Better Diagnostics AI Corp. · Class II · Cleared Mar 11, 2025
| K-number | K241725 |
| Device name | Better Diagnostics Caries Assist (BDCA) Version 1.0 |
| Applicant | Better Diagnostics AI Corp. |
| Product code | MYN |
| Device class | Class II |
| Decision date | Mar 11, 2025 |
| Decision | Substantially Equivalent |
| Regulation | 892.2070 |
Better Diagnostics Caries Assist (BDCA) Version 1.0 is a computer-aided detection (CADe) software that automatically identifies and localizes carious lesions on bitewing and periapical dental radiographs for patients aged 18 and older. The software is intended for use by board-licensed dentists as an adjunct tool to assist in diagnosis, not as a replacement for clinical judgment or comprehensive dentist review.
BDCA v1.0 detects caries using colored bounding boxes rather than segmentation outlines used by the predicate device. It accepts BMP, JPG, and PNG image formats (predicate supports additional formats including DICOM, TIFF, EOP). BDCA has a three-layer architecture: APIs connecting dental PMS systems, cloud-hosted computer vision models performing image processing, and a presentation layer displaying AI-marked annotations. The age restriction is 18+ years versus the predicate's 12+ years.
IEC 62304 Edition 1.1 (medical device software life cycle), IEC 62366-1:2015 (usability engineering), ISO 14971 Third Edition (risk management), ISO 15223-1:2021 (medical device symbols). FDA guidance documents on computer-assisted detection devices and cybersecurity in medical devices were also applied.
Both BDCA and the predicate (Overjet Caries Assist) are Class II devices with product code MYN designed to detect caries on bitewing and periapical radiographs in a concurrent-read workflow. Although BDCA uses bounding boxes for localization instead of segmentation, the submission argues both approaches achieve identical clinical outcomes—detection of caries. The standalone and MRMC clinical studies demonstrate BDCA meets or exceeds predefined performance goals for sensitivity (88–89% at surface level) and specificity (99%+), with reader improvement studies showing AI assistance significantly enhances dentist detection capability. Minor differences in supported image formats and patient age range are deemed inconsequential to safety and effectiveness, supported by rigorous validation per IEC 62304.
View the full FDA submission: accessdata.fda.gov