Five Takeaways From the FDA’s List of AI-Enabled Medical Devices

The Food and Drug Administration (FDA) has authorized 91 AI/ML medical devices in 2022 alone, according to data released on October 5. Some of the devices that received FDA clearance this year include ‘an atrial fibrillation history feature for the Apple Watch, and Tel Aviv-based startup Aidoc’s feature to analyze X-ray exams to detect and triage collapsed lungs.’

As the number of devices seeking FDA approval is constantly increasing, the agency is looking to adapt its regulatory framework to the new technology.

A report in MedTech Dive analyzed FDA data on AI/ML devices and summarized five key takeaways:

1. Number of AI/ML enabled medical devices reviewed by the FDA more than doubled from 2017 to 2018

In 2021, the FDA authorized a record 115 submissions, an 83% increase from 2018. 

Earlier this year, the FDA’s Centre for Devices and Radiological Health issued the final results of its Software Precertification Pilot Program. The pilot program explored a new approach to regulating digital health software technologies based on regulating companies’ internal processes, cultural and organizational standards, instead of the individual apps or devices.

2. More AI tools for radiology than for any other specialty

More than three-quarter of the submissions the FDA has authorized to date, have been in radiology and 11% in cardiology. One of the reasons could be the wide array of data available for device developers to draw on from imaging and electrocardiograms.

3. Till date, 96% of authorized AI/ML enabled devices have 510(k) clearance

Even though FDA-approved devices form just a fraction of the healthcare devices, till date, 96% of authorized AI/ML medical devices have 510(k) clearance. Only three devices have gone through the FDA’s more rigorous premarket approval process.

Most of these devices don’t provide a medical diagnosis but offer suggestions or recommendations to clinicians or patients.   

4. GE Healthcare and Siemens leading the AI/ML device market 

Currently, GE Healthcare has 42 FDA-cleared AI/ML enabled medical devices, mostly in radiology, while Siemens Healthineers has 29 cleared devices.

GE Healthcare’s devices include an ‘algorithm to help clinicians detect collapsed lung cases’ and Siemens’s include a ‘feature to help with quick quantification and accurate visualization of calcified coronary lesions from CT scans.’ 

5. Algorithms to get increasingly complex and regulators are looking to adapt

 With the increasing challenges and demands, AI/ML devices will become increasingly complex as more and more companies would want their algorithms to “learn” and adapt than simply be “locked and deployed”. The FDA had taken some steps to keep up with such changes, including its Software Precertification Pilot Program. 

The new guidance provides clarity on the FDA’s use of enforcement discretion for Software as a Medical Device (SaMD), mobile medical apps and categories of clinical decision support software. These regulatory support tools provide developers of these software with a framework for assessing whether or not their technology is subject to FDA regulation as a medical device.