Method for generating a dental arch model

Abstract

The present invention provides a method for generating a three-dimensional digital model, referred to as an "updated model," of a patient's dental arch during the course of treatment of the dental arch with orthodontic aligners, referred to as "active aligners," wherein the treatment is simulated by a treatment scenario generated at a start time t1 and including a plurality of intermediate models, each of the intermediate models being a three-dimensional digital model of the dental arch, the intermediate models being divided into tooth models and generated to represent the dental arch at each intermediate time following the start time, the method comprising: 1) generating an updated model at an update time t1 during the course of treatment; a 2) determining one of the intermediate models, i.e., the "active intermediate model", depending on the update time, and then identifying one or more tooth models in the active intermediate model, each representing a non-matching tooth; and 3) deforming the active intermediate model until an updated model, called an "updated deformed image", is obtained that matches at least one of the updated images.

Description

[Technical Field]

[0001] The present invention relates to a method for generating a three-dimensional digital model of a dental arch.

[0002] The present invention also relates to a method for manufacturing aligners, particularly for the purpose of fitting orthodontic treatment with orthodontic aligners, by the manufacturing method according to the present invention.

[0003] Finally, the present invention relates to a computer system for carrying out these methods. [Background technology]

[0004] As shown in Figures 1 and 2, the orthodontic aligner 10 conventionally takes the form of a one-piece removable device, is conventionally made of a transparent polymer material, and has a shape corresponding to the continuous teeth of the dental arch to which the orthodontic aligner 10 is attached. The orthodontic aligner 10 is equipped with a groove 12, the general shape of which is U-shaped, and has a shape such that multiple teeth of the dental arch, generally all teeth of the dental arch, can be accommodated in the groove 12.

[0005] The groove shape is determined to ensure the aligner is securely attached to the tooth, but also in accordance with the desired target position of the tooth. More precisely, the shape is determined so that when the aligner is in its service position, it applies a stress that tends to move the treated tooth toward the target position.

[0006] Traditionally, at the start of orthodontic treatment, the shapes that various aligners must have at different times during treatment are determined, and then all corresponding aligners are manufactured. For this purpose, the following steps are taken: - At initial time t1, generally at the start of treatment, a 3D digital model called the "initial model" is generated of the patient's dental arch, which is the initial configuration. The initial model is then cut to create a tooth model. - The dental arch at each intermediate time t nModify from the initial configuration to the final configuration at the final time t through an intermediate configuration N+1 Determining an orthodontic treatment suitable for modifying to the final configuration at, where n is between 2 and N - Modifying the initial model to generate an intermediate model and a final model representing the dental arch in the intermediate configuration and the final configuration, respectively - Based on the initial model, the intermediate model, and the final model, determining a series of N aligners, where the first aligner is worn until time t2 and the nth aligner is intended to be worn from time t n to time t n+1 until - Manufacturing at least some of the aligners

[0007] Then, all of the manufactured aligners are provided to the patient so that the patient can replace the aligner at a predetermined intermediate time

[0008] During treatment, the patient visits an orthodontist for visual inspection at regular intervals to verify, in particular, whether the tooth movement is as expected and whether the aligner the patient is wearing is still suitable for treatment

[0009] Specifically, the orthodontist may visually diagnose whether the aligner has come off. Specifically, the bottom 20 of the groove has a shape that is substantially complementary to the shape of the free end ............... (FIG. 5). Therefore, the outer shape of the bottom of the groove may be compared with the outer shape of the tooth D, and the gap between the bottom of the groove and one or more free ends of the tooth may be evaluated

[0010] If detachment is detected, the orthodontist takes a new impression of the teeth or, similarly, performs a new scan of the teeth and then repeats the above process to design and manufacture a new series of aligners

[0011] Patients have to be forced to visit an orthodontist. The trust of patients in orthodontists may also decline. Finally, this incurs additional costs. Therefore, the number of examinations by orthodontists has to be restricted.

[0012] Furthermore, misfitted aligners may look bad.

[0013] To solve these problems, the applicant proposed in EP3412245 a method for evaluating the shape of an orthodontic aligner worn by a patient.

[0014] This method advantageously enables the remote detection of aligner detachment. This method considerably facilitates the assessment of the suitability of the aligner for treatment. Specifically, this method may be carried out, for example, by a patient, based on a simple image obtained without particular attention, especially based on a photograph or video. Therefore, the number of appointments with the orthodontist may be restricted.

[0015] However, when teeth that are not suitable for treatment are detected, especially when the aligner has come off the teeth, an appointment with the orthodontist has to be made to have a new series of aligners made. Apart from the inconvenience this causes to the patient, this appointment delays treatment. Specifically, treatment has to be interrupted from the time the detachment is detected until a new aligner is received.

[0016] Solutions to address these problems are needed.

Prior Art Documents

Patent Documents

[0017]

Patent Document 1

Patent Document 2

Patent Document 3

[0018] [Non-Patent Document 1] “Unpaired image-to-image translation using cycle-consistent adversarial networks” by Zhu, Jun-Yan et al. [Overview of the Initiative] [Problems that the invention aims to solve]

[0019] One object of the present invention is to satisfy this requirement at least partially. [Means for solving the problem]

[0020] The present invention provides a method for generating a three-dimensional digital model of a patient's dental arch, called an "updated model," in the course of treatment of the dental arch with orthodontic aligners called "active aligners," particularly in the context of treatment with a series of orthodontic aligners intended to be sequentially attached to the dental arch, wherein the treatment is simulated by a treatment scenario generated at an initial time t1, for example, at the start of treatment, and the treatment scenario includes a plurality of intermediate models, each intermediate model being a three-dimensional digital model of the dental arch, and the intermediate models are determined to be cut to create tooth models and represent the dental arch at each of the intermediate times after the initial time.

[0021] The production method according to the present invention consists of the following steps, namely, 1) A step of acquiring at least one update image during the update time between treatments, wherein each update image is an aligner image representing an active aligner attached to the dental arch at the position of use, or an image of exposed teeth representing a dental arch without an aligner, 2) Before step 4), and preferably before step 3), a step of determining the intermediate model or "active intermediate model" according to the update time, 3) A step of searching for the representation of one or more teeth that do not fit the treatment scenario in the updated image called the “analysis update image”, If one or more mismatched teeth are detected, 4) The steps include identifying one or more tooth models that represent one or more mismatched teeth in the active intermediate model, 5) The step of deforming the active intermediate model until an updated model is obtained that fits at least one of the updated images, called a “deformed updated image”.

[0022] This invention is based on the premise that, when misfitted teeth are ignored, an intermediate model of the treatment scenario designed before treatment commencement correctly models the dental arch at the corresponding intermediate time. Specifically, if the aligner is not dislodged from a tooth, it indicates that treatment is progressing as planned with respect to that tooth. Therefore, at the intermediate time close to the update time, the corresponding intermediate model is realistically fitted for the "fitted" teeth, which are generally almost all teeth. Thus, conventionally, this active intermediate model, initially generated for the manufacture of the aligner, can be used as a starting point to create the updated model.

[0023] Specifically, all parts of the active intermediate model for the fitted teeth can be immediately utilized. Therefore, the updated model may be assembled based on the active intermediate model, and only the actual positions of the tooth models for the non-fitted teeth are pursued.

[0024] Therefore, advantageously, patients no longer need to undergo new scans to adapt the aligner to the negative progression of their treatment. One or preferably more updated images are sufficient for the treatment scenario. This significantly simplifies and accelerates treatment.

[0025] Furthermore, the determination of the updated model is simplified because the number of tooth models that need to be moved is very limited, generally only 1 to 5 tooth models represent misfit teeth, and such moves are constrained by the tooth models of the fit teeth.

[0026] This method may include a step before step 1) of generating treatment scenarios that include the multiple intermediate models. The treatment scenarios are generated after the treatment itself has been determined. The treatment scenarios may also be generated at the same time as the treatment itself is determined.

[0027] The method according to the present invention may include one or more of the following features: - At least some, or possibly all, of the intermediate models of the treatment scenario represent the dental arch in the configuration expected at the intermediate time showing aligner replacement. - In step 1), a reminder is preferably sent to the patient on the patient's mobile phone, thereby prompting the patient to take at least one updated image, preferably at least one aligner image and preferably at least one image of an exposed tooth. - In step 1), more than two updated images, preferably more than four updated images, - In step 1), at least two updated images are acquired under different acquisition conditions, particularly with different orientations of the acquisition device, and preferably the angle of the optical axis of the acquisition device relative to the patient's front differs by more than 20°, more than 30°, more than 45°, more than 60°, or possibly more than 90° between the acquisitions of at least two updated images. - In step 1), at least one updated image, preferably each updated image, is an image extracted from a photograph or video. - In step 1), at least one updated image, preferably each updated image, is an extraoral image. - In step 1), at least one updated image, preferably each updated image, is preferably taken by the patient themselves using a mobile phone, and possibly after the dental retractor has been fitted. - In one embodiment, in step 1), at least one updated image, preferably each updated image, is taken preferably by the patient himself using a mobile phone, after the mobile phone and dental retractor are mounted in the holder and then the dental retractor is placed in the patient's mouth. - The holder takes the form of an exclusive casing opening on the retractor and on the mobile phone. - All updated images are taken at time intervals of less than 5 days, preferably less than 1 day, preferably less than 1 hour, preferably less than 10 minutes. - The patient preferably sends one or more updated images to a computer using a mobile phone that acquired the aligner image. - The computer is configured to receive and process updated images of multiple patients, preferably more than 100, more than 1,000, or more than 10,000 patients. - Before step 2), an intermediate model of the dental arch is generated from the intermediate model of the treatment scenario, and then added to the treatment scenario via the intermediate model. - In step 2), the active intermediate model is determined manually by an operator, preferably a dental professional, more preferably an orthodontist, preferably using a computer that allows the operator to visualize the treatment scenario, or is performed automatically by a computer, preferably the computer receiving updated images of multiple patients and processing them, preferably by comparing the update time with the intermediate time of the intermediate model of the treatment scenario. - In step 2), the active intermediate model is an intermediate model in which the difference between the intermediate time and the update time is less than 4 weeks, less than 2 weeks, preferably less than 1 week. - In step 2), the active intermediate model is the intermediate model whose intermediate time is closest to the update time. - In step 3), the search for mismatched teeth in the updated images is performed manually by an operator, preferably a dental professional, more preferably an orthodontist, preferably using a computer that allows the operator to view one or more updated images, or preferably automatically by a computer, preferably the computer receiving and processing updated images of multiple patients, the computer preferably implementing a deep learning device, and preferably a neural network. - In step 3), the tooth display in the analysis update image is considered not to fit the treatment scenario if, when the analysis update image coincides with and overlaps with the view of the active intermediate model that fits the analysis update image, which is on the same scale and life-size as the view of the active intermediate model (the actual tooth dimensions are the same as the tooth display dimensions), at least one point of the display is located at a distance greater than 1 / 10 mm, 3 / 10 mm, 5 / 10 mm, or 1 mm, preferably less than 7 mm or 5 mm, from the corresponding point in the view. - In step 4), tooth models of misfitted teeth in the active intermediate model are identified manually by an operator, preferably a dental professional, more preferably an orthodontist, preferably using a computer that allows the operator to view the active intermediate model, or automatically by a computer, preferably the computer receiving and processing updated images of multiple patients, the computer preferably implementing a deep learning device and preferably a neural network. - In step 3), to detect mismatched teeth, the updated analysis image is analyzed independently of the active intermediate model, in which case the active intermediate model may be determined after step 3). - Alternatively, in step 3), to detect mismatched teeth, the updated analysis image is compared to the view of the active intermediate model, in which case the active intermediate model must be determined before step 3). - In step 3), in order to detect mismatched teeth, The positioning, localization, and calibration of the virtual acquisition device are pursued to enable the virtual acquisition device to have a view on the active intermediate model that is as close as possible to the analysis update image, i.e., a view of the active intermediate model that has the greatest degree of fit with the analysis update image (i.e., the best fit), and then, The view and the updated analysis image are compared, or the updated map representing the distinction information of the updated analysis image is compared with the reference map representing the distinction information in the view. - In step 3), in order to detect mismatched teeth, The updated analysis image is an aligner image, and the external shape of at least one tooth and the external shape of the aligner are determined in the updated analysis image, and then the external shapes are compared with each other. - Teeth are considered unfitted if they are outside the aligner by a threshold in the aligner image. - Before step 5), the active intermediate model is processed to improve its accuracy. - In step 5), the deformation of the active intermediate model includes, and preferably consists of, the movement of the tooth model of the active intermediate model. - In step 5), the movement of the tooth model is continued until the positional error of each tooth model relative to the deformed updated image is less than 1 mm, preferably less than 5 / 10 mm, preferably less than 3 / 10 mm, preferably less than 2 / 10 mm, preferably less than 1 / 10 mm. - In step 5), the tooth model of the misfitted tooth is moved manually by an operator, preferably a dental professional, more preferably an orthodontist, preferably using a computer that allows the operator to view an active intermediate model, or is moved automatically by a computer, preferably the computer receiving and processing updated images of multiple patients, the computer preferably implementing a deep learning device, and preferably a neural network, or an optimization method, and preferably a metaheuristic optimization method, - In step 5), the movement of the tooth model of the misfitted tooth is restricted by the tooth model of the misfitted tooth, which remains stationary. - In step 5), the movement of the tooth model of the misfitted tooth is an iterative process, and in each iteration, - One or more of the tooth models of the mismatched teeth are moved to obtain the dental arch model to be tested, and then, - The model to be tested is tested by evaluating the degree of fit between the model and the deformed updated images, particularly the images of exposed teeth. The updated model is the model with the highest degree of compliance among all tested models. - Prior to the iterative process, the position, localization, and calibration of the virtual acquisition device, or "constrained virtual acquisition conditions," are pursued, which enable the observation of the active intermediate model to generate a view in which the representation of the fitted teeth can coincide with and overlap the representation of the fitted teeth in the deformed updated image, and then, During the aforementioned iterative process, in each iteration, the degree of fit between the model under test and the deformed updated image is evaluated by comparing the deformed updated image with a view of the model under test obtained under the constrained virtual acquisition conditions. - The iteration cycle is interrupted when the number of iterations exceeds a predetermined number or when the goodness-of-fit value exceeds a predetermined threshold. - In step 5), during the deformation of the active intermediate model, the tooth model to be moved is only the tooth model of the misfitted tooth. - At the end of step 3) or 4), preferably step 4), the degree of mismatch of one or more mismatched teeth, and in particular the amount of movement of one or more dislodged teeth, is measured based on at least one updated image, preferably by comparing the analyzed updated image with the active intermediate model, and then, In step 5), one or more tooth models of misfitted teeth are moved, according to the measured values, preferably until the placement error for each tooth model relative to the deformed updated image is less than 1 mm, preferably less than 5 / 10 mm, preferably less than 3 / 10 mm, preferably less than 2 / 10 mm, preferably less than 1 / 10 mm. - In step 5), the deformation of the active intermediate model includes the movement of one or more tooth models of misfitted teeth, the amplitude and / or direction of the movement determined according to a measurement of the degree of misfit of the one or more misfitted teeth, the measurement being performed based on at least one updated image, in particular an aligner image, preferably by comparing the updated image with the active intermediate model.

[0028] The present invention relates to a method for manufacturing an orthodontic aligner, comprising steps 1) to 5), followed by the following steps, namely, 6) A step of designing an "updated" aligner suitable for modifying the dental arch from the actual configuration at update time to the theoretical final configuration at the point of use, based on the updated model and the final model representing the dental arch in the theoretical final configuration, 7) The method also includes the steps of manufacturing a replacement aligner and providing the replacement aligner to a patient.

[0029] The theoretical final configuration planned for the dental arch in the final time after the last intermediate time is generally the theoretical final configuration of the dental arch that is targeted at the end of treatment.

[0030] The methods according to the present invention may be partially performed by a computer, particularly with respect to the steps of modifying a model, calculating or investigating a model to explore constrained virtual acquisition conditions in particular, or analyzing an image or map to determine, for example, the outline.

[0031] The present invention - A computer program that, when executed by a computer, includes program code instructions for executing one or more, preferably all, of steps 2), 3), 4), 5), and possibly step 6), - A computer-readable storage medium on which such a program is stored, for example, memory or a CD-ROM, and - This also relates to the computer on which such a program is loaded.

[0032] The present invention - A personal device, preferably a mobile phone, configured to acquire one or more updated images in step 1), - A computer is loaded with a program containing program code instructions to perform one or more, preferably all, of steps 2) to 5) and possibly step 6) when executed by a computer, i.e., a computer "configured" to perform these steps, - In some cases, the system also includes a computer loaded with a program configured to manufacture an aligner in step 7).

[0033] definition "Patient" or "user" means any person on whom the method according to the present invention is performed, regardless of whether or not that person has a disease.

[0034] "Tooth" refers to a set of teeth in the dental arch.

[0035] "Dental healthcare professional" refers to any person qualified to provide dental treatment, and specifically includes orthodontists and dentists.

[0036] The aligners used by the patient during treatment are called "active aligners." In treatment with multiple aligners, each aligner is expected to become active sequentially.

[0037] A 3D scanner, or "scanner," is a device that makes it possible to obtain a model of the dental arch.

[0038] "Position of use" refers to the position of the aligner when it is attached to the dental arch for treatment. Traditionally, this attachment can be deactivated by the patient by simply pulling the aligner.

[0039] When an aligner is attached to the dental arch in its place of use, teeth that are not properly positioned within the aligner and teeth that are properly positioned within the aligner are referred to as "out-of-place teeth" and "in-place teeth," respectively. An orthodontist can perfectly distinguish between out-of-place and in-place teeth. This distinction may also be made by a computer evaluating the distance between the tooth and the bottom of the groove in the aligner to which it is attached.

[0040] More generally, a tooth is said to be "fit" or "ill-fit" when, at the time of update, it is in or out of the position planned for the treatment scenario. A loose tooth is an example of an ill-fitting tooth.

[0041] "Update time" is the time it takes for updated images to be acquired. This time is short enough that the tooth structure hardly changes during this period.

[0042] The configuration of the dental arch is called the “actual” configuration when it is the configuration that the patient’s dental arch actually has. The configuration of the dental arch is called the “theoretical” configuration when it is a “simulated” configuration of the patient’s dental arch or a configuration “planned” for the future.

[0043] "Model" refers to a three-dimensional digital model. A model consists of a set of voxels. A "dental arch model" is a model that represents at least a portion of the dental arch, preferably at least two, preferably at least three, and preferably at least four teeth. Figure 3 shows an example view of a dental arch model.

[0044] A "tooth model" is a three-dimensional digital model of the teeth in a patient's dental arch. The dental arch model may be cut to define at least some teeth, preferably all teeth represented in the dental arch model. Thus, the tooth model is a model within the dental arch model. Figure 4 shows an example view of a cut dental arch model. Computational tools exist for manipulating the tooth model in a dental arch model. Such tools allow for setting constraints, in particular, to limit the movement of the tooth model to realistic movement, for example, to prevent adjacent tooth models from penetrating each other.

[0045] A "scenario" is a sequence of dental arch models that represent a continuous set of dental arch configurations. Specifically, a "treatment scenario" or "treatment plan" includes models that represent the configuration of the dental arch at various points in time during dental treatment. These times are traditionally the initial time before treatment begins, the intermediate time during treatment, and the final time when treatment is completed. Each model in the scenario that represents the expected configuration of the dental arch at the intermediate time is called an "intermediate model." Figure 7 shows an example of a treatment scenario.

[0046] The configuration of the dental arch at intermediate and final time points is theoretical, as it is derived from simulations of the future. Therefore, these configurations are expected or "planned" configurations and may differ from reality at intermediate time points. By displaying the scenario model chronologically, it becomes possible to simulate the effects of dental arch treatment.

[0047] Treat is an example of a software package for manipulating tooth models and creating treatment scenarios, and it is described on the page https: / / en.wikipedia.org / wiki / Clear_aligners#cite_note-invisalignsystem-10. U.S. Patent No. 5975893A also describes the creation of treatment scenarios.

[0048] An "image" refers to a two-dimensional image, such as a photograph or video. Images are made up of pixels.

[0049] The "acquisition conditions" are: - The spatial position, spatial localization, and calibration of the actual device for acquiring images of the patient's dental arch (e.g., aperture and / or exposure time and / or focal length and / or sensitivity values) (actual acquisition conditions), or - Specify the spatial position, spatial localization, and calibration (e.g., aperture and / or exposure time and / or focal length and / or sensitivity values) of a virtual device for acquiring images of a model of the patient's dental arch (virtual acquisition conditions).

[0050] The "calibration" of an acquisition device consists of all the values ​​of its calibration parameters. Calibration parameters are specific to the acquisition device (unlike the position and localization of the acquisition device) and their values ​​affect the image being acquired. For example, aperture is a calibration parameter that corrects depth of field. Exposure time is a calibration parameter that corrects the brightness (or "exposure") of the image. Focal length is a calibration parameter that corrects the field of view, i.e., the amount of "zoom". Sensitivity is a calibration parameter that corrects the response of the digital acquisition device's sensor to incident light.

[0051] The calibration parameters are preferably selected from the group formed by aperture, exposure time, focal length, and sensitivity.

[0052] An observation of the model under determined conditions, angles, and distances of virtual acquisition (particularly through calibration of the virtual acquisition device) is called a "view."

[0053] "Image of the dental arch," "View of the dental arch," "Display of the dental arch," "Scan of the dental arch," or "Model of the dental arch" means an image, view, display, scan, or model of all or part of the aforementioned dental arch.

[0054] A model of the patient's dental arch "fits" to the image when there is a view of this model corresponding to the image, i.e., a view in which the representation of teeth in the view is similarly positioned to the representation of teeth in the image. Thus, the outline of the tooth model represented in the view can substantially coincide with and overlap the outline of the representation of teeth in the image.

[0055] This view of the model may be considered to "fit" or "coincide and overlap" with the aforementioned image.

[0056] A deep learning algorithm is a deep learning device known to those skilled in the art. A deep learning algorithm includes a "neural network" or "artificial neural network."

[0057] Those skilled in the art can select a neural network depending on the task to be performed. Specifically, the neural network can be selected from the following: - A network called a convolutional neural network (CNN), which is specialized for image classification, for example, - AlexNet (2012) - ZF Net (2013) - VGG Net (2014) - GoogleNet (2015) - Microsoft ResNet (2015) - Caffe: BAIR Reference CaffeNet, BAIR AlexNet - Torch: VGG_CNN_S, VGG_CNN_M, VGG_CNN_M_2048, VGG_CNN_M_1024, VGG_CNN_M_128, VGG_CNN_F, VGG ILSVRC-2014 16-layer, VGG ILSVRC-2014 19-layer, Network-in-Network(Imagenet & CIFAR-10) - Google: Inception (V3, V4) - A network specialized in identifying and detecting objects in an image (object detection network), for example, - R-CNN (2013) - SSD (single-shot multibox detector: object detection network), Faster R-CNN (faster region-based convolutional network method: object detection network) - Faster R - CNN (2015) - SSD (2015) - RCF(Richer Convolutional Features for Edge Detection)(2017) - A network specialized in image generation, for example, - Cycle-Consistent Adversarial Networks(2017) - Augmented CycleGAN (2018) - Deep Photo Style Transfer(2017) - FastPhotoStyle (2018) - pix2pix (2017) - Style-Based Generator Architecture for GANs(2018) - SRGAN (2018).

[0058] The list above is not exhaustive.

[0059] Training a neural network involves having the neural network deal with a training database containing information about two types of objects that the neural network must "match," that is, learn to connect with each other.

[0060] Training may be performed based on a training database in which each record comprises a first object of a first type and a corresponding second object of a second type.

[0061] Alternatively, training may be performed based on a training database in which each record contains either a first object of a first type or a second object of a second type, but each record contains records containing information about the type of object it contains. Such training techniques are described, for example, in the article "Unpaired image-to-image translation using cycle-consistent adversarial networks" by Zhu, Jun-Yan et al.

[0062] Training a neural network with these records teaches the neural network to feed corresponding objects of the second type from any object of the first type.

[0063] The quality of analysis performed by a neural network directly depends on the number of records in the training database. It is preferable that the training database contains more than 10,000 records.

[0064] To evaluate the "placement error" of the tooth model, the distance between each point of the tooth modeled by the tooth model in the deformation update image and the corresponding point in the view (when the deformation update image at a 1:1 scale coincides with and overlaps with the view of the active intermediate model at a 1:1 scale that fits the deformation update image) is measured. The placement error is the longest of these distances when considering all points of the display that have corresponding points in the view. A 1:1 scale means that the tooth display is life-size, in which case the deformation update image and the view of the active intermediate model represent teeth with actual dimensions.

[0065] Unless otherwise indicated, "to have," "to include," or "to possess" must be interpreted broadly and non-restrictively.

[0066] Other features and advantages of the present invention will become clearer when you read the detailed description below and examine the accompanying drawings. [Brief explanation of the drawing]

[0067] [Figure 1] This is a perspective view of an orthodontic aligner. [Figure 2] Figure 1 is a top view of an orthodontic aligner. [Figure 3] This figure shows an example of an initial model (intermediate, final, and updated models may take a similar form). [Figure 4] This figure shows an example of a tooth model that has been cut (only the tooth model is shown). [Figure 5] This diagram schematically shows the holding arch that supports orthodontic aligners. [Figure 6] This figure shows a system suitable for carrying out the method according to the present invention. [Figure 7] This is a diagram showing the treatment scenario. [Figure 8] This figure schematically illustrates the method according to the present invention. [Figure 9] This figure shows an example of an aligner image. [Figure 10] This diagram schematically illustrates the acquisition of aligner images and / or exposed teeth. [Figure 11] This figure schematically shows a retractor that may be used with the acquisition kit shown in Figure 10. [Figure 12] This diagram schematically illustrates a first method for detecting outliers in an image. [Figure 13] This diagram schematically illustrates a second method for detecting outliers in an image. [Modes for carrying out the invention]

[0068] In one embodiment, the method according to the present invention includes the following steps before carrying out steps 1) to 5): Initial time t1, generally at the start of treatment, a) The steps include generating a 3D digital model called the "initial model" of the patient's dental arch, which is the actual initial configuration, and then cutting the initial model to create a tooth model, b) The dental arch at each intermediate time t n A step of determining the dental arch treatment to modify the initial configuration from the initial configuration to the final configuration at the end of treatment, generally the theoretical final configuration at the end of treatment, where n is between 2 and N, c) A step of modifying the initial model to generate a treatment scenario that includes a final model and an intermediate model representing the dental arch in the final and intermediate configurations, respectively, d) The step of designing a series of aligners based on the initial model, intermediate model, and final model, e) The step of manufacturing one or more of the aligners and providing these aligners to the patient (Figure 8).

[0069] In one embodiment of the present invention, the method comprises steps 1) to 5) of the method for generating an updated model according to the present invention during treatment, and then preferably the following steps, i.e. 6) A step of designing an updated aligner based on the updated model and the final model, which is shaped to modify the dental arch from the actual configuration of the dental arch at the time of update to the final configuration, 7) Further comprising the steps of manufacturing a replacement aligner and providing the replacement aligner to the patient.

[0070] In step a), an initial model is created at an initial time t1, which is prior to the start of orthodontic treatment with an orthodontic aligner, preferably less than 6 months before the start of treatment, preferably less than 3 months, less than 1 month, or less than 2 weeks before.

[0071] The initial model may be created based on measurements taken on the patient's teeth or a physical model of the patient's teeth, such as a plaster model.

[0072] The initial model is preferably created using specialized equipment, such as a 3D scanner, which is preferably operated by a dental professional, such as an orthodontist or orthodontic laboratory. In an orthodontic clinic, the physical model of the patient or the patient's teeth may be advantageously positioned in precise locations, and the specialized equipment may be state-of-the-art. Thus, a very accurate initial model may be obtained. The initial model preferably provides information about the position of the teeth with an error of less than 5 / 10 mm, preferably less than 3 / 10 mm, and preferably less than 1 / 10 mm.

[0073] The initial model is, for example, a point cloud type or one of the following types: STL, OBJ, 3D DXF, IGES, STEP, or VDA. Advantageously, such a so-called "3D" model may be observed from any angle.

[0074] The initial model is conventionally observable and manipulable by computer. The initial model is then cut to define the tooth model.

[0075] Creating a tooth model by cutting a 3D model is a conventional operation in which the model is cut so that it defines the representation of one or more teeth in the initial model. Other elements of the dental arch, such as the gingiva, may also be modeled.

[0076] The initial model may be disconnected manually by an operator using a computer, or it may be disconnected automatically by a computer, which preferably implements a deep learning device and, more preferably, a neural network.

[0077] Specifically, the tooth model may be defined, for example, as described in international patent application PCT / EP2015 / 074896.

[0078] Figure 4 shows an example of an initial model in which tooth model 32 has been cut (only the tooth model is shown, and different shading has been applied for easier identification).

[0079] After cutting, the tooth model may be moved. The computer may then simulate the movement of the tooth from the initial time to the final time when the tooth is in its final configuration, without modifying the tooth model, by deforming the initial post-cut model through the movement of the tooth model in this way. The final time may, in some cases, indicate the end of treatment.

[0080] In step b), a treatment is decided upon to move one or more teeth from the initial configuration through the intermediate configuration to the final configuration.

[0081] A set of dental arch models, which allows for the visualization of each stage of treatment, forms the treatment scenario. The treatment scenario is displayed using a computer, and an initial model, modified to simulate the configuration of the dental arch at various intermediate time points, is stored in memory.

[0082] Conventionally, there are multiple possible potential scenarios for a given treatment. In one embodiment, a computer determines the potential scenarios and selects a treatment scenario from them. In another embodiment, a computer determines the potential scenarios and presents them to a dental professional, who then selects a treatment scenario from them. In another preferred embodiment, a dental professional, preferably an orthodontist, determines the potential scenarios and selects a treatment scenario from them. The computer advantageously allows the dental professional to visualize a simulation of the effect of the potential scenarios on the dental arch.

[0083] In step c), the initial model is transformed to generate a final model representing the dental arch in the theoretical final configuration, and an intermediate model showing the stages between the initial and final models.

[0084] The deformation may be determined by a dental professional, preferably an orthodontist, in accordance with orthodontic treatment, using a computer that allows the dental professional to visualize the effect of the treatment to be conceived on the dental arch, or it may be determined automatically by a computer, which preferably implements a deep learning device and, more preferably, a neural network.

[0085] Steps b) and c) are preferably performed simultaneously. Specifically, the treatment scenario is determined as a result of one or more simulations performed by moving the tooth model, thereby deforming the initial model into the configuration of the final model. Once a treatment is selected, it is sufficient to simulate this treatment and save the deformed initial model at the intermediate time to generate an intermediate model.

[0086] Figure 9 shows an example of a treatment scenario including an initial model, two intermediate models, and a final model.

[0087] In step d), the series of aligners are designed to deform the dental arch through tooth movement, depending on the treatment scenario.

[0088] In step e), one or more of the first aligners are manufactured. Conventionally, all of the aligners are manufactured. These aligners are provided to the patient, thereby allowing the patient to begin treatment.

[0089] The method, including steps a) through e), is publicly known and widely used to design and manufacture a series of orthodontic aligners.

[0090] Traditionally, patients are monitored by their orthodontist. As explained in the introduction, the orthodontist periodically checks the fit of the aligners, and if they do not fit the treatment, especially if the aligners have come loose, a new dental arch model is generated by scanner, and then steps b) to e) are repeated with the original model replaced by this new model, thereby creating a new set of aligners to be used for the remainder of the treatment.

[0091] According to the present invention, this method comprises steps 1) to 5) and preferably steps 1) to 7) during treatment.

[0092] In step 1), at the update time, an updated image called an “analyzed” updated image is acquired, which enables the detection of non-conformity, particularly tooth detachment from the aligner that was fitted at the time of the update.

[0093] The update time may be, for example, more than two weeks, more than four weeks, more than eight weeks, or more than twelve weeks after the initial period.

[0094] Preferably, at least one reminder is sent to the patient before the update time, for example, less than two weeks before the update time, informing the patient that they need to take an analysis update image. This reminder may be in paper form, or preferably in electronic form, such as email, an automated alert from a dedicated mobile application, or SMS. Such a reminder may be sent from the orthodontic clinic or orthodontic laboratory, or by the dentist, or by a dedicated application on the patient's mobile phone.

[0095] The analysis update images are taken by an image acquisition device, preferably a personal acquisition device, preferably a mobile phone, so-called "connected" camera, so-called "smart" watch, tablet, or (desktop or laptop) personal computer equipped with an image acquisition system such as a webcam or camera.

[0096] The updated images for analysis are preferably extraoral images.

[0097] In one embodiment, a photographic kit 15 is used, as shown in Figures 10 and 11. Such a kit preferably comprises a holder 17, a dental retractor 19, and an image acquisition device, preferably a mobile phone 21. The dental retractor 19 and the acquisition device, preferably the mobile phone, are preferably detachably attached to the holder 17.

[0098] The retractor 19 may have the characteristics of a conventional retractor.

[0099] As shown in Figure 11 (with the retractor 19 separated from the holder), the retractor 19 preferably has a groove 23 that extends around the retractor opening of axis X and is positioned so that the patient's lips are located inside while the patient's teeth are visible through the retractor opening.

[0100] The retractor 19 may be attached to the holder by one or more clips 27a and 27b, for example, magnetic clips.

[0101] Preferably, the retractor is equipped with lugs 26a and 26b for moving the cheek out of the way, so that an acquisition device attached to the holder can take a photograph of the vestibular surface of teeth located at the back of the mouth, such as molars, through the retractor opening.

[0102] The holder preferably takes the form of an exclusive casing opening over the retractor opening and over the mobile phone. Thus, the mobile phone observes the patient's dental arch through the casing. Advantageously, the holder allows the position of the mobile phone relative to the dental arch to be pre-set.

[0103] The acquisition is preferably performed by the patient or a friend or relative of the patient, but may also be performed by another person, particularly a dentist or orthodontist, preferably without the need to precisely position the image acquisition device against the teeth.

[0104] The analysis update image is preferably a photograph or an image extracted from video. The analysis update image is preferably a color image, and preferably a true-color image. The analysis update image is more preferably a photograph showing the actual dental arch as perceived by the human eye (unlike tomographic or panoramic images obtained by X-ray).

[0105] The updated analysis image is then preferably sent to a central computer by the mobile phone that acquired the updated analysis image.

[0106] A dedicated application is loaded onto the mobile phone, preferably providing auditory and / or visual guidance to the patient regarding various actions to be performed, and transmitting updated analysis images.

[0107] In one embodiment, the analysis update image is an aligner image, which is advantageous as it becomes possible to detect mismatches, particularly outliers, by analyzing only this image. This detection may be performed by comparing this image with an active intermediate model.

[0108] In one embodiment, the analysis update image is an image of an exposed tooth, which is advantageous as it allows for the detection of mismatch with high accuracy, particularly without interference from the display of the aligner. However, in this case, it is necessary to compare this image with an active intermediate model.

[0109] In step 1), it is preferable that aligner images and exposed tooth images are acquired in such a way that the advantages of two complementary analyses can be obtained.

[0110] In step 1), it is preferable that at least one image of the exposed tooth is obtained, which will be used as the deformed update image in step 5).

[0111] In step 2), the intermediate model or "active intermediate model" is preferably determined by a computer in accordance with the update time. Step 2) may be performed following step 3) if the analysis update image has been analyzed independently of the active intermediate model in order to detect misfit teeth in step 3).

[0112] From the treatment scenario, preferably by computer, an intermediate model is selected that should best represent the dental arch at the update time according to this scenario. Assuming that the treatment is progressing according to the treatment scenario, the intermediate model whose intermediate time is closest to the update time may be selected. This intermediate model is said to be "active".

[0113] The active intermediate model is an intermediate model in which the intermediate time is close to the update time, preferably with a difference of less than 4 weeks, less than 2 weeks, and preferably less than 1 week from the update time. The intermediate time of the active intermediate model preferably precedes the update time.

[0114] In one preferred embodiment, the active intermediate model is not modified before step 5). Thus, tooth movement is evaluated by comparing the updated analysis image with the intermediate model designed at the initial time.

[0115] In a preferred embodiment, the active intermediate model may be roughly corrected manually, for example, to take into account any modification of the patient's dental arch between the initial time and the update time, which is not due to drift during the course of treatment, in order to remove the tooth model of a tooth that has been extracted or pulled out.

[0116] In step 3), the display of teeth that do not conform to the treatment scenario in the updated analysis images, particularly teeth that have fallen outside the aligner, is investigated.

[0117] The centralized computer is preferably programmed to automatically detect and identify mismatched teeth.

[0118] Detection of incompatibility with the active intermediate model The computer may use analytical update images, particularly those of exposed teeth, to compare with an active intermediate model.

[0119] It is preferable to pursue the position, localization, and calibration of the virtual acquisition device (this position, localization, and configuration are collectively referred to as the "virtual acquisition conditions") that best correspond to the actual acquisition conditions of the updated images for analysis ("best fit").

[0120] Next, the view under the aforementioned virtual acquisition conditions is compared with the updated analysis image.

[0121] By comparing the aforementioned view with the updated analysis image, it is possible to detect not only misaligned teeth, i.e., teeth that are detached, but also teeth that are not detached but whose position does not correspond to the treatment scenario.

[0122] The comparison between the view and the updated analysis image may be performed, for example, by comparing corresponding maps with respect to distinguishing information representing the tooth outline. The comparison procedure described later may be used to compare the test map and the updated map.

[0123] The search for the virtual acquisition conditions and the comparison may be carried out specifically in accordance with the teachings of PCT / EP2015 / 074896.

[0124] In a preferred embodiment, a tooth representation in the analysis update image is considered unsuitable for the treatment scenario if at least one point of this representation, which has a 1:1 scale, is located at a distance greater than 1 / 10 mm, 3 / 10 mm, or 5 / 10 mm from a corresponding point in the view, which also has a 1:1 scale (i.e., a point representing the same point on the tooth), and is preferably less than 7 mm or 5 mm.

[0125] This distance may also be measured at the pixel level, which is advantageous because it eliminates the need to establish a scale.

[0126] Therefore, the advantage of mismatch is that it allows for the detection of drift in the execution of treatment. Specifically, distances that are too long are considered to be related not to drift in treatment, but rather to anomalies caused by poor tooth determination, for example, due to teeth being masked in the updated analysis image.

[0127] Out-of-line detection in aligner images, independent of active intermediate models. To detect misalignment using a computer, an analysis update image in the form of an aligner image may be analyzed to determine, in particular, the outer shape of the bottom of the grooves in the aligner and the outer shape of the free ends of the teeth.

[0128] Those skilled in the art know how to process an image or view to isolate its outline. This processing includes, for example, the application of known masks or filters supplied with image processing software. Such processing operations enable, for example, the detection of high-contrast areas.

[0129] These processing operations are specifically carried out in the following known preferred methods, namely: - Apply a Canny filter and explore edges using the Canny algorithm in particular. - Applying the Sobel filter and, in particular, calculating the derivative using the extended Sobel operator, - Applying a Laplace filter to calculate the Laplacian of an image, - Detecting blobs in images ("Blobdetector") - Apply a threshold value, applying the set threshold value to each element of the vector. - Resize using the relationship between pixel regions ("Resize(Area)") or perform bicubic interpolation in the vicinity of a pixel. - Shrinking an image using specific structured elements, - Expanding an image using specific structured elements, - In particular, perform modifications using the region adjacent to the restored region. - Apply a bilateral filter. - Apply Gaussian blur, - Apply Otsu's method to find a threshold that minimizes intraclass variance. - Apply the A* filter to find the path between points. - Apply an adaptive threshold and apply the adaptive threshold to the vector. - Applying an equalization filter, especially to the histogram of a grayscale image. - Detect blur ("BlurDetection") and use its Laplacian to calculate the entropy of the image. - Detecting the outline of a binary image ("FindContour") - This includes one or more of the following actions: performing a "floodfill," which involves filling connected elements with a determined color.

[0130] The following non-restrictive methods, namely, - Apply the "MeanShift" filter to find objects in the image projection. - Applying "CLAHE," where CLAHE stands for "Contrast Limited Adaptive Histogram Equalization." - Apply the "K-means" filter to determine the cluster center and the groups of samples around the cluster. - Applying a DFT filter to perform the Discrete Fourier Transform, Forward Fourier Transform, or Inverse Fourier Transform of a vector. - Calculating the moment, - Calculate the Hu moment invariant by applying the "HuMoments" filter. - Calculating the integral of an image, - By applying a Scharr filter and implementing the Scharr operator, it becomes possible to calculate the derivative of an image. - Searching for the convex hull of a point ("ConvexHull") - Searching for convex points in the outer shape ("ConvexityDefects") - Comparing shapes ("MatchShapes") - Checking whether there are points on the outline ("PointPolygonText") - Apply Harris corner detection ("CornerHarris") - Detecting corners by searching for the minimum eigenvalue of the gradient matrix ("CornerMinEigenVal") - Apply the Hough transform to find circles in a grayscale image ("HoughCircles") - "Active contour modeling" (tracing the outline of an object based on a potentially "noisy" 2D image), - Calculate the force field ("gradient vector flow" or GVF) in a portion of the image. - Applying a cascade classifier ("CascadeClassification") is not preferred, but may be done.

[0131] The determination of the tooth's external shape may be optimized according to the teachings of PCT / EP2015 / 074900.

[0132] In one embodiment, the outer shape of the groove bottom and the outer shape of the free end of the tooth are divided to define segments of these outer shapes for each tooth. The segments of the outer shape of the groove bottom and the segments of the outer shape of the free end of the tooth are respectively called "outer tooth outline" 24 i and "internal tooth outline" 30 iThese are called (Figures 12 and 14). The adjectives “inner” and “outer” are used here solely for clarity. In Figure 12, the dotted lines indicate the boundaries of consecutive segments.

[0133] Next, the comparison may be performed by any means, particularly using the technique for comparing the medial and lateral tooth profiles described in EP3412245.

[0134] Specifically, for each of the multiple teeth whose medial and lateral tooth shapes have been determined, the following steps are taken: i) A step of determining the distance between the inner tooth outline and the outer tooth outline, ii) Preferably a step of determining a distance threshold based on the distance determined in step i), iii) For each of the aforementioned teeth, - The distance between the inner tooth outline and the outer tooth outline and - It is possible to perform the step of determining the distance score according to the distance threshold.

[0135] In step i), the distance d between the inner tooth outline and the outer tooth outline is determined for each of the teeth (Figure 13).

[0136] The distance between the inner and outer tooth outlines of a tooth may be, for example, the average distance or the longest distance between pixels of the outlines corresponding to the same point on the tooth.

[0137] This distance is preferably measured in pixel units, which is advantageous as it eliminates the need to establish a scale.

[0138] In step ii), preferably, the distance threshold Sd is determined based on the distance determined in step i).

[0139] In step ii), it is preferable that the distance threshold Sd is substantially equal to the shortest distance among the distances determined in step i). min)。Conventionally, at least one of the teeth to be treated contacts the bottom of the groove into which the teeth are inserted. In that case, the distance between the inner tooth profile and the outer tooth profile of this tooth is the minimum distance d corresponding to the normal situation min and is equal to. Therefore, this distance may be used as a standard for evaluating the distance between the inner tooth profile and the outer tooth profile of the outer tooth in step iii).

[0140] In step iii), a score called the "distance score" S(d,Sd) is determined for each tooth - according to the distance d between the inner tooth profile and the outer tooth profile and - the distance threshold Sd.

[0141] The distance score of the tooth is preferably equal to (d - Sd), that is, equal to the difference between the distance between the inner tooth profile and the outer tooth profile of this tooth and the distance threshold. The higher the distance score, the more the affected tooth deviates from the groove.

[0142] FIG. 12 shows an example of the implementation of steps i) to step iii), in which the tooth D1 deviates from the bottom of the aligner to a point where d - d min > Sd.

[0143] For each of the plurality of teeth for which the inner tooth profile and the outer tooth profile have been determined, the following steps, namely i’) For each pair consisting of the left teeth adjacent to the right teeth of at least the first, second, and third adjacent tooth triplets, for each tooth for which the inner tooth profile and the outer tooth profile have been determined, the first and third teeth are adjacent to the second tooth, determine an offset called the "inner offset" between the inner tooth profile of the left tooth and the inner tooth profile of the right tooth, and determine an offset called the "outer offset" between the outer tooth profile of the left tooth and the outer tooth profile of the right tooth, then determine the difference between the inner offset and the outer offset called the "offset difference", and ii') Preferably a step of determining an offset difference threshold based on the offset difference determined in step i'), iii') It is also possible to perform the step of determining at least one offset score for at least one tooth, preferably for each of the three teeth, in accordance with the offset difference with the adjacent tooth and the offset difference threshold.

[0144] In step i'), at least one triplet consisting of the first, second, and third teeth D1, D2, and D3 is considered to be continuous along the dental arch, with the first and third teeth being adjacent to the second tooth.

[0145] The inner tooth outlines 301, 302, and 303 and the outer tooth outlines 241, 242, and 243 of teeth D1, D2, and D3 are determined, respectively.

[0146] Medial "offset" or lateral "offset" refers to the distance between the medial or lateral tooth contours of two adjacent teeth, respectively.

[0147] The following, namely, - "First Inner Offset" Δ 1-2 An offset between the inner tooth profile of the first tooth 301 and the inner tooth profile of the second tooth 302, referred to as i, - "Second inner offset" Δ 2-3 An offset between the inner tooth profile of the second tooth 302 and the inner tooth profile of the third tooth 303, referred to as i, - "First outer offset" Δ 1-2 An offset between the outer tooth profile of the first tooth 241 and the outer tooth profile of the second tooth 242, referred to as e. - "Second outer offset" Δ 2-3 An offset called e is determined between the outer tooth profile of the second tooth 242 and the outer tooth profile of the third tooth 243.

[0148] The medial offset between the medial tooth outlines of two adjacent teeth is preferably equal to the longest distance between the medial tooth outlines of these two teeth.

[0149] The lateral offset between the outer tooth profiles of two adjacent teeth is preferably equal to the longest distance between the outer tooth profiles of these two teeth.

[0150] The inner and outer offsets are preferably measured on a pixel-by-pixel basis, which is advantageous as it eliminates the need to establish a scale.

[0151] Next, the following, namely, - "First offset difference" Δ 1-2 ( =Δ 1-2 i-Δ 1-2 e) The first inner offset Δ 1-2 i and the first outer offset Δ 1-2 The difference between e, - "Second offset difference" Δ 2-3 ( =Δ 2-3 i-Δ 2-3 e) The second inner offset Δ 2-3 i and the second outer offset Δ 2-3 The difference between e and the given value is determined.

[0152] In the example in Figure 12, Δ 1-2 is, Δ 2-3 It's much smaller than that.

[0153] In step ii'), the offset difference SΔ is preferably the first and second offset difference Δ determined in step i'). 1-2 and Δ 2-3 The determination is made based on the following criteria.

[0154] In step ii'), it is preferable that the offset threshold is substantially equal to the smallest difference between the offsets determined in step i').

[0155] Traditionally, at least two adjacent teeth undergoing treatment are in contact with the bottom of the groove in which they are inserted. In this case, the offset difference between these two teeth undergoing treatment is virtually zero. Therefore, since this zero offset difference corresponds to a normal situation, this zero offset difference may be used as a standard to evaluate the offset difference between adjacent teeth undergoing treatment.

[0156] In Figure 13, the offset difference between the two teeth D1 and D2 is virtually zero.

[0157] In step iii'), for each pair of teeth in the triplet, at least one score called an "offset score" is determined according to the offset difference between the tooth and the adjacent tooth and the offset difference threshold.

[0158] Specifically, the offset difference between the first tooth and the second tooth may be compared with an offset difference threshold SΔ, where SΔ is, for example, zero. More specifically, the offset score of the first tooth and the second tooth may be determined by subtracting the offset difference threshold from the offset difference between the first tooth and the second tooth.

[0159] This offset score, if positive, for example, indicates that one or both of the first and second teeth are likely to be out of place at the bottom of the groove.

[0160] In Figure 13, the offset difference between tooth D2 and tooth D3 is positive, which indicates that the second or third tooth is dislodged.

[0161] In Figure 13, the offset difference between the two teeth D1 and D2 is virtually zero, and therefore, a positive offset difference between the two teeth D2 and D3 indicates that the third tooth is missing.

[0162] Generally, when the first offset score for the first and second teeth indicates that one of these two teeth is out of place, a second offset score is determined for the second tooth and the third tooth adjacent to it. If the second offset score is lower than the first offset score, it is likely that the first tooth is out of place at the bottom of the groove. Otherwise, it is likely that the second tooth is out of place.

[0163] Alternatively, the identification of misaligned teeth, particularly dislodged teeth, may be performed by an operator, preferably a dental professional, more preferably an orthodontist, simply by observing the aligner image displayed on a computer screen.

[0164] Analysis using deep learning devices A deep learning device, preferably a neural network, may be used to identify mismatched teeth in the updated images.

[0165] Specifically, this can be done as described in EP3432218.

[0166] The following steps, namely, I. The step of creating a training database in which each record includes a history image and a history description, and the database includes more than 1,000, preferably more than 5,000, preferably more than 10,000, preferably more than 30,000, preferably more than 50,000, preferably more than 100,000 history records, and each history image includes one or more zones representing one tooth, i.e., "history tooth zones", and the associated history description specifies a tooth attribute value for at least one tooth attribute for each history tooth zone, II. A step of training at least one deep learning device, preferably a neural network, with a training database, III. Transmitting an updated analysis image, preferably an aligner image, to the at least one deep learning device, thereby determining the attribute values ​​of the tooth represented within a zone representing at least one tooth in the updated analysis image, i.e., a "tooth analysis zone," and at least one probability for the attribute values ​​relating to the fit of the represented tooth. IV. It is preferable to perform the steps of determining whether the teeth of the dental arch are located in the positions represented by the tooth analysis zone, and determining the attribute values ​​of the teeth, according to the probability.

[0167] The deep learning device may be a neural network specifically designed for identifying and detecting objects in an image (i.e., an object detection network), and may be selected from the examples of such networks described above.

[0168] Deep learning devices not only enable the identification of teeth in updated analysis images, but also allow for the determination of whether or not the teeth are in proper fit.

[0169] In Step I, tooth attributes are attributes whose values ​​are unique to each tooth.

[0170] For example, the tooth attribute "out of place" has a value of "fit" or "not fit" depending on whether the tooth is properly positioned or abnormally positioned relative to the aligner.

[0171] In Step II, each record can be input into the deep learning device, which includes a history image and a description of each tooth represented in the history image, including the tooth's shape and any misalignments. The deep learning device then gradually learns to recognize patterns in the image and associates these patterns with tooth zones, as well as with tooth attribute values ​​for misalignments, particularly tooth detachment.

[0172] In step III, the deep learning device recognizes the pattern in the updated image. Specifically, the deep learning device, - The presence of a zone representing at least partially a tooth ("tooth analysis zone") at the location in the aforementioned analysis image, - The probability of the attribute value of the tooth represented within the tooth analysis zone may be determined.

[0173] For example, in an aligner image, a deep learning device can determine that there is a 99.5% probability that the shape in the analyzed image represents a tooth, and a 99% probability that this tooth is not aligned with the aligner.

[0174] It is preferable for the deep learning device to analyze the entire updated image and determine the probability for all identified tooth analysis zones.

[0175] In step IV, preferably by computer, for each tooth represented in the analysis update image, it is determined, according to the probability determined in step III, whether the represented tooth should be considered a aligner or a misaligned tooth. For example, if the probability of a tooth being misaligned is higher than a threshold, for example 98%, then this tooth may be considered to be out of the aligner.

[0176] Analysis and combination of aligner images and exposed tooth images Misaligned teeth are preferably detected by comparing images of exposed teeth with an active intermediate model and by analyzing only the aligner images. This improves the final result.

[0177] Specifically, when the analysis update image is an aligner image, certain outliers may be difficult to detect by analyzing only this image. Therefore, if only the analysis of the aligner image is considered, misfitted teeth may be considered to be fitted. Using images of exposed teeth and an active intermediate model may make it possible to detect misfits that cannot be detected by the aligner image alone.

[0178] If no misaligned teeth are detected, treatment may continue without modifying the treatment scenario. Therefore, the aligner designed in step d) is always suitable for treatment. It is preferable that an informational message be sent to the patient indicating that the treatment is progressing correctly. It is also preferable that an informational message be sent to the orthodontist treating the patient.

[0179] If one or more mismatched teeth are detected, the process proceeds to step 4).

[0180] In step 4), an active intermediate model tooth model is identified that represents the teeth identified in step 3 as misfit teeth, particularly detached teeth.

[0181] This identification may be performed by the active intermediate model and any computer that has access to identifiers of mismatched teeth, particularly detached teeth, determined in step 3). Such identifiers may be sent to a computer that has access to the active intermediate model, or vice versa, i.e., the active intermediate model may be sent to a computer that has access to the identifiers.

[0182] Preferably, in step 3), the active intermediate model is sent to a central computer that has analyzed the updated analysis image and detected misfitted teeth, particularly detached teeth, which then performs step 4).

[0183] As a variation, step 4) may be performed by an operator who has access to the active intermediate model.

[0184] In step 5), the active intermediate model is deformed until a configuration is found that fits the updated image taken at the update time. This updated image, called the "deformed updated image," is preferably an image of exposed teeth without aligners, i.e., an image of an exposed dental arch. This facilitates the deformation of the active intermediate model.

[0185] The deformation update image may be the same as or different from the analysis update image. The deformation update image is preferably acquired and transmitted by a mobile phone, preferably by an acquisition kit comprising a mobile phone and a holder to which a dental retractor is attached, as described above with respect to the analysis update image. The deformation update image is preferably an extraoral image.

[0186] Implementing the deformation achieved by acting on all tooth models of the active intermediate model is extremely time-consuming. In particular, the inventors have observed that this deformation does not result in a model that accurately represents the actual configuration of the dental arch during the update time. Specifically, the inventors have observed that in extraoral update images, the precise position of teeth (especially molars) in the posterior part of the mouth cannot always be determined when all tooth models are movable during deformation. Therefore, such deformations can result in substantial deviations from the active intermediate model in areas where the active intermediate model accurately represents the actual position of the teeth. Consequently, aligners manufactured based on such deformations may be unusable.

[0187] In one embodiment of the present invention, the active intermediate model is not deformed except in the areas of the tooth model where misfitted teeth, particularly detached teeth, are located. Thus, the general shape of the active intermediate model is preserved, and only these tooth models are moved.

[0188] Therefore, deformation of the active intermediate model is limited to the movement of the tooth model of the active intermediate model that represents misfitted teeth.

[0189] Specifically, by deforming the intermediate model by moving only the tooth models of misfitted teeth, the relative deviation of the misfitted teeth during the update time relative to the active intermediate model before deformation can be evaluated with great accuracy.

[0190] In a preferred embodiment, the movement of the tooth model is continued until the positional error of each tooth model relative to the deformed updated image is less than 1 mm, preferably less than 5 / 10 mm, preferably less than 3 / 10 mm, preferably less than 2 / 10 mm, and preferably less than 1 / 10 mm.

[0191] constraints It is preferable to pursue virtual acquisition conditions that correspond to the actual acquisition conditions of the deformed update image and enable the observation of the active intermediate model so that the view of the active intermediate model is as close as possible to the deformed update image. This search may be performed in step 3) if the deformed update image is an analysis update image.

[0192] In one embodiment, such virtual acquisition conditions are determined by considering all tooth models. Therefore, mismatched teeth negatively affect the accuracy of the virtual acquisition conditions. However, such teeth are few in number, and their impact is generally low.

[0193] It is preferable that the virtual acquisition conditions are determined without considering the tooth model of misfitted teeth, particularly detached teeth. Therefore, it is possible to determine "constrained" virtual acquisition conditions that accurately correspond to the actual acquisition conditions of the deformed update image and allow the active intermediate model to be observed such that the view of the active intermediate model, called the "constrained view," is very close to the deformed update image. The resulting constrained view is not degraded by the effects of misfitted teeth and is therefore particularly accurate for fitted teeth.

[0194] Specifically, in the latter embodiment, the representation of a fitted tooth in the deformed update image may be advantageously used as a reference whose position in the active intermediate model is known. Specifically, in the view obtained under the virtual acquisition conditions, such a reference has the same relative position as it has in the deformed update image.

[0195] Preferably, at least three misaligned points, such as the apex of a fitted tooth, particularly a tooth that is not misaligned, are used as references. In this case, analyzing the distances between these references in the deformed updated image makes it possible to evaluate the virtual acquisition conditions through a simple calculation.

[0196] Advantageously, in that case, the virtual acquisition conditions may be used to test the positions of all tooth models of misfitted teeth, particularly detached teeth.

[0197] Movement of misfitted tooth models via optimization The active intermediate model is preferably modified by an optimization algorithm.

[0198] Preferably, in each iteration, one or more misfitted teeth, particularly detached teeth, are moved in the tooth model, and the degree of fit between the thus modified active intermediate model and the deformed updated image is then evaluated, and the iteration continues until it is found that the modified model and the deformed updated image are well fitted.

[0199] The number of repetitions may be, for example, more than 10, more than 100, more than 1000, more than 10000, and / or less than 1,000,000.

[0200] The following steps, namely, A) A step of analyzing the deformed updated image and generating an updated image with respect to the distinction information, B) A step of searching for an updated model corresponding to the tooth position when a deformed updated image is acquired, based on the updated map and through the movement of the tooth model of misfitted teeth, in particular detached teeth, wherein the search is preferably carried out by an evolutionary metaheuristic method, preferably quasi-annealing.

[0201] After step A), the deformed updated image is analyzed to create an updated map with respect to at least one form of distinguishing information.

[0202] "Distinguishing information" is characteristic information (i.e., image features) that is sometimes extracted from an image through computer processing.

[0203] Distinction information may have a variable number of values. For example, outline information may be equal to 1 or 0 depending on whether a pixel belongs to a certain outline. Gloss information may take on a large number of values. Image processing makes it possible to extract and quantify the distinction information.

[0204] The updated map represents the distinction information within the reference frame of the deformed updated image. The distinction information is preferably selected from the group consisting of outline information, color information, density information, distance information, gloss information, saturation information, and information regarding the reflection and combination of these forms of information. The distinction information is preferably outline information.

[0205] The objective of step B) is to modify the active intermediate model until an updated model corresponding to the deformation is obtained. Therefore, ideally, the updated model is a 3D digital dental arch model, and if this 3D dental arch model is real, it is possible to obtain a deformed updated image based on this model, and if the aligner is represented in the deformed updated image, the aligner is ignored.

[0206] Therefore, various models are continuously tested, and the selection of the model to be tested preferably depends on the correspondence level between the already tested models and the deformed updated images. This selection is preferably made using an optimization method known in the art, particularly preferably an evolutionary metaheuristic optimization method, specifically, an optimization method selected from simulated annealing methods.

[0207] Preferably, the metaheuristic optimization method is - Preferably, an evolutionary algorithm selected from evolutionary strategies, genetic algorithms, differential evolution algorithms, distribution estimation algorithms, artificial immune systems, path relinking, shuffled complex evolution, simulated annealing, ant colony optimization algorithms, particle swarm optimization algorithms, tabu search, and GRASP methods, - kangaroo algorithms, - Fletcher-Powell methods, - noise methods, - probabilistic tunneling, - random-restart hill climbing, - cross-entropy methods, and selected from the group formed by hybrid methods combining the above-mentioned metaheuristic methods.

[0208] Preferably, step B) is the following steps, namely, B1) defining the model to be tested through the movement within the active intermediate model of the tooth models of the non-conforming teeth, particularly the missing teeth; B3) generating a view of the model to be tested under the constrained virtual acquisition conditions; B4) processing the view to generate at least one test map that at least partially represents the discrimination information; B5) A step of testing the model to be tested by comparing the updated map with the test map, thereby measuring the difference between the updated map and the test map, wherein this difference is called the "fit" or "match". B6) Depending on the difference, for example, if the difference is smaller than the threshold, - Modify the model to be tested by moving one or more teeth in the model that are not fitting, especially teeth that are out of place, and then return to step B3), or - Includes the step of defining the updated map as the tested model with the smallest difference from the updated map.

[0209] The measurement of the difference depends on the distinction information used. The difference may be measured, for example, by the ratio of the number of points belonging to both the outline of the test map and the outline of the updated map to the total number of points in the outline of the updated map, or by the product of the reciprocal of the average distance between the outlines represented in the updated map and the test map and the length of the outline represented in the updated map.

[0210] Therefore, the updated model obtained at the end of step B) is a model resulting from the continuous modification of the active intermediate model and is highly accurate because it is obtained by the deformation of the initial model. Thus, advantageously, the updated model is highly accurate even though it is based on simple photographs or video images taken without any particular attention.

[0211] Tooth models cannot penetrate each other. Therefore, the movement of the tooth model of a detached tooth, particularly in step 5), is restricted by the tooth model of a stationary, non-detached tooth. This further accelerates the search for the updated model.

[0212] By moving the tooth model of the active intermediate model, an updated model can be obtained that may be observed under conditions where the view of the updated model fits the deformed updated image. In other words, this view may coincide with and overlap the deformed updated image, thereby the teeth represented in the view and the teeth represented in the deformed updated image overlap substantially exactly.

[0213] The updating of the active intermediate model may be refined by repeating the previous operation using multiple updated images as analyzed and / or deformed updated images. This yields an updated model that represents the teeth, which is substantially the actual configuration at update time. As an alternative to or addition to the use of optimization methods, the search for the updated model is performed using deep learning devices, and preferably neural networks.

[0214] Movement of the tooth model of ill-fitting teeth according to the assessment of non-fitness. In one embodiment, non-compliance can be quantified by "degree of non-compliance" through analysis of the updated analysis image.

[0215] Specifically, in one embodiment, the variation in detachment along the edge of a detached tooth, preferably a detached tooth, can be measured by analyzing the updated analysis image.

[0216] This information regarding non-conformity is preferably used in step 5) to move the tooth model of the mis-conformed teeth, particularly the detached teeth. For example, in Figure 12, the measurement of d is used to move, for example, the translation dd min The tooth D1 may be moved downward via this method.

[0217] It is particularly useful to use the movement of the tooth model in accordance with the nature and amplitude of the mismatch, which is evaluated based on the updated analysis image, as the initial movement in step 5). After this rough movement, it is preferable to perform finer movement, preferably through optimization or using a deep learning device, preferably a neural network.

[0218] In step 6), at least one post-update aligner is designed that is suitable for modifying the dental arch from its actual configuration at update time toward the final configuration. "Along toward the final configuration" means that the post-update aligner is shaped to modify the dental arch configuration toward the final configuration. However, multiple new aligners may be required to reach the final configuration.

[0219] It is preferable that the new set of aligners be designed with the updated model in mind rather than the initial model. Steps a) to d) described above are used to do this, and it is preferable that the initial time and initial model are replaced with the updated time and updated model, respectively.

[0220] In step 7), at least one or more aligners from the new series of aligners are preferably manufactured in the same manner as in step e).

[0221] The aligner may be manufactured, for example, according to the teachings of EP1835864.

[0222] These new aligners can be given to patients, for example, by mail.

[0223] The patient then continues treatment with these new aligners.

[0224] system The method according to the present invention is carried out at least partially, preferably entirely, by a computer. Any computer, specifically a PC, server, or tablet, may be envisioned.

[0225] Conventionally, a computer includes, in particular, a processor, a memory, a human-machine interface that conventionally includes a keyboard, a screen, and a mouse, a module for communication via the Internet, via WiFi, via Bluetooth® or via a telephone network, and a communication bus. The memory conventionally includes a ROM memory and a RAM memory. A software package configured to implement part of the method of the present invention is loaded into the memory of the computer.

[0226] <000097​​​​​​​​​​​​​​​​​​​​​​​​​​​​​​

[0229] The operator may specifically be a dental professional, preferably an orthodontist. The computer may implement one or more deep learning devices, preferably a neural network.

[0230] Figure 6 shows a system in a preferred embodiment of the present invention.

[0231] This system includes multiple mobile phones belonging to patient P, each communicating with a central computer 50, for example, more than 1,000, preferably more than 10,000 mobile phones 21.

[0232] The centralized computer 50 is preferably configured to receive and process updated images, particularly aligner images Ig and exposed tooth images Id, from multiple patients, preferably more than 100, more than 1,000, or more than 10,000 patients. The centralized computer 50 may also be configured to receive and process updated images from patients, all of whom are being treated by the same orthodontist. The centralized computer 50 is preferably configured to receive and process updated images from multiple different orthodontists, for example more than 10, more than 100, or more than 1,000 orthodontists.

[0233] The central computer 50 includes a communication module for communicating with a plurality of local computers 52, preferably located within the orthodontic clinic, for example, more than 10 or more than 100 local computers, via Wi-Fi, Bluetooth®, optical fiber, or telephone network.

[0234] The system also comprises multiple scanners 54, each scanner 54 communicating with one or more local computers, preferably each scanner communicating with a single local computer, for example, via a wired link, via Wi-Fi, via Bluetooth®, via optical fiber, or via a telephone network. Each scanner 54 is preferably located in the same location as its respective local computer, preferably within the same orthodontic clinic.

[0235] The system also includes a manufacturing unit equipped with a manufacturing computer 56 that communicates with a central computer and / or local computers, for example, via a wired link, via Wi-Fi, via Bluetooth®, via optical fiber, or via a telephone network.

[0236] Examples In step a), an initial model M0 is generated using the scanner 54 during the initial time. The initial model M0 is then sent to the local computer 52. A software package loaded onto the local computer preferably enables the initial model to be automatically cut to create a tooth model.

[0237] After examining the patient, the orthodontist moves the tooth model using a local computer to create the final model M f Define the area.

[0238] Preferably, in steps b) and c), the local computer 52 is programmed to determine one or more scenarios for modifying the dental arch so that the dental arch reaches a final configuration corresponding to the final model. Preferably, the local computer also allows the orthodontist to visualize the potential scenarios and select a treatment scenario. Preferably, the orthodontist can also create or modify the scenarios proposed by the local computer 52.

[0239] Intermediate times and corresponding intermediate models may be defined by a local computer or proposed to the orthodontist by a local computer 52 so that the orthodontist can verify and / or modify the intermediate times and intermediate models.

[0240] Local computer 52 is used for treatment scenarios, particularly at least the intermediate model M. i Also, the initial and final models M0, M f This is sent to the central computer 50.

[0241] In step d), the software package loaded into the central computer 50 determines the shape of the aligner to be manufactured based on these models, and then transmits this information I0 to the manufacturing unit for the purpose of manufacturing the aligner G0 in step e).

[0242] Alternatively, the initial model, intermediate model, and final model may be transmitted by the central computer 50 to the manufacturing computer 56, which in step d) determines the shapes of N aligners and controls the manufacturing process.

[0243] Alternatively, step d) may be performed by the local computer 52. The local computer then transmits the information necessary to manufacture the aligner to the manufacturing unit.

[0244] The Alaina G0 is sent to the patient, and the patient begins treatment.

[0245] The patient preferably receives a reminder requesting them to take one or more updated images, preferably at least one aligner image and preferably one or more images of exposed teeth.

[0246] In step 1), the patient takes photographs at the update time, both with and without the active aligner that needs to be worn at that time. The patient uses their mobile phone 21 for this purpose and transfers these photographs to the central computer 50.

[0247] In step 2), the central computer identifies the active intermediate model according to the update time of the photograph acquisition. Preferably, the central computer identifies the intermediate model whose intermediate time is closest to the update time as the active intermediate model. More preferably, the central computer identifies the intermediate model used to design the active aligner that was fitted by the patient at the update time as the active intermediate model.

[0248] In step 3), the central computer 50 analyzes photographs, particularly photographs showing the aligner in use, with the aim of automatically detecting non-fitting teeth, especially loose teeth, and identifying teeth that do not fit properly, especially loose teeth. If no teeth that do not fit properly, especially loose teeth, are detected, the central computer 50 sends a message to the patient and / or orthodontist to inform them that the treatment is progressing normally.

[0249] The centralized computer 50 is preferably programmed to identify such teeth.

[0250] If not programmed to do so, in step 4), the central computer 50 may, in some cases, with the help of the operator, identify the tooth model of misfitted teeth, especially detached teeth.

[0251] In step 5), the central computer matches the updated Model M to the photos taken at the time of the update. a The active intermediate model is modified by moving the tooth model of mismatched teeth, especially detached teeth, until a suitable match is found.

[0252] In step 6), the central computer designs one or more new aligners G1 taking the updated model into consideration, and then transmits the information necessary to manufacture the aligners G1 to the manufacturing computer 56 of the manufacturing unit.

[0253] In step 7), the manufacturing unit produces a new aligner G1. The new aligner is then sent to the patient, and the patient may continue treatment with these new aligners.

[0254] Variation In one embodiment, the method according to the present invention uses an aligner image to update the determined active intermediate model at an initial time and an intermediate time t indicating the replacement of the aligner. i This represents the dental arch with the expected configuration. Therefore, the intermediate model of the treatment scenario is each intermediate time t that the patient is required to replace the aligner. n This represents the expected dental arch configuration. Therefore, at the start of treatment, i.e., at the initial time t1 when the initial model is generated, the patient begins to wear the first aligner of the series of aligners, and then changes the aligners at intermediate times t2, t3, etc. Therefore, at intermediate time t i t is the time at which it is expected to replace the (i-1)th orthodontic aligner in a series with the ith orthodontic aligner in a series, in which case i is 2 or greater. If, for example, 30 aligners are planned to be used in the treatment (N=30), the patient will have time t 30 We began installing the 30th aligner, and this aligner was the final one. 31 Attach it until it's attached.

[0255] The number of aligners N may be more than 5, more than 10, more than 20, or more than 30, and / or less than 60, preferably less than 50.

[0256] The time interval between two consecutive aligner changes, i.e., between two consecutive intermediate periods, may be greater than 7 days, greater than 15 days, and / or less than 60 days, preferably less than 30 days.

[0257] In one embodiment, the treatment scenario is not limited to a series of intermediate models for aligner replacement, but includes other intermediate models where the intermediate time does not indicate aligner replacement. Preferably, the treatment scenario is a substantially continuous series of intermediate models. Thus, the treatment scenario is similar to a video that allows visualization of the progression from the initial model to the final model.

[0258] Therefore, advantageously, the suitability of the active aligner may be examined at any time. Update images may be taken at any update time during treatment. Then, depending on the treatment scenario, an active intermediate model is selected that corresponds to the update time, i.e., with respect to the update time, represents the dental arch in the configuration expected according to the treatment scenario.

[0259] In one embodiment, an additional intermediate model may be generated from the intermediate model of the treatment scenario and added to the treatment scenario. Specifically, the update time is two intermediate time t i and t i+1 If it is between, for example by interpolation, the intermediate time t i and t i+1 From the intermediate model, an additional intermediate model may be created to act as an active intermediate model. This additional intermediate model may be generated, in particular, during treatment.

[0260] As is now clear, the method according to the present invention makes it possible to obtain a highly accurate updated model corresponding to the actual configuration of the dental arch at the time of update, based on a simple photograph or video, without the need to perform a new scan. Therefore, this method may be carried out without the need to make an appointment with an orthodontist.

[0261] Of course, the present invention is not limited to the embodiments described and illustrated above.

[0262] Orthodontic treatment may be for therapeutic and / or cosmetic purposes.

[0263] Multiple updated images may be used in step 3) and / or step 5).

[0264] Tooth model placement errors may be used to detect drift in orthodontic treatment, i.e., to detect situations where the progression of tooth position does not follow the treatment scenario.

[0265] Finally, the patient is not necessarily human. Specifically, the method according to the present invention may be used on other animals. [Explanation of Symbols]

[0266] 10. Orthodontic Aligner 12 grooves 15 Photography Kit 17 Holder 19 Dental retractors, retractors 20 bottom 21 Mobile phone 22 Free end 23 Groove 241 Outer tooth outline, first tooth 242 Outer tooth outline, second tooth 243 Outer tooth outline, third tooth twenty four i External tooth shape 26a, 26b lag 27a, 27b clips 301 Inner tooth outline, first tooth 302 Inner tooth outline, second tooth 303 Inner tooth outline, third tooth 30 i Medial tooth shape 32 Tooth Model 50 Centralized Computer 52 Local Computer 54 Scanners 56 Manufacturing Computers D1, D2, D3 teeth d distance d min Shortest distance G0, G1 Alaina I0 Information ID: Image of exposed teeth Ig Alaina image M0 initial model M i Intermediate model M f Final model M a Updated model P patient Sd Distance Threshold S(d,Sd) Distance score t1 initial time t n Intermediate time t N+1 Last time X-axis

Claims

1. A method for generating a three-dimensional digital model of a patient's dental arch, called the "updated model," during the treatment of the dental arch using an orthodontic aligner called the "active aligner," wherein the treatment is performed during the initial time (t １ ) generates multiple intermediate models (M ｉ The treatment scenario is simulated by including the following: each intermediate model is a three-dimensional digital model of the dental arch, the intermediate model is cut to create a tooth model, and each intermediate time (t) after the initial time ｉ In a method determined to represent the dental arch in ), 1) A step of acquiring at least one updated image by an image acquisition device during the update time in the process of the treatment, wherein each updated image is an aligner image (Ig) representing the active aligner attached to the dental arch at the position of use, or an image of an exposed tooth (Id) representing the dental arch without an aligner, 2) A step of determining the intermediate model or "active intermediate model" according to the update time, 3) A step of searching for the representation of one or more teeth that do not fit the treatment scenario in the updated image called the “analysis update image”, If one or more mismatched teeth are detected, 4) The steps of identifying one or more tooth models that represent one or more mismatched teeth in the active intermediate model, 5) A step of automatically deforming the active intermediate model by computer until an updated model that conforms to at least one updated image called a "deformed updated image" is obtained, wherein during the deformation of the active intermediate model, the tooth models moved are only those of teeth that do not conform. Includes, In the above method, the non-fit of at least one of the misfit teeth is measured by comparing the updated image with the active intermediate model, and then, in step 5), the tooth model of the at least one misfit tooth is moved in accordance with the measurement. Steps 2) to 5) are performed by the computer, in this manner.

2. In step 3), in order to detect mismatched teeth, the position, localization, and calibration of the virtual acquisition device are pursued so that the virtual acquisition device has a view on the active intermediate model that is as close as possible to the analysis update image, and then, The view and the analysis update image are compared, or the updated map representing the distinction information of the analysis update image is compared with the reference map representing the distinction information in the view, and / or The method according to claim 1, wherein the analysis update image is an aligner image, and the external shape of at least one tooth and the external shape of the active aligner are determined in the analysis update image, and then the external shapes are compared with each other.

3. In step 5), the movement of the tooth model of the misfitted tooth is an iterative process, and in each iteration, One or more of the tooth models of the mismatched teeth are moved to obtain the dental arch model to be tested, and then, The dental arch model to be tested is tested by evaluating the degree of fit between the dental arch model and the deformed updated image. The method according to claim 1, wherein the updated model is the model having the best fit among all the tested models.

4. Prior to the iterative process, the position, localization, and calibration of the virtual acquisition device, or "constrained virtual acquisition conditions," are pursued to enable observation of the active intermediate model to generate a view in which the representation of the fitted teeth in the deformed updated image can coincide with and overlap the representation of the fitted teeth, and then, The method according to claim 3, wherein, during the iterative process, in each iteration, the degree of fit between the dental arch model under test and the deformed updated image is evaluated by comparing the deformed updated image with a view of the dental arch model under test under the constrained virtual acquisition conditions.

5. The method according to any one of claims 1 to 4, wherein in step 1), the updated image is an image extracted from a photograph or video and is acquired using a mobile phone.

6. The method according to any one of claims 1 to 5, wherein in step 5), the deformation of the active intermediate model includes movement of the tooth model of the active intermediate model, the movement is continued until the positional error for each tooth model, taking into account the deformation update image, is less than 1 mm, preferably less than 5 / 10 mm, preferably less than 3 / 10 mm, preferably less than 2 / 10 mm, preferably less than 1 / 10 mm.

7. The method according to any one of claims 1 to 6, wherein in step 5) the deformed updated image is an image of an exposed tooth.

8. The method according to any one of claims 1 to 7, wherein the active intermediate model is selected such that the difference between the intermediate time and the update time of the active intermediate model is less than two weeks.

9. The method according to any one of claims 1 to 8, wherein the intermediate model of the treatment scenario represents the dental arch in the configuration expected at an intermediate time showing aligner replacement.

10. The method according to any one of claims 1 to 9, wherein the treatment scenario is a substantially continuous series of intermediate models.

11. The method according to any one of claims 1 to 10, wherein, prior to step 2), an intermediate model of the dental arch is generated from an intermediate model of the treatment scenario and then added to the treatment scenario via the intermediate model.

12. The method according to any one of claims 1 to 11, wherein, in step 3), the representation of a tooth in the analysis update image is not suitable for the treatment scenario if, when the analysis update image coincides with and overlaps with the view of the active intermediate model that is adapted to the analysis update image which is at the same scale and actual size as the view of the active intermediate model, at least one point of the representation is located at a distance of more than 1 / 10 mm and less than 7 mm from a corresponding point in the view.

13. A method for manufacturing an orthodontic aligner, comprising a method for generating an updated model according to any one of claims 1 to 12, 6) The steps of designing an “updated” aligner suitable for modifying the dental arch from the actual configuration at the update time to the theoretical final configuration, based on the updated model and the final model representing the dental arch in the theoretical final configuration, and transmitting the information necessary to manufacture the updated aligner to a manufacturing unit, 7) A step of manufacturing the updated aligner using the manufacturing unit, Includes, The method is such that step 6) is performed automatically by a computer.

14. The method according to claim 1, wherein step 2) is performed following step 3).

Images