Artificial intelligence may have already been used to diagnose you

By Heather Close Published February 6, 2024 FEATURED, HEALTH BUSINESS, YOUR HEALTH

Kentucky Health News

As Congress and federal regulators begin to take steps to monitor and control the use of artificial intelligence, “The technology has already been put to use in unseen and unregulated ways in health care,” Ben Leonard and Chelsea Cirruzzo report for Politico Pulse, drawing on a Politico story by Daniel Payne and Ruth Reader on how A.I. is affecting Americans. Here are some ways:

Medical imaging: A.I. is is commonly used in radiology, to help “decipher medical images like X-rays — and some systems work on par with humans, according to recent studies,” Politico reports. “The algorithms also measure and improve radiologists’ performance.”

Cancer: The progression of a cancer can be hard to evaluate, and A.I. can help, especially with prostate cancer, in which “tumors grow in multiple locations and extend in ways that might not get picked up on imaging,” Politico reports.

Record keeping: “A.I. systems help doctors make better sense of their notes. Many health-tech companies target ‘low-risk, high-reward projects’ that can summarize information and act more like a secretary than a co-pilot.

Billing and coverage: “Firms are toying with technology that can change how bills are generated and processed. . . . A.I. can scrutinize bills more quickly than humans, leading to potential labor and time savings.” Advocates say it will make claims move faster, “but skeptics worry AI could reduce transparency and accountability in decision-making” by health-care providers and insurers.”

The Food and Drug Administration “has taken an experimental approach to regulating A.I. in medical devices, mostly issuing strategic plans and policy guidance,” Politico reports. “The biggest challenge is that A.I. evolves over time, and companies don’t want to re-apply for FDA clearance to market their technology.” The Department for Health and Human Services finalized a rule in December that would require more transparency about A.I. in clinical settings.