Skip to main content

News

AI Diagnostic Software Secures FDA Green Light

By Steven Porter  
   May 29, 2018

An official touted the 'tremendous potential' of AI-powered tools in clinical settings, while cautioning that such tools should not supplant human review and judgment.

The Food & Drug Administration gave a software company the go-ahead to market a diagnostic program powered by artificial intelligence.

The software is designed to detect wrist fractures in adults, according to the FDA's announcement last week, which also emphasized that AI should not be used as a replacement for clinical review or judgment.

"Artificial intelligence algorithms have tremendous potential to help health care providers diagnose and treat medical conditions," said Robert Ochs, PhD, acting deputy director for radiological health for the Office of In Vitro Diagnostics and Radiological Health in the FDA's Center for Devices and Radiological Health, in a statement.

  • What is it? OsteoDetect is a computer-aided program that uses an algorithm to review X-ray images for indications of a common type of wrist fracture, according to the FDA's announcement.
     
  • How'd it win approval? The company proved its product by submitting (1) a study of its performance against that of three board-certified orthopedic hand surgeons in a retrospective study of 1,000 radiograph images, and (2) a study of 200 patient cases reviewed by 24 providers. The FDA determined both studies supported the software's claimed benefit.
     
  • Low-risk approval process: OsteoDetect, which was submitted by Imagen, secured approval through the de novo premarket review pathway, which is used for new devices deemed low- or moderate-risk.
     
  • More on machine learning: OsteoDetect relies on an approach to software that's gaining steam in healthcare applications, making it easier for doctors to identify ailments and predict outcomes with advance notice, as The Wall Street Journal reported earlier this month.
     
  • Building on a solid foundation: In order for AI tools to live up to their potential, they need solid building blocks. That's a riddle researchers are actively working to solve, and they seem to be finding success, as Alexandra Wilson Pecci wrote earlier this year for HealthLeaders Media.

For more on AI in healthcare, be sure to revisit the September edition of HealthLeaders magazine.

Steven Porter is editor at HealthLeaders.

Tagged Under:


Get the latest on healthcare leadership in your inbox.