FDA Using AI in Drug Surveillance; Health Systems Move Ahead With AI; RAISE Guidance

Welcome to MedAI Roundup, highlighting the latest news and research in healthcare-related artificial intelligence each month.

The FDA said it’s modernizing its drug surveillance efforts by using AI to help sift through the more than 2 million adverse event reports it receives each year. (Endpoint News)

AI industry leaders are worried that policymakers are “behind the times” with AI in healthcare, which could lead to regulation that fails to protect patients. (KFF Health News)

While policymakers and regulators catch up, health systems across the country have been figuring out when and how to implement the emerging technology without compromising patient safety or privacy, according to a STAT investigation.

Following the Responsible AI for Social and Ethical Healthcare (RAISE) conference, experts published guidance for the development of healthcare AI tools simultaneously in Nature Medicine and NEJM AI.

A survey of more than 1,000 physicians revealed that 93% feel burned out on a regular basis, but 83% believe AI-based tools could eventually alleviate many of the problems facing healthcare, according to Athenahealth’s third Physician Sentiment Survey.

Mass General Brigham pediatricians used GPT-4 to translate educational videos on pediatric care — including instructions for surgical procedures and best practices for intubation and intensive care — into Spanish so they could be used by clinicians in low-resource areas of Guatemala and Colombia, they reported in Frontiers.

The NIH National Center for Advancing Translational Sciences (NCATS) announced a new AI tool called the Biomedical Data Translator that integrates machine learning tools with large biomedical databases to identify existing treatments that can be repurposed for rare diseases.

AI can quickly and accurately measure fat around a patient’s heart during routine low-dose CT scans, potentially helping clinicians identify patients at risk of a heart attack, according to a study published in NPJ Digital Medicine.

A study of 177 individuals — including 78 with Alzheimer’s disease — identified a new set of 21 diagnostic metabolomic biomarkers that could help train new AI tools to predict patients at higher risk of Alzheimer’s disease, according to a paper published in the Journal of the Neurological Sciences.

Please enable JavaScript to view the

comments powered by Disqus.