FDA Offers Glimpse into New Work in Device Cybersecurity, 3D Printing, Artificial Intelligence
Posted 26 March 2018 | By
Officials from US Food and Drug Administration (FDA) revealed peeks into new work items in the areas of medical device cybersecurity, 3D-printing and artificial intelligence at the Association for the Advancement of Medical Instrumentation (AAMI) conference last Friday.
FDA’s Center for Devices and Radiological Health (CDRH) is working to create new industry policies and regulations, as well as to further develop existing ones for each of the three topics.
Suzanne Schwartz, CDRH associate director for science and strategic partnerships, said at the AAMI conference that the agency has had discussions with manufacturers and health care providers, among other stakeholders, on certain limitations in the FDA cybersecurity policies set forth in guidance issued in 2014
on medical device design and development recommendations.
“There is room for further iteration, further refinement and understanding” on what FDA is looking to see in terms of ensuring effective cybersecurity management, Schwartz said. She added CDRH has “a plan in place” to update the premarket guidance to “better bridge what FDA is looking to see by way of analysis, risk assessments, design, documentation and how to organize all of this.”
She also highlighted the first medical device development tool to be FDA qualified – the 23-item Kansas City Cardiomyopathy Questionnaire (KCCQ)
for measuring self-reported outcomes from patients with congestive heart failure. CDRH plans to qualify many more of these tools by using different use cases, she said.
At the AAMI session on additive, or 3D, manufacturing, CDRH staff pointed to recent actions in the area, such as the 2017 finalized guidance
, the more than 100 510(k)s cleared for devices designed using 3D-printers, and engaging a new working group, which consists of both internal and external experts.
“3D-printing is an interesting area because it’s a bit different from what FDA has typically done,” said James Coburn, a senior researcher at CDRH. “Patient-specific devices is a huge area” of new opportunities within 3D-printing because an “infinite number of shapes” can be created at a lower cost for industry.
Coburn also pointed to a new CDRH work area – anatomical models. These are not medical devices when they are merely intended to help patients understand a surgical procedure, he said, but they become medical devices once they are used by surgeons to improve their surgical planning.
Hospitals across the US have already begun 3D-printing anatomical models and providing these to their surgeons, he said. This has prompted FDA to begin thinking about how these 3D-printed models should be considered from a regulatory point of view, according to Coburn.
Another area that is relatively new within the regulatory landscape is that of artificial intelligence.
CDRH Senior Biomedical Research Scientist Berkman Sahiner updated conference attendees on the specific issues FDA is focusing on, including the need for continuous learning.
“FDA does not have a formal policy yet for continuous learning systems,” he said. “But we would be able to define some of those conditions or requirements with input from all stakeholders.”