| | | | |  | | By Daniel Payne, Ruth Reader and Erin Schumaker | | | | | 
It's even harder to get into Harvard Med's AI doctoral program than the school itself. | Maddie Meyer/Getty Images | Harvard Medical School is bringing artificial intelligence into its curriculum. For the first time this fall, students starting a health science and technology track will take a one-month introductory course on AI in health care. Those students, who often study to become physician-engineers or physician-scientists, will consider the uses of AI and examine the technology’s diagnostic limits in medicine. Tackling the issue from the start of students’ education is unique among medical schools, Dr. Bernard Chang, the school’s dean for medical education, said in an announcement. That approach also alerts students that AI is changing the practice of medicine in significant ways. In addition to the introductory course, Harvard Medical School announced a new Ph.D. track on AI in medicine will begin this fall semester, which Chang attributed in part to student demand. (The program has seven slots and received 400 applications.) The school also offers grants of up to $100,000 for projects looking to innovate medical practice using AI. Why it matters: Health systems are increasingly looking for clinicians who understand AI, and AI education could set the tone for how the tech is used and how much doctors trust it — both crucial to AI’s success in the health sector.
| | | | 
Paris | Jeffrey Diaz | | This is where we explore the ideas and innovators shaping health care. There are microplastic particles in dolphin breath, researchers have found. It suggests the dolphins inhale the contaminants when they come up for air. 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.
| | | |  Hospitals are using both AI and other tools to evaluate patients, providing for ready comparisons. | AFP via Getty Images | Artificial intelligence tools that predict whether a patient is about to take a potentially fatal turn are already in use in hospitals nationwide, but they’re not yet clearly better than traditional methods, a new study in JAMA Network Open found. Researchers from the University of Chicago, Yale and other institutions compared six tools — some AI-powered, some not — and gave top marks to eCARTv5, a machine-learning algorithm developed by Dr. Dana Edelson, the study’s principal investigator and executive director of rescue care at the University of Chicago Medical Center. But second place went to the National Early Warning Score developed by Britain’s Royal College of Physicians. It uses physiological findings and patient observation to make predictions. Another model reliant on AI, Epic’s Deterioration Index, performed poorly, the researchers found. The study involved nearly 363,000 patient encounters across seven sites in the Yale New Haven Health system. It included big academic medical centers, teaching hospitals and smaller community health clinics. Why it matters: Neither regulators nor academics agree on how AI should be vetted, and health systems, many strapped for resources, have difficulty identifying useful AI. Edelson’s study is notable for comparing multiple methods of understanding in different settings when a patient might enter a critical state. Most interesting of all, she and her colleagues found that simple calculations can outperform AI. “Just because it’s AI doesn’t mean it is a better tool,” said Edelson. “Health systems may not be able to do these analyses themselves, and that’s why we publish this stuff.”
| | |  During the next pandemic, the haves need to share with the have-nots, an international panel says. | AP | The world is still “acutely vulnerable” to another pandemic. That’s according to the latest report from the Global Preparedness Monitoring Board, a group of public health specialists first convened by the World Health Organization and the World Bank in response to the Ebola epidemic in Africa. “We all, including heads of state, need to abandon the idea that pandemics are rare, once-in-a-century events,” Kolinda Grabar-Kitarović, co-chair of the monitoring board, said at a press conference announcing the report. “In reality, the risk of an epidemic is ever present, and the frequency of outbreaks has increased significantly in this century.” Grabar-Kitarović, a former president of Croatia, recommended building “resilient systems” to respond to health threats instead of focusing on specific pathogens, our Giedre Peseckyte reports. The board also underscored the importance of international cooperation as evidenced by the lessons learned from the Covid-19 pandemic and the recent outbreaks of mpox. “If the reason for not sharing the technology is you want to put your own people first, I think that is very unfortunate and myopic … the disease is going to come back and bite you still,” said Joy Phumaphi, a board co-chair and Botswana’s former health minister. What’s next? An alliance of health activist groups said the diplomats negotiating an agreement under WHO auspices to set national obligations during the next pandemic should consider the report. Mohga Kamal-Yanni, policy co-lead for the People’s Medicines Alliance, said the accord should include binding commitments to share resources. | | | | Follow us on Twitter | | | | Follow us | | | | |
No comments:
Post a Comment