What AI could mean for healthcare

Researchers at the University of South Florida are using AI and virtual reality to study Alzheimer's disease and autism, mapping the brains of mice as they develop. A team from IBM and the Cleveland Clinic published a strategy for using AI to find new targets for immunotherapy. And a new startup is incorporating AI into CRISPR, the revolutionary gene-editing technique, to identify novel gene alterations not already found in nature to expand the possibilities of new treatments.

Meanwhile, a slew of new AI-powered cancer screenings have now been approved by the US Federal Drug Administration and are available for concerned patients, according to CNBC. That follows a trend: about 600 of the 900 AI and machine learning devices and programs approved by the FDA in the past five years have been radiology applications.

And patients could even find that their most ornery doctors are suddenly friendlier. Many report that ChatGPT and other tools have improved their virtual bedside manner over medical chat-based programs.

“It’s hard to overstate the potential that AI breakthroughs could have for nearly every aspect of healthcare, from our understanding of disease and designing new drugs to aspects of healthcare delivery like improving access through more efficient allocation of limited resources,” said Laura Yasaitis, a healthcare consultant for Eurasia Group.

Of course, she has concerns, such as overreliance on AI-generated output, privacy, and exacerbating inequities such as in approving or denying insurance claims.

And, at least for now, Yasaitis thinks that the most promising use of AI is, yes, the boring stuff.

“For every hour of direct patient care, doctors spend around two hours on paperwork during the day, and another one to two hours at night,” she said. “AI applications that can generate much of that content, and then only require review by clinicians, could dramatically reduce that burden.”

While these technologies offer hope for improving healthcare options and extending lives, insurers have yet to catch up. Medicare and private insurers tends not to cover AI-based tests. So when AI makes medical breakthroughs, there might be a delay for those who cannot afford to pay out of pocket.