The researchers are urging the FDA and industry to enhance the credibility of device authorization by conducting clinical validation studies on these technologies and making the results publicly available. They also found that the latest FDA guidance does not clearly distinguish between different types of clinical validation studies in its recommendations to manufacturers. The team has proposed definitions for clinical validation methods to be used as a standard in the field of medical AI.
Key takeaways:
- A multi-institutional team of researchers analyzed clinical validation data for over 500 medical AI devices and found that about half of the tools authorized by the U.S. Food and Drug Administration (FDA) lacked reported clinical validation data.
- Since 2016, the average number of medical AI device authorizations by the FDA per year has increased from two to 69, indicating significant growth in the commercialization of AI medical technologies.
- Of the 521 device authorizations, 226 or approximately 43%, lacked published clinical validation data, with some devices using 'phantom images' or computer-generated images that did not technically meet the requirements for clinical validation.
- The researchers recommend that the FDA and device manufacturers should clearly distinguish between different types of clinical validation studies in its recommendations to manufacturers, and they have laid out definitions for the clinical validation methods which can be used as a standard in the field of medical AI.