The best variant sleuth could be AI

The ideas and innovators shaping health care

Jan 04, 2024 View in browser   By Ruth Reader, Carmen Paun, Daniel Payne and Erin Schumaker PANDEMIC AI tools could help identify dangerous Covid variants earlier, researchers believe. | CDC via AP As new variants of Covid-19 emerge, scientists want to know which strains will become dominant and lead to waves of illness. Artificial intelligence offers a potential model, according to a new paper by researchers at the Massachusetts Institute of Technology and the Hebrew University Hadassah Medical School in Jerusalem. To create a predictive model, the researchers analyzed 9 million SARS-CoV-2 genetic sequences from 30 countries and the characteristics and conditions that led to heavy infection rates. In addition to genetic data, the model uses epidemiological information about how fast other variants spread amid various public health interventions. It found three strong predictors of a dominant variant: — The number of infections a strain causes in its first week relative to the number of times it appears in sequencing among other variants — The number of mutations in the spike protein — The number of weeks since the current dominant variant began circulating After a week of surveillance, the resulting model could detect 72.8 percent of variants that will cause at least 1,000 new cases per million people within three months. Given another week of surveillance, the percentage goes up to 80 percent. Why it matters: The Centers for Disease Control and Prevention uses a method to estimate current levels of circulating variants that requires two to three weeks of processing time. The new method could potentially get ahead of those estimates. Researchers suggest the model’s framework could also be applied to other coronaviruses and the flu. WELCOME TO FUTURE PULSE Boulder, Colo. | Shawn Zeller/POLITICO This is where we explore the ideas and innovators shaping health care. If you’re one of the estimated one in three adults who's grinded their teeth or clenched their jaw, scientists in the U.K. are making something for you: a headband to detect the behavior and vibrate, making you aware of it and helping you relax your jaw muscles, The Independent reports. Share any thoughts, news, tips and feedback with Carmen Paun at cpaun@politico.com, Daniel Payne at dpayne@politico.com, Ruth Reader at rreader@politico.com or Erin Schumaker at eschumaker@politico.com. Send tips securely through SecureDrop, Signal, Telegram or WhatsApp. INNOVATORS If tasty food is becoming a little less so, could it be a sign of Alzheimer's risk? A researcher is investigating. | Cheyenne Cohen via AP If a university professor in England is right, people could be using their taste buds a few years from now to gauge their Alzheimer’s risk. You read that right: Professor Alan Chalmers from the University of Warwick in Coventry, U.K., thinks taste could be used to detect someone’s risk of developing the neurodegenerative disease decades before the first symptoms appear. How so? “If you have any form of neurodegeneration — it could be Alzheimer’s, it could be some sort of brain trauma, it could be Parkinson’s — you lose the ability to taste and smell, and this occurs very early in the disease,” Chalmers told Carmen. A visualization professor, Chalmers has worked with private companies, hospitals and the U.K. nonprofit Alzheimer’s Society to create artificial flavors using food-safe chemicals. When testing for Alzheimer’s, people would taste and smell different flavors dispensed through a device and be further screened if they perform poorly in comparison to people of the same age and sex, he said. Chalmers is gathering more data from volunteers who want to test the method, with plans to put it through clinical trials. He hopes the device will become an at-home screening test. Why it matters: If Chalmers’ method works, it could allow people to receive treatment with new drugs that work in only the disease’s early stages. POLICY PUZZLE Gottlieb's urging his former agency to take a cautious approach to regulating AI. | J. Scott Applewhite/AP | AP Photo The Food and Drug Administration should narrow its ambitions when it comes to regulating the new artificial intelligence in health care. That’s according to Dr. Scott Gottlieb, former FDA commissioner in the Trump administration, in JAMA Health Forum, reports POLITICO’s Lauren Gardner. Gottlieb recommends the agency carve out low-risk artificial intelligence devices from regulation to focus its resources on high-risk use cases. Gottlieb, who sits on the boards of health care companies already using AI — Pfizer, Illumina, Tempus and Aetion — suggests the agency consider adopting “authoritative benchmarks” that can be used to help evaluate new technologies. He also argues the agency should review AI devices through a “familiar regulatory pathway.” Why it matters: The debate over how, or whether, to regulate the advanced AI that drug companies already use to speed the development of products is just heating up. The FDA uses its existing rules for medical devices to evaluate some AI products, but much of the latest technology goes unregulated. Last year, the FDA issued guidance and discussion papers on how the agency plans to oversee artificial intelligence in medical products, but it hasn’t issued any new rules. What’s next? In October, President Joe Biden tasked the FDA’s parent agency, the Department of Health and Human Services, with developing a regulatory plan this year.   Follow us on Twitter Carmen Paun @carmenpaun Daniel Payne @_daniel_payne Ruth Reader @RuthReader Erin Schumaker @erinlschumaker   Follow us

To change your alert settings, please log in at https://www.politico.com/_login?base=https%3A%2F%2Fwww.politico.com/settings This email was sent to rouf@idiot.cloudns.cc by: POLITICO, LLC 1000 Wilson Blvd. Arlington, VA, 22209, USA Unsubscribe | Privacy Policy | Terms of Service