FDA Places ML-Based Quantitative Imaging Software in Class II

What changed

The FDA classified radiological machine learning-based quantitative imaging software with a predetermined change control plan into class II, with special controls now being codified into the classification regulation.

Who this touches

Medical device makers and health systems building or deploying AI/ML-based quantitative imaging software as a medical device.

Why it matters

Class II status with special controls attaches defined expectations to the predetermined change control plan itself, not just to the software's intended use.

Suggested next step — and why

Device sponsors with quantitative imaging AI in development may want to compare existing change control documentation against the newly codified special controls, since this classification changes what the FDA expects on file.

The clock

No additional compliance date is specified; the special controls become part of the codified classification language for this device type as of this action.

The editor's read

The duty isn't to build a change control plan — it's to document each change against it. An undocumented change is indistinguishable from an unauthorized one. — IDV