Taking an optical impression of a patient's dental arch
View Patent ↗By combining the wide field of the image sensor of a smartphone ( 61 ) or the like, with the precision of the image sensor of an IOS ( 62 ) or the like, the process of generating a three-dimensional reconstruction of a patient's dental arch, referred to as extended optical impression, is improved and simplified. More specifically, the practitioner can take, in a single image, a global but not very precise 3D reconstruction ( 30 ) of the arch, then substitute, for the areas (or sectors) of this reconstruction which require a finer precision, precise local 3D reconstructions ( 41,42,43 ) obtained with the IOS. A related algorithm ( 263 ) is responsible for merging the global reconstruction with each of the local reconstructions, in order to produce a single precise 3D reconstruction ( 50 ) which corresponds to the desired optical impression.
1. A method for generating a three-dimensional reconstruction of a patient's dental arch, called an extended optical impression, comprising the following steps:
obtaining at least one global two-dimensional image of the acquired dental arch with a first image sensor having a relatively wide field of view and relatively low precision, and generating by first digital data processing means a global, relatively coarse three-dimensional reconstruction of the dental arch from said two-dimensional image;
acquiring a plurality of respective local two-dimensional images of respective associated sectors of the dental arch with a second image sensor having a relatively more restricted field of view and a relatively higher precision than the first image sensor, and generating by second digital data processing means respective local three-dimensional reconstructions, relatively more precise than the global three-dimensional reconstruction, of said associated sectors of the dental arch;
merging, by a merging algorithm performed by third digital data processing means, the global three-dimensional reconstruction and local three-dimensional reconstructions into a single total three-dimensional reconstruction wherein the local three-dimensional reconstructions are respectively substituted for portions of the global three-dimensional reconstruction that correspond to the associated sectors of the dental arch, the global reconstruction playing a role of absolute spatial guide, said merging comprising a setting of the local three-dimensional reconstructions in the global three-dimensional reconstruction, said setting being adapted to minimize a distance between the point clouds corresponding to each of the associated sectors of the dental arch and the corresponding point cloud of the global three-dimensional reconstruction.
2. The method according to claim 1 , wherein the first image sensor is an image sensor of a portable digital apparatus of the patient, such as a mobile phone, and wherein the global two-dimensional image is acquired by the patient itself using his portable digital apparatus and is transmitted from said portable digital apparatus to the practitioner to generate the global three-dimensional reconstruction.
3. The method according to claim 1 , comprising acquiring the global two-dimensional image by the practitioner using the first image sensor.
4. The method according to claim 3 , wherein the first image sensor is comprised in augmented reality glasses further comprising the first digital processing means for generating the global three-dimensional reconstruction.
5. The method according to claim 1 , wherein the second image sensor is an intraoral scanner, comprising the second digital data processing means for generating local three-dimensional reconstructions.
6. The method according to claim 1 , wherein the global three-dimensional reconstruction and/or local three-dimensional reconstructions are performed by generating, for each one, a depth map from a single 2D image acquired by the first sensor or by the second image sensor, respectively, based on a deep learning architecture trained beforehand for that purpose.
7. A computer program comprising instructions stored in a non-transitory medium, which, when the computer program is loaded into the memory of a computer and is executed by a processor of said computer, cause the computer to perform a method for generating a three-dimensional reconstruction of a patient's dental arch, called an extended optical impression, comprising the following steps:
obtaining at least one global two-dimensional image of the acquired dental arch with a first image sensor having a relatively wide field of view and relatively low precision, and generating by first digital data processing means a global, relatively coarse three-dimensional reconstruction of the dental arch from said two-dimensional image;
acquiring a plurality of respective local two-dimensional images of respective associated sectors of the dental arch with a second image sensor having a relatively more restricted field of view and a relatively higher precision than the first image sensor, and generating by second digital data processing means respective local three-dimensional reconstructions, relatively more precise than the global three-dimensional reconstruction, of said associated sectors of the dental arch;
merging, by a merging algorithm performed by third digital data processing means, the global three-dimensional reconstruction and local three-dimensional reconstructions into a single total three-dimensional reconstruction wherein the local three-dimensional reconstructions are respectively substituted for portions of the global three-dimensional reconstruction that correspond to the associated sectors of the dental arch, the global reconstruction playing a role of absolute spatial guide, said merging comprising a setting of the local three-dimensional reconstructions in the global three-dimensional reconstruction, said setting being adapted to minimize a distance between the point clouds corresponding to each of the associated sectors of the dental arch and the corresponding point cloud of the global three-dimensional reconstruction.
8. A device for generating a three-dimensional reconstruction of a patient's dental arch, called extended optical impression, comprising:
a first image sensor having a relatively wide field of view and relatively low precision for acquiring at least one global two-dimensional image of the dental arch, and first digital data processing means for generating a global, relatively coarse three-dimensional reconstruction of the dental arch from said two-dimensional image;
a second image sensor having a relatively more restricted field of view and a relatively higher precision than the first image sensor, for acquiring a plurality of local two-dimensional images of respective associated sectors of the dental arch, and second digital data processing means for generating respective local three-dimensional reconstructions, relatively more precise than the global three-dimensional reconstruction, of said associated sectors of the dental arch; and
third digital data processing means configured to implement a merging algorithm adapted to merge the global three-dimensional reconstruction and local three-dimensional reconstructions into a single total three-dimensional reconstruction wherein the local three-dimensional reconstructions are respectively substituted for portions of the global three-dimensional reconstruction that correspond to the associated sectors of the dental arch, the global reconstruction playing a role of absolute spatial guide, the merging algorithm comprising a setting of the local three-dimensional reconstructions in the global three-dimensional reconstruction, said setting being adapted to minimize a distance between the point clouds corresponding to each of the associated sectors of the dental arch and the corresponding point cloud of the global three-dimensional reconstruction.
9. The device according to claim 8 , comprising augmented reality glasses which comprise the first image sensor and, furthermore, the first digital processing means for generating the global three-dimensional reconstruction.
10. The device according to claim 8 , comprising an intraoral scanner that comprises the second image sensor and, furthermore, the second digital data processing means for generating local three-dimensional reconstructions.
11. The device according to claim 8 , wherein the first digital processing means and/or the second digital processing means are configured to generate the global three-dimensional reconstruction and the local three-dimensional reconstructions, respectively, by generating for each one a depth map from a single 2D image acquired by the first sensor or by the second image sensor, respectively, based on a deep learning architecture previously trained for this purpose.