Second Opinion CC

Second Opinion CC is computer-aided detection (CADe) software that uses machine learning to identify potential caries (cavities) on dental radiographs. It displays detected lesions as polygonal overlays on bitewing and periapical X-rays to aid dentists as a second reader for patients 19 years or older, though it does not replace the dentist's clinical judgment.

Second Opinion CC uses neural network-based computer vision algorithms for caries detection via polygonal contours, whereas the primary predicate (Second Opinion K210365) uses bounding boxes. Both employ supervised machine learning, cloud-based processing, and graphical overlays on radiographs. The key difference is the localization method: polygons versus boxes.

The device was evaluated per FDA's July 2012 "Guidance for Industry and Food and Drug Administration Staff Computer-Assisted Detection Devices Applied to Radiology Images." Clinical testing used the Weighted Alternative Free-Response Receiver Operating Characteristic (wAFROC) paradigm and Jaccard Index (0.4 threshold) for lesion localization, with Dice coefficient for segmentation accuracy assessment.

Second Opinion CC is substantially equivalent because it shares the same intended use as the predicate (aiding caries detection in dental radiographs), employs the same underlying machine learning technology, produces near-identical clinical performance (non-inferior wAFROC-FOM with 95% CI lower bound exceeding -0.05 margin), and raises no new safety or effectiveness questions. The only material difference—polygons instead of boxes—does not affect the fundamental caries detection capability or clinical utility.