Wednesday, June 19, 2019

FDA Grapples with Artificial Intelligence & Machine Learning


It’s an interesting time to be alive, no doubt about it.  And it is *especially* an interesting time to be practicing in the healthcare arena.

The world is seeing an incredible increase in systems that are using Artificial Intelligence and Machine Learning to take human beings out of the loop and increase the efficiency that computers bring to many processes that require pattern recognition.  For more than a year I’ve been working with a company called Parallel Dots that is working on a system called Dentistry A.I. that is bringing computer processing power to the realm of helping to read radiographs (x-rays).

The problem with these types of systems is to make sure they truly work as promised and that is the job of the FDA.  The federal agency reviews data and studies submitted by manufacturers and determines the safety and efficacy of them.  However, programs that learn and constantly change/update themselves are way different than a device used, for example, to deliver medications to a patient.

Due to this difference, the FDA is looking at ways to evaluate these new devices and programs that promise to greatly enhance and speed up diagnostics and other parts of the healthcare system.  Recently the FDA put a post on their website where the agency outlined some ideas on how they plan to approve these types of devices and systems.  The post states, in part, that:

Artificial intelligence and machine learning technologies have the potential to transform health care by deriving new and important insights from the vast amount of data generated during the delivery of health care every day. Medical device manufacturers are using these technologies to innovate their products to better assist health care providers and improve patient care. The FDA is considering a total product lifecycle-based regulatory framework for these technologies that would allow for modifications to be made from real-world learning and adaptation, while still ensuring that the safety and effectiveness of the software as a medical device is maintained.

The thoughts and points outlined in the post are interesting and forward thinking.  If you would like to read the entire post, follow this link.

1 comment:

  1. I really like this blog- It is very helpful. Thanks for sharing your blog