DTX Studio Clinic 3.0

DTX Studio Clinic 3.0 is an AI-driven medical image analyzer that employs supervised machine learning to process intraoral radiographs and highlight potential dental pathologies for dentists. The device aims to enhance diagnostic accuracy by providing concurrent decision support in identifying findings such as caries, bone loss, and root canal deficiencies in patients 15 years and older.

DTX Studio Clinic 3.0 is a computer-assisted detection (CADe) software designed to aid dentists in analyzing intraoral radiographs, specifically bitewing and periapical images, for patients aged 15 years and older. The device utilizes an AI-powered Focus Area Detection algorithm, based on supervised machine learning, to identify and localize six specific dental findings: caries, calculus, periapical radiolucency, root canal filling deficiency, discrepancy at the margin of an existing restoration, and bone loss. The system functions as a concurrent review tool, presenting detected focus areas as togglable bounding boxes for the dentist's final diagnostic assessment. Performance data indicates that the AI algorithm, when assessed standalone, demonstrated an overall sensitivity of 0.79 and precision of 0.45, with specific sensitivities varying by finding type (e.g., 0.70 for caries, 0.95 for root canal filling deficiency). A clinical performance study involving 30 readers showed that using DTX Studio Clinic 3.0 significantly improved diagnostic performance, with an 8.7% overall increase in the Area Under the Curve (AUC) for detecting dental findings compared to unaided reading. Sensitivity for detecting existing dental findings improved by 22.4% overall with AI assistance, though a mild decrease in image-level specificity (8.7%) was observed.

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 algorithm's performance vary across different types or resolutions of intraoral radiographic images (e.g., sensor vs. phosphor plates, different brands)?; What is the clinical significance of the 8.7% decrease in image-level specificity with aided reading, and are there strategies to mitigate potential increases in false positives?; How does the AI model differentiate between subtle pathological findings and normal anatomical variations or artifacts, particularly for bone loss and periapical radiolucency?; What was the range of experience levels among the 30 readers in the clinical study, and did the AI's benefit differ significantly based on reader experience?; Are there plans to expand the detectable dental findings or to adapt the algorithm for different patient populations, such as pediatric patients or those with specific systemic conditions affecting oral health?