Dental AI
Building AI that reads dental radiographs and supports clinical decision-making at the chair
Industry Challenge
AI in dentistry is rapidly evolving from concept to clinical tool — detecting caries, assessing periodontal bone loss, identifying impacted teeth, and planning orthodontic treatment from radiographs. The bottleneck is annotated training data: dental AI models need thousands of labeled panoramic and periapical radiographs reviewed by qualified dental professionals.
How SCILabel Serves This Industry
Data Collection
We source dental radiograph datasets — panoramic (OPG), periapical, bitewing, and CBCT — from dental clinics, dental schools, and oral health research centres through direct purchase and revenue-sharing agreements. All data is de-identified.
Data Annotation & Labeling
Our dental-trained annotators label caries (per tooth, per surface, per severity), alveolar bone levels for periodontal AI, root morphology, impacted tooth position (Winter's classification), crown and restoration status, and periapical lesion detection. We support standard dental ontologies including FDI and Universal numbering systems.
Data & Model Evaluation
Dental evaluators benchmark caries detection AI against dentist ground truth, assess sensitivity on early (enamel) caries versus advanced lesions, and test for performance variation across radiograph quality and imaging system.
Annotation Types & Formats
- Tooth-level bounding box and polygon annotation on OPG
- Caries classification per surface and severity (ICDAS)
- Alveolar bone level measurement for periodontal AI
- Periapical lesion detection and segmentation
- Impacted third molar classification (Winter's / Pell-Gregory)
Specialist Workforce Tracks
Track 2 (Medical Imaging): with dental specialisation: Dentists, Dental Therapists, Dental Radiographers.