Method for constructing a restoration

Abstract

The invention relates to a method for constructing a restoration, in which a dental situation is measured by means of a dental camera and a 3D model of the dental situation is generated. In this case, a computer-assisted detection algorithm is applied to the 3D model of the dental situation, wherein a type of restoration and/or at least a tooth number and/or a position of the restoration to be inserted are automatically determined.

Claims

1. A method comprising the steps of: measuring by a dental camera a dental situation; automatically generating a 3D model of the dental situation from the measured dental situation; applying to the 3D model of the dental situation, responsive to the generating, and without user interaction after said generating, a computer-assisted detection algorithm to automatically determine, a type of restoration to be inserted into the dental situation; wherein the dental situation has at least one preparation or an implant-supported mesostructure for inserting the restoration to be produced, wherein the computer-assisted detection algorithm has an artificial neural network for machine learning, and wherein, based on the 3D model of the dental situation, a shape of the preparation or of the implant-supported in mesostructure is analyzed by a machine learning system and said type of restoration is selected, wherein the machine learning system comprises one or more convolutional neural networks (CNN).

2. The method according to claim 1, wherein the type of restoration is an inlay, a crown, a bridge, an abutment, a pontic or a veneer.

3. The method according to claim 1, wherein using a surface of at least one residual tooth of a respective tooth corresponding to the restoration and/or of adjacent teeth relative to the respective tooth, the tooth number and/or a position of the tooth are additionally determined for the restoration to be inserted.

4. The method according to claim 3, wherein color information of the residual tooth of the respective tooth and/or of the adjacent teeth is used for the restoration to be inserted to specify a color for the restoration to be inserted.

5. The method according to claim 3, wherein the tooth number is determined and the determined type of restoration and/or the tooth number are used to specify a material for the restoration to be produced.

6. The method according to claim 1, wherein the determined type of restoration and/or tooth number are displayed to a user with the aid of a display device.

7. The method according to claim 1, wherein the determined type of restoration and/or tooth number are used to construct the restoration.

8. The method according to claim 1, wherein a machine learning system is trained by generating input data through analyzing a plurality of known 3D models of dental situations having a known type of restoration and a known tooth number; and augmenting the plurality of known 3D models by rotating said plurality of 3D models about defined degrees of freedom and/or scaling said plurality of 3D models along said degrees of freedom.

9. The method according to claim 8, wherein an individual CNN is applied to an individual 3D data of an individual 3D model of the plurality of known 3D models to train said individual CNN.

10. The method according to claim 1, further comprising: generating a plurality of heightfields of the 3D model of the dental situation for use as input to said machine learning system.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention is explained with reference to the drawings.

(2) The drawings show:

(3) FIG. 1 a sketch for illustrating the method for constructing a restoration,

(4) FIG. 2 a sketch for explaining the heightfields in occlusal direction,

(5) FIG. 3 a sketch for explaining the heightfields in labial and mesial direction.

DETAILED DESCRIPTION OF THE INVENTION

(6) FIG. 1 shows a sketch for illustrating the method for constructing a restoration 1, such as a bridge, in which a dental situation 3 is measured by means of a dental camera 2 and a 3D model 4 of the dental situation 3 is generated. The measurement of the dental situation 3, which is indicated by a rectangle, by means of the dental camera 2 is indicated by the dashed lines 5. The 3D model 4 of the dental situation 3 is displayed by means of a display device 6, such as a monitor, which is connected to a computer 7. The image data of the three-dimensional camera 2 are forwarded to the computer 7. Input devices such as a mouse 8 and keyboard 9 are connected to the computer 7, so that a user can navigate within the graphical display of the 3D model 4 with the aid of a cursor 10. The dental situation 3 includes a missing incisor 11 of the upper jaw 12 with a tooth number 11 in accordance with the dental chart. A first preparation 13 of the adjacent tooth with the tooth number 12 in accordance with the dental chart is arranged next to it. A second preparation 14 in the form of a tooth stump of the adjacent tooth with the tooth number 21 is arranged on the right side. The restoration 1 to be constructed and produced is shaped such that a first recess 15 is made to fit the first preparation 13 and a second recess 16 is made to fit the second preparation 14. The restoration 1 is thus placed onto the two preparations 13 and 14 and adhesively bonded. Using a computer-assisted detection algorithm, the 3D model 4 of the dental situation 3 is analyzed and a type of restoration 17 and a tooth number 18 are determined automatically, whereby the type of restoration 17 and the tooth number 18 are displayed in a menu 19 by the display device 6. A 3D model 20 of the restoration 1 to be produced is generated automatically on the basis of the measured 3D model 4 of the dental situation 3, the determined type of restoration 17 and the determined tooth number or position of the restoration 18 to be inserted, whereby significant structures, such as the shape of the preparation 15 and the preparation 16, the shape of a first adjacent tooth 21 and a second adjacent tooth 22, are taken into account. The first adjacent tooth 21 is a cuspid having the tooth number 13 and the second adjacent tooth 22 is a tooth having the tooth number 22 in accordance with the dental chart. Using the constructed 3D model 20, the restoration 1 can be carved out of a blank fully automatically by means of a CAM processing machine. The advantage of the discussed method is thus that the restoration 1 can be produced fully automatically after the measurement by means of the dental camera 2, without any need for user interaction.

(7) FIG. 2 shows a sketch for explaining the heightfields 30 which serve as input data for a CNN, whereby the CNN is a computer algorithm running on the computer 7 of FIG. 1. The 3D model 4 is cut at equal intervals perpendicular to an occlusal direction 31, such that sectional images or heightfields 30 are generated. When training a CNN, a large number of different 3D models of different dental situations are analyzed.

(8) In FIG. 3, the 3D model 4 is cut perpendicular to a labial direction 40, such that heightfields 41 or sectional images perpendicular to the labial direction 40 are produced. The 3D model 4 is also cut perpendicular to a mesial direction 42 such that height images 43 perpendicular to the mesial direction 42 are produced. The sectional images or height images 30, 41 and 43 serve as input data for the CNN, whereby the type of restoration 17 and the tooth number 18, for example, of the restoration 1 to be inserted of FIG. 1 are determined as output data of the CNN.

REFERENCE SIGNS

(9) 1 Restoration 2 Camera 3 Dental situation 4 3D model 5 Lines of the recording region 6 Display device 7 Computer 8 Mouse 9 Keyboard 10 Cursor 11 Incisor 12 Upper jaw 13 Preparation 14 Second preparation 15 Recess 16 Second recess 17 Type of restoration 18 Tooth number of restoration to be inserted 19 Menu 20 3D model 21 Adjacent tooth 22 Second adjacent tooth 30 Heightfields 31 Occlusal direction 40 Labial direction 41 Heightfields 42 Mesial direction 43 Height image