Overjet Iris Intelligent Imaging System

The Overjet Iris Intelligent Imaging System is a web-based dental image management software (PACS) that acquires, stores, displays, and distributes digital dental images from intraoral sensors, panoramic machines, or consumer cameras. It includes AI/ML algorithms that automatically identify and report image capture quality issues (cone cuts, overlapping contacts, inadequate coverage, foreshortening, elongation, and intraoperative tools) to assist clinicians, though these quality indicators are not substitutes for professional clinical judgment.

The subject device is a web-application platform focused exclusively on 2D dental images, while the predicate (VistaSoft 4.0 and VisionX 4.0) is PC-based and handles both 2D and 3D images. Both share core features for acquiring, viewing, storing, and managing images with similar basic manipulation tools. The subject device provides AI-based reporting of image capture quality issues, whereas the predicate includes in-line automatic image plate quality checks and additional diagnostic support modules (nerve canal tracing, panoramic curve detection).

Software verification and validation (V&V) testing per FDA guidance 'Content of Premarket Submissions for Device Software Functions' (June 2023); IEC 62304:2006+A1:2015 for software lifecycle processes; FDA Guidance 'Cybersecurity in Medical Devices' (June 2025). Testing included unit-level validation, integration testing, system verification, and user acceptance testing on 1888 dental images (955 bitewing, 933 periapical).

Both devices share the same classification (21 CFR 892.2050), intended use (medical image management for dental professionals), and core functionality for acquiring and managing dental images with AI-limited to image quality assessment. The differences—web versus PC platform, 2D versus 2D/3D support, and different AI quality indicators—do not raise new safety or effectiveness questions because both AI functionalities are restricted to image quality reporting rather than diagnosis or interpretation. Performance testing demonstrated specificity of 98–99% and sensitivity of 91–94% across image types, meeting or exceeding pre-specified goals.