Method for generating a 3D model of a dental arch
AssigneeDental Monitoring
InventorsPhilippe Salah, Guillaume Ghyselinck, Thomas Pellissard, Laurent Debraux
Life of this patent
Estimating the term…
Abstract
A method for generating a three-dimensional digital model of a dental arch of a patient. The method includes projection of at least one light beam onto the arch, so as to draw at least one light mark on the arch, and simultaneously, displacement of the arch across the beam and acquisition, during the displacement, of a series of updated images of the arch each showing an updated projection defined by the projected mark. The method includes identification of the updated projection on each updated image, then search for a three-dimensional digital model, called “updated model”, exhibiting a best fit with all the updated projections.
Claims15
A method for generating a three-dimensional digital model of a dental arch of a patient, by means of an acquisition device comprising:
- a support;
- a dental retractor fixed to the support and defining a retractor aperture;
- an updated image acquisition unit fixed to the support in a position in which it observes the retractor aperture along an optical axis;
- a projector suitable for projecting a light beam toward the retractor aperture, so as to draw, when the retractor is worn by the patient, at least one light mark on the arch of the patient, and
- a processing module,
- said method comprising the following steps:
- A) projection, by means of the projector, of at least one light beam onto the arch, so as to draw at least one light mark on the arch;
- B) simultaneously with the step A), displacement of the arch across the light beam and acquisition, during said displacement, of a series of said updated images of said arch each showing a representation of the projected light mark, or “updated projection”;
- C) identification, by the processing module, of said updated projection on each updated image, then production, by the processing module, of a three-dimensional digital model, called “updated model”, exhibiting a best fit with all the updated projections, the updated model being sought by means of an optimization method and/or a deep learning device.
Dependencies are inferred from the claim text.