionals,” said Vishwanatha. “The outcome will be better and perhaps the surgery will be done much faster than what’s being done. But at no point I estimate that a robot will take over what a surgeon can do.”

Vishwanatha’s team is currently working on a few national projects on artificial intelligence machine learning to “advance health equity, and researcher diversity.”

Part of that mission includes making sure AI tools are trained in a diverse set of data. A potential drawback of AI falls on what the tool is learning from, according to Vishwanatha.

If data from a single demographic group based on race, ethnicity or sex is collected and subsequently fed into a learning system, it may give doctors and medical professionals skewed information.

“I think that, in surgical care, there is a tremendous opportunity to use AI learning models,” Vishwanatha said. “With that said, I’m also very concerned about how this could affect different communities. Not all our communities get the surgical care that is necessary. So, if surgical AI models are developed based on existing surgical data, it could affect some communities because they’re not represented in that data.”

Fresquez agrees with that sentiment citing an issue plaguing marginalized communities in the 80s.

Jamboor Vishwanatha, Ph. D. is the vice president for Minority Health and Health Disparities at the UNT-Health Science Center in Fort Worth. UNT-Health Science Center at Fort Worth

“You want to have databases that are inclusive of all nationalities and genders,” Fresquez said. “Looking back into the early days of imaging, Computer Aided Detection, or CAD, that was used predominantly in mammography. And there were definite differences in incidence in cancer so it’s not detection, across the different nationalities and ethnic groups that were being assessed.”

Those cancer databases were then updated to better serve black and brown communities. When hospitals first collected data, only white affluent women were able to afford treatments.

Fresquez says Methodist Mansfield is committed to contributing diverse results to these databases.

“The more data that you collect, the better,” said Fresquez. “The size of the database really closes the gap in the standard deviation. So, the more data points you have the sharper your data and your outcome reporting will be.”

RAPID-AI is currently being used in over 2,000 healthcare facilities in over 100 countries which will help in the diverse collection of data.

KERA has reached out to RAPID-AI about their data collection methods but has not received a statement at the time of writing.

Now, Methodist Mansfield is only using AI in their Neuro ICU center. But Fresquez says the hospital is more than willing to utilize AI tools in the future.

“We want to make sure we don’t wait too long to implement change,” said Fresquez. “I wouldn’t have thought this technology would be available. We’re fortunate that I think we’re the first ones in our health system to actually implement it. I’m very optimistic and excited about what the future holds.”

If you found the reporting above valuable, please consider making a donation to support it here. Your gift helps pay for everything you find on texasstandard.org and KERAnews.org. Thanks for donating today.

Most viewed right now