Mohamed Havez spends a lot of time examining brain scans, looking for signs of illness that could lead to stroke or dementia.
That’s a big job. The tell-tale signs the UT Health San Antonio researchers are looking for could be hiding among her more than 500 similar tiny spaces on her MRI of a middle-aged person’s brain.
“Think of a neuroradiologist sitting down and trying to count them all,” said Harves, director of the Glenn Biggs Institute for Alzheimer’s and Neurodegenerative Diseases at the health center. “It doesn’t really happen. Each scan can take him an hour or two, or more, which is impossible in a busy clinic workflow. is.”
To speed up diagnosis, he developed an artificial intelligence-powered device that counts lesions in seconds.
Harves said these symptoms are “very difficult to quantify without AI.”
Related: UT Health San Antonio supports pioneering use of AI-powered diabetes management
His research highlights the use of artificial intelligence in San Antonio healthcare as academic institutions, nonprofits, and private healthcare facilities ramp up investments in technology that can assist with tasks ranging from tedious digital paperwork to diagnostics. This is just one example of how usage is accelerating.
Leaders of major healthcare organizations say they support the rise of AI as a health care companion, but it will not be able to accurately make notes in electronic medical records, scan medical images, and diagnose and treat patients. He stresses that humans are still needed to reaffirm the capabilities of supporting technology.
University Health System’s efforts also include a pilot program that uses AI voice recognition to transcribe patient notes into electronic medical records, according to senior vice president and chief information officer Bill Phillips. That’s what I’m talking about.
human factor
Health systems are starting to use AI chatbots to process data such as billing information, he said. The company is also looking at other types of AI, such as connecting phones to hospital rooms and allowing patients to book appointments and contact doctors.
“I don’t think it’s affecting any jobs,” he said of the potential for increased use of AI to reduce the workforce. “We found that we could handle large workloads faster.”
AI-embedded electronic medical records will help researchers and doctors “predict whether a patient will eventually be hospitalized based on symptoms and outcomes,” Phillips said. However, human staff and doctors are expected to recheck the AI and make the final decisions.
“Here, doctors and patients just have a conversation, and AI captures the key elements and creates a graph for the provider,” he said. “AI will always have a human element, so the job of the provider will be to review what the AI system documents and make sure it is accurate.”
At the Methodist Healthcare System, doctors are using AI to detect blockages in brain arteries that can cause strokes. They also use the technology to detect pulmonary embolism, a potentially fatal condition.
Related: Rackspace Launches Effort To Train Thousands Of Employees On Artificial Intelligence
“AI in healthcare is an essential capability to quickly identify potentially life-threatening health conditions,” said Eddie Queller, the longtime chief information officer of the healthcare system. “Methodist Healthcare uses several healthcare AI imaging capabilities to provide the highest quality patient care.”
Earlier this year, San Antonio’s Baptist Health System announced that it uses AI technology to speed up scans by 40%, performing the most common spine and joint imaging in about 12 minutes. The Baptist M&S Orthopedic and Neuroimaging Institute, located on 1604 Ring Road, said it was the first facility in the country to use the technology.
Dr. Carlos Morales, a radiologist at Baptist, said the AI tool uses deep learning techniques trained on large batches of relevant data to create the MRI images. Such scans require the patient to be as still as possible during the scanning process. Movement can blur the image, which is the enemy of any radiologist. However, AI tools are faster, which limits the effects of motion and creates sharper images.
“This technology is a game changer,” he said.
Dr. Rajeev Sri of UT Health’s Long School of Medicine last year participated in a study collecting data from 60 radiologists reading x-rays at a university hospital. By feeding large amounts of data into AI, he hoped it could train the software to mimic human reading patterns and provide a machine-like second opinion.
UT Health’s newest AI tool for identifying brain lesions is not yet in clinical use. But Harves plans to start applying it to patient data next year to aid diagnosis.
“The neuroradiologist’s job will be much easier,” he says. “That’s the diagnosis. But the final decision rests with the radiologist and the doctor.”
Beyond hospital efforts, medical professionals are also prescribing AI-powered medical devices. A local doctor has started using an AI-powered device to regulate blood sugar levels in young people with type 1 diabetes.
Jane Lynch, Ph.D., professor of pediatric endocrinology and diabetes at UT Health San Antonio, who treats children at University Health’s Texas Diabetes Institute, says the device will help young people in central and southern Texas. one of the first doctors to provide
“A dizzying pace”
Despite optimism about the future of AI-powered medicine, researchers and doctors are coming at a time of regulatory uncertainty and may express safety and privacy concerns. He said he expects a range of reactions from patients. We all agreed that when using AI, we need to keep humans under constant surveillance.
Baptiste Health’s Morales said AI tools are continually improving, but it’s “not uncommon for the technology to mark an arterial blockage in a medical image before you realize it’s just a blood vessel. No,’ he said.
“I don’t think there will ever come a time when patients and other doctors will be 100 percent happy with what the machine is producing, even if a human doesn’t say, ‘This is absolutely right,'” he said. “But this will allow us to look at these scans more efficiently. At this point, we can look at things a little faster. I can imagine what it will be like in 10, 20 years.”
Related: Using AI to Help San Antonio Radiologists Reduce X-Ray Reading Errors
“At a dizzying pace, we are capturing electronic medical records that can impact patient history, test results, diagnoses, prescriptions and bills,” said Dr. In our review, we stressed the importance of adopting “validated” technology. About the use of AI in the medical industry. ”
“I still want people to make the final decisions in conversation with patients, rather than letting machines make the final decisions,” he says. “This is just an evolution of what we have, but perhaps we need to be more informed in making these recommendations, so we need to work hard. .”
University Health still uses AI technology to scan medical images. But the company is looking at how the technology could help radiology, predictive analytics, and virtual nursing.
“All these things are happening, but we have to be very careful,” said Phillips, CIO of the university’s health department, while pointing out the pros and cons of new technology, including cybersecurity risks. . “We’re pursuing it, but we’re doing it at a manageable pace.”
Investing in AI
While other industry sectors in San Antonio are adopting or being forced to adopt AI technology, the healthcare sector is starting to take action.
The San Antonio Police Department uses AI to fly drones for large crowd observation, traffic collision analysis, and tactical operations. Rackspace Technology Inc., the city’s largest tech company, is training its employees on AI tools.
Local real estate professionals are using ChatGPT, a popular AI-based chatbot, to describe residential listings and draft financing terms for deals. The non-profit Southwest Research Institute has developed an AI markless motion capture system. The system not only allows pitchers to throw faster without injury, but is a tool to aid clinics looking to detect diseases and conditions such as dementia based on how a person walks.
Related: SwRI develops AI-powered tools to take motion capture analytics out of the lab and into the field
Medical experts say AI deployment must be cost-effective. Because AI research is expensive and time-consuming, it is seeking funding from the National Institutes of Health to develop algorithms and train them on data, UT Health researcher Harves said. But as health systems in San Antonio, Austin, Dallas and Houston scramble to build out the technology, the city’s medical institutions are pouring money into hiring AI experts.
UT Health spokesman Will Sansom said the health system is recruiting “core strengths with promising new research directions, including AI.” He named Dr. Harves. A former adjunct assistant professor at the University of Pennsylvania’s Perelman School of Medicine, he said the health care system was “a prime example” of “embracing faculty with deep learning expertise.”
The city’s medical center declined to disclose the exact amount each medical institution is investing in AI technology.
When asked how much UT Health San Antonio is investing in AI research and technology, Sansom said, “We provide an estimate of the time (resources) spent on AI research, exploration, development, integration, and implementation. That’s not possible at the moment,” he said.
He added that the health system’s annual research portfolio is $360 million and is “moving rapidly” toward $400 million.
Sansom said the Gleehey Childhood Cancer Institute and the Glenn Biggs Institute for Alzheimer’s and Neurodegenerative Diseases at UT Health San Antonio have invested nearly $1 million for a supercomputer that can train AI tools with deep learning models. pointed out that it was spent
A so-called Genie supercomputer could be used to develop new AI tools to better count brain lesions in MRI scans. Sansom said it has a “graphical processing unit, a very important component for the development of deep learning models.” “These GPUs enabled the development and training of AI tools for detecting and counting brain lesions.”
UT Health San Antonio also said it hopes Genie will be studied at the Alzheimer’s Research Center, a National Institute on Aging-designated Center of Excellence. The South Texas ADRC, the only ADRC of its kind in Texas, is a collaboration between UT Health’s Biggs Research Institute and the University of Texas at the Rio Grande Valley.
The idea of using AI for Alzheimer’s-related research was born in May when UT Health San Antonio began building a $100 million patient care and clinical trial facility called the Center for Brain Health. I was. This center he plans to complete in 2025.