Only 20% of AI devices for children used pediatric data to train: 3 notes

Just one in five devices approved by the FDA for children actually included pediatric data in their algorithms, a recent Boston Children's study found.

1. The study, published Dec. 16 in JAMA Pediatrics, analyzed 876 artificial intelligence and machine learning-enabled medical devices. Of those devices, 549 were authorized after 2021 and 853 were authorized through the FDA 510(k) clearance pathway. The most common medical specialties were devices were radiology (76.1%), cardiovascular use (10.4%) and neurology (3.5%).

2. Researchers found that only 17%, or 149 devices, were explicitly labeled for use in children, and of those, 19% explicitly reported development using pediatric datasets. The rest were validated using only adult data or failed to report whether children were studied.

3. Some devices could cause harm if inappropriately extended to children. "For example, a radiology device that screens head CTs and flags those of particular concern may not fit kids, whose anatomy or disease processes are different," Florence Bourgeois, MD, senior associate physician of the division of emergency medicine at Boston Children's, said in a Dec. 18 system news release. "Conditions may not be detected with the same specificity and sensitivity. Measuring adverse effects from these devices will require new approaches."

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.

 

Articles We Think You'll Like

 

Featured Whitepapers

Featured Webinars