Second Opinion® 3D

Second Opinion® 3D is a radiological automated image processing software device that uses machine learning to identify and mark clinically relevant anatomy in dental CBCT radiographs, including dentition, maxilla, mandible, inferior alveolar canal, maxillary sinus, nasal space, and airway. It is designed to aid dental health professionals as a concurrent and second reader for patients 12 years of age or older, and should not be used as a substitute for full patient evaluation or sole basis for diagnosis.

Both the subject device and predicate use neural network-based computer vision algorithms developed from open-source models using supervised machine learning techniques to perform 3D modeling of patient anatomy. Both devices process CBCT radiographs, display mask overlays on the radiograph, produce near-instantaneous identification results, and require only basic documentation levels. Both passed all verification and validation testing requirements.

Not stated in this summary.

Second Opinion® 3D is substantially equivalent to Relu Creator (K233925) because both devices have identical intended use (aiding dental clinicians in identifying clinically relevant anatomy on CBCT radiographs), employ the same fundamental technology (automatic segmentation using machine learning algorithms), pose the same primary hazard (inaccurate anatomical segmentations causing temporary non-serious injury), and demonstrate comparable bench performance accuracy. The subject device's Dice Similarity Coefficient scores (0.76–0.97 across anatomies) are statistically significant and comparable to the predicate's reported performance of 0.92±0.02 for dental implant segmentation.