The technology
Dental segmentation
An analysis trained on 18,000+ intra-oral photos annotated by Dr Romain Lateur, collected in real clinical conditions.
Study carried out by iTransform GP · February 2026
18k+
Annotated photos
6
Shots per patient
8
Architectures evaluated
92%
IoU (Voting Classifier)
The dataset
6 standardised shots per patient
Two arches, three views each — a reproducible protocol, annotated pixel by pixel by Dr Romain Lateur.
Maxillary
Left
Centre
Right
Mandibular
Left
Centre
Right
The model
Le Voting Classifier
Rather than trusting a single model, MouthCare combines the predictions of 8 deep learning architectures through a pixel-level vote. The result is more robust and reliable than any single model.
- Pixel-level majority vote across 8 evaluated architectures.
- Segmentation into 6 clinical classes (teeth, gums, tartar, cavities…).
- 92% IoU — accuracy validated with industry-standard metrics.
🔒 Your photo is analysed confidentially — never stored.
TeethGumsTartarCavityGingivitisCrown
Download the app
MouthCare, in your pocket
Available on iOS and Android — free.