Tuesday, Feb. 18, 2025

AdvertiseDonateSubmit
NewsSportsArtsOpinionThe QuadPhotoVideoIllustrationsCartoonsGraphicsThe StackPRIMEEnterpriseInteractivesPodcastsGamesClassifiedsPrint issues

IN THE NEWS:

Black History Month 2025

UC Regents committee evaluates implications of AI in health care

The Ronald Reagan UCLA Medical Center is pictured. During the UC Board of Regents’ July meeting, the Public Engagement and Development Committee discussed the use of artificial intelligence in health care. (Daily Bruin file photo)

By Dylan Winward

July 22, 2024 8:39 p.m.

This post was updated July 26 at 1:36 a.m.

The UC Regents Public Engagement and Development Committee discussed the use of artificial intelligence in health care Wednesday.

Three leading data scientists answered questions from the UC Board of Regents about how the UC Health system uses artificial intelligence, the implications of AI usage on underrepresented communities and oversight for artificial intelligence usage. The panel also discussed funding for research relating to AI and health care.

Dr. Atul Butte, the chief data scientist at UC Health, said an increasing number of AIs are training using clinical data which has been anonymized following approvals from the Food and Drug Administration. He added that existing algorithms developed within the UC system predict intestinal lung disease, immunodeficiency and Covid-19 patient reactions.

Butte, a professor at UC San Francisco, said UC Health is also responsible for systemwide governance of AI use in health care and has been involved in briefing Congress on the ramifications of artificial intelligence usage.

“It’s not just going to be about the physicians. It’s also going to be about the patients, the nurses and all of our clinical disciplines,” Butte said.

Dr. Christopher Longhurst, the chief clinical and innovation officer at UC San Diego Health, said AI is being used to help clinicians draft messages to patients for virtual care. Within the trial, the AI generated messages would then be edited by physicians before being sent, he added.

Patients participating in the study were told that the responses were partly generated by AI, Longhurst said. He added that trials found the AI would help create longer messages to patients, which were made to seem more empathetic in nature than ones written solely by physicians.

“Many of our patients told us they know that our physicians are busy. They’re glad that they have a co-pilot helping them out, but also glad that the messages are being reviewed by a doctor before they leave,” Longhurst said.

Although the study found that the tool did not save significant time for doctors, it did help reduce their cognitive load, Longhurst said, adding that the health system is now expanding the pilot into dozens of new clinics and hundreds of additional clinicians.

Janet Reilly, the chair of the UC Board of Regents, asked the panelists how the UC as a public university plans to stay at the cutting edge of AI despite gaps in funding.

Dr. Matthew Lungren, the chief data science officer for health and life sciences at Microsoft, said during the panel that one way the University can ensure continuous funding is through public-private partnerships. After Regent Alfonso Salazar asked the panel about what they were doing to protect patient privacy within public-private partnerships, Butte said privacy contracts are drawn up on a case-by-case basis.

Longhurst said existing funding for AI within UC Health is being prioritized to areas where there are existing opportunities to improve, such as in postoperative sepsis treatment. Butte said the UC is also trying to obtain more funding from the National Institutes of Health.

The systemwide integration of AI into health care can help eliminate biases in healthcare, Butte said. Student Regent-Designate Sonya Brooks asked the panel to elaborate on UC Health’s commitment to using AI in a responsible way when it comes to underrepresented communities.

Butte replied by saying the University is working to increase study participation by rural patients, including through the UC Davis Health system.

Longhurst also said the use of AI in remote patient monitoring can increase access to preventative care, an important aspect of increasing healthcare access. One potential implication of the AI rollout is that it will increase access to affordable healthcare, Butte said.

“We can learn what we do at UC (with) our best doctors, our best nurses, treat patients, create those AI models, and then put those AI models wherever we can, so that that level of care is delivered, not necessarily here in Mission Bay at UCSF but in all of our partners around the world,” Butte said.

However, Mayté Frías – the staff advisor-designate to the Regents – said she was skeptical about the ability of AI models to avoid negatively impacting health equity.

“This is moving really quickly, and at the same time, we’ve been carrying systems of oppression for centuries, and we can’t quite figure that out,” she said. “I am very suspicious that with AI, we’re going to be able to account for all of that.”

Share this story:FacebookTwitterRedditEmail
Dylan Winward | News editor
Winward is the 2024-2025 News editor and an Arts, Copy, Photo, PRIME and Sports contributor. He was previously the 2023-2024 features and student life editor. Winward is a third-year English and statistics student from London in the United Kingdom.
Winward is the 2024-2025 News editor and an Arts, Copy, Photo, PRIME and Sports contributor. He was previously the 2023-2024 features and student life editor. Winward is a third-year English and statistics student from London in the United Kingdom.
COMMENTS
Featured Classifieds
More classifieds »
Related Posts