A recent report details the steady growth of AI and machine learning.
Artificial intelligence is a growing trend in the world of life sciences industry. According to a recent report, the number of medical devices that use AI or machine learning (ML) in some capacity approved by the FDA continues to increase.
The report also looks into the methods that companies go about getting these approvals, and what sorts of approvals they’re attempting to receive. As more and more devices are approved, clear trends are starting to emerge.
Antoine Manson, regulatory policy and innovation specialist at Bayer AG, authored the report A Review of Artificial Intelligence and Machine Learning–Enabled Medical Devices Approvals in the US from 1995 to 2022, which was presented at the European DIA 2023 conference. In it, he reviews FDA AI/ML approvals, along with the most common approval pathways, the average approval time, what sorts of machines are being approved, and what area is seeing the most approvals.
According to his results, the vast majority of these devices are approved under the 501(k) pathway, which suggests that most devices bear similarities to previously approved devices. Most devices were Class II devices and there is an average of 152 days for review. Devices that used alternative pathways, such as De Novo or PMA, have longer wait times that reached up to 370 days.
The report also details a yearly increase of 50% for the number of approved devices. This significant increase in approvals began in 2015 and does not appear to be slowing down. In 2015, five devices with AI/ML capabilities were approved. By 2021, that number has quickly, but steadily, grown to 115.
Out of the devices researched for the report, the vast majority are used for radiology, with 392 devices receiving approval. Devices used in the cardiovascular field come in second with 57 devices.