At University Health, AI is used to improve mammography, magnetic resonance imaging, or MRI, and computed tomography, or CT scans, Bill Phillips, executive vice president and chief operating officer said.
What’s happening?
One of the key areas being advanced with AI is the ability to more accurately read mammograms, allowing specialists to find potential issues sooner, Phillips said.
“It is producing higher accuracy rates, better predictability and focusing. Because if you look at a mammogram, it is a pretty large area that you have to read where [AI] really shows density changes in tissue and highlights it,” he said.
Alongside improvements to mammograms, AI helps speed up MRIs to be at least three times faster, allowing more patients to be seen in a day.
Utilizing licensed software, the clarity of the MRI can be improved.
Phillips said an abdominal MRI that usually requires a nine to 14-second breath hold time could be reduced down to four seconds, which reduces patient stress.
While AI can make mistakes, Phillips said the systems used for health care are more accurate due to information pulled from a specialized factual data set.
“Ultimately, radiologists still have to look at it and sign off on it,” he said. “But we are finding it extremely accurate.”
Looking ahead
AI is still new in the industry, improving to cover more specialties each year.
Projects that are being developed include advancements in pathology and electrocardiogram, or EKG, results.
Various organizations and health care leaders are developing AI, including local researchers at The University of Texas at San Antonio, who are developing improvements for how clinicians handle biomedical and social data.
This development is funded through the National Institutes of Health, which created a program to support AI-powered advancements in the health care industry.
Phillips said the rate of AI advancements is increasing each year as the technology is better understood.
For University Health, AI improvements are integrated at all hospitals in the region, and will be implemented at new hospitals, including University Health Retama Hospital, upon completion.
Phillips said that by the time new AI is implemented, it is vetted to ensure accuracy and benefits to both patients and specialists.
“You have to vet the good from the bad and look at the real meaningful impact of it, and are you really getting better patient outcomes with it,” he said.
