Videa Perio Assist

Videa Perio Assist is an AI-powered, cloud-based software designed to assist dental professionals by semi-automating the measurement and visualization of interproximal alveolar bone levels from bitewing and periapical radiographs. The system employs image processing algorithms to identify dental landmarks and calculate bone levels, presented as linear distances or percentages, to support clinical assessment of periodontal health in patients with permanent dentition.

Videa Perio Assist is a radiological semi-automated image processing software developed to aid dental professionals, including dentists and dental hygienists, in evaluating periodontal health. Its primary purpose is to measure and visualize mesial and distal alveolar bone levels associated with each tooth, utilizing bitewing and periapical radiographs from patients aged 12 and older with permanent dentition. The system operates as a cloud-based, AI-powered API that integrates with existing image viewers, allowing users to upload radiographs and receive outputs displaying identified bone levels as linear distances in millimeters or as percentages of root length. It is designed as an assistive tool and should not replace comprehensive patient evaluation or be the sole basis for diagnosis. The device's performance was evaluated through non-clinical bench testing and clinical studies. Bench testing on 996 radiographs demonstrated high recall (94.4% bitewing, 91.9% periapical) and precision (84.3% bitewing), though periapical precision was lower (79.1%) in subgroups with obscured interproximal points due to tooth overlap. Clinical testing on 189 radiographs, with results compared against a reference standard set by periodontists, showed good sensitivity (92.8% bitewing, 88.3% periapical) and specificity (89.4% bitewing, 87.0% periapical), with mean absolute error meeting pre-specified criteria (less than 1.5mm). While generally robust across different sensor manufacturers and patient age groups, some specific subgroups, such as periapical CEJ-ABL measurements for specificity and certain sensor/age combinations for specificity on bitewings, showed performance below acceptance criteria, often attributed to challenges like tooth overlap, though these instances were not deemed to impact overall millimeter accuracy when a prediction was made.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the Videa Perio Assist algorithm differentiate between true alveolar bone loss and anatomical variations or radiographic artifacts, particularly in the periapical CEJ-ABL subgroup where specificity was lower?; What specific image quality parameters (e.g., contrast, resolution, angulation) are most critical for the optimal performance of Videa Perio Assist, and how does the system address or flag suboptimal images?; Beyond the three sensor manufacturers tested, how robust is the device's performance across a wider range of intraoral X-ray sensors and imaging systems commonly found in diverse dental practices?; Could the AI be enhanced to provide a confidence score for its measurements, especially in challenging cases with overlapping teeth or obscured landmarks, to better guide clinical interpretation?; What is the typical learning curve for dental professionals to effectively integrate Videa Perio Assist into their daily workflow and accurately interpret its outputs, particularly when the software's measurements might differ from initial visual assessments?