Quality control system based on appliance image

Abstract

A method of inspecting a manufacturing defect of a dental appliance may include obtaining an image of the dental appliance, identifying an identifier of the dental appliance, determining a digital file associated with the dental appliance from digital files based on the identified identifier, the digital file associated with the dental appliance includes a digital model of an intermediate component used in the dental appliance manufacturing process, determines an expected characteristic of the dental appliance by digitally manipulating the digital model of the intermediate component used in the dental appliance manufacturing process, determines an actual characteristic of the dental appliance from an image of the dental appliance, and displays the actual characteristic of the dental appliance. Determining whether a manufacturing defect is present in the dental appliance by comparing the expected characteristic of the dental appliance to the actual characteristic of the dental appliance, and outputting an output associated with the determination of whether a manufacturing defect is present.

Description

[0001] This application is a divisional application of the invention patent application with Chinese Application No. 202210280323.5, filing date of September 28, 2018, and invention title of "Quality Control System Based on Appliance Images". Application 202210280323.5 is a divisional application of the invention patent application with Chinese Application No. 201880074124.0 (corresponding to PCT International Application No. PCT / US2018 / 053564), filing date of September 28, 2018, and invention title of "Quality Control System Based on Appliance Images". Technical Field

[0002] Embodiments of the present invention relate to the field of manufacturing customized products, and in particular, to image-based quality control systems and methods for customized manufactured products. Background Art

[0003] For some applications, a shell is formed around a mold to achieve the female surface of the mold. The shell is then removed from the mold for further use in various applications. An example application where a shell is formed around a mold and then used later is orthodontic or orthodontic treatment. In such an application, the mold can be a male mold of a patient's dental arch, and the shell can be an aligner for aligning one or more of the patient's teeth. When using attachments, the mold can also include features related to planned orthodontic attachments and virtual fillings.

[0004] The mold can be formed using casting or rapid prototyping devices. For example, a 3D printer can use additive manufacturing techniques (e.g., stereolithography) or subtractive manufacturing techniques (e.g., milling) to manufacture the mold. The aligner can then be formed on the mold using a thermoforming device. Once the aligner is formed, it can be trimmed manually or automatically. In some cases, a computer-controlled 4-axis or 5-axis trimming machine (e.g., a laser trimming machine or a grinder) is used to trim the aligner along the cutting lines. The trimming machine uses electronic data identifying the cutting lines to trim the aligner. Thereafter, the aligner can be removed from the mold and delivered to the patient. Although much of this process has been automated, there can be further improvements. Summary of the Invention

[0005] A first aspect of the present disclosure may include a method that includes: receiving a digital file associated with a plastic shell customized for a patient's dental arch; generating a first image of the plastic shell using one or more imaging devices; determining an inspection plan for the plastic shell based on at least one of first information associated with the first image of the plastic shell or second information associated with the digital file. The inspection plan specifies one or more additional images of the plastic shell to be generated. The method may further include executing the inspection plan to capture one or more additional images of the plastic shell, determining whether one or more defects are included in the plastic shell at least in part based on the one or more additional images, and performing quality control of the plastic shell in response to determining that one or more defects are included in the plastic shell.

[0006] A second aspect of the present disclosure may further expand the first aspect of the present disclosure. In the second aspect of the present disclosure, the first image is a top view image of the plastic shell, and determining the inspection plan includes using at least one of the first information or the second information to determine one or more features of the plastic shell, and determining one or more additional images to be generated based on the one or more features. The one or more features include at least one of an exact cut line of the plastic appliance, a cavity of the plastic appliance associated with an attachment, an angle of the cut line of the plastic appliance, a distance between cavities of the plastic appliance associated with teeth, or a distance between cavities of the plastic appliance associated with attachments. Determining the inspection plan may further include determining the size of the plastic shell from at least one of the first information or the second information, and determining settings for generating one or more additional images based on at least one of the one or more features or the size of the plastic shell. The settings may include at least one of an orientation of one or more imaging devices, a zoom of one or more imaging devices, or a focus of one or more imaging devices. Executing the inspection plan may include capturing one or more additional images using the settings.

[0007] A third aspect of the present disclosure may further expand the first and / or second aspects of the present disclosure. The third aspect of the present disclosure may include applying the digital file as an input to a model, generating, by the model, an output that identifies one or more locations of the plastic shell, the one or more locations being identified as high-risk areas for one or more defects, and determining one or more additional images based on the one or more locations identified as high-risk areas by the output.

[0008] The fourth aspect of the present disclosure can further expand the first to third aspects of the present disclosure. The fourth aspect of the present disclosure can include applying a digital file as an input to a trained machine learning model. The trained machine learning model is trained to identify one or more high-risk regions of one or more defects at one or more locations of a plastic housing. The fourth aspect can also include generating an output by the trained machine learning model, the output identifying one or more locations of the plastic housing that are identified as high-risk regions of one or more defects, and determining one or more additional images based on the one or more locations identified as high-risk regions by the output.

[0009] The fifth aspect of the present disclosure can further expand the first to fourth aspects of the present disclosure. The fifth aspect of the present disclosure can include applying a digital file as an input to a prediction model. The prediction model performs finite element analysis using the geometry of the plastic housing to calculate one or more values of one or more strains of the plastic housing, and identifies one or more high-risk regions of one or more defects at one or more locations based on one or more values of one or more strains at one or more locations of the plastic housing exceeding a threshold. The fifth aspect can also include: generating an output by the prediction model that identifies one or more locations of the plastic housing that are identified as high-risk regions of one or more defects, and determining one or more additional images based on the one or more locations identified as high-risk regions by the output.

[0010] The sixth aspect of the present disclosure can further expand the first to fifth aspects of the present disclosure. The sixth aspect of the present disclosure can include: determining that an inspection scheme includes applying a digital file to a rule engine that uses one or more rules, the one or more rules specifying generating one or more additional images of the plastic housing when one or more features are included in the plastic housing at a location; and performing the inspection scheme includes capturing one or more additional images.

[0011] The seventh aspect of the present disclosure can further expand the sixth aspect of the present disclosure. The seventh aspect of the present disclosure can include generating one or more rules based on at least one of: a) historical data that includes reported defects of a first set of plastic housings and locations of the reported defects on the first set of plastic housings, b) digital files of a second set of plastic housings with labels indicating whether each of the second set of plastic housings has been subjected to a defect, or c) digital files of a third set of plastic housings with labels indicating the likelihood of the presence of a defect in each of the third set of plastic housings.

[0012] The eighth aspect of the present disclosure can further expand the first to seventh aspects of the present disclosure. In the eighth aspect of the present disclosure, determining the inspection plan includes obtaining the inspection plan from a memory location. The settings of one or more imaging devices for generating one or more additional images are preset in the inspection plan obtained from the memory location. The settings include at least one of one or more positions of one or more imaging devices for generating one or more additional images, one or more orientations of one or more imaging devices for capturing one or more additional images, one or more depths of focus of one or more imaging devices for capturing one or more additional images, or the number of one or more additional images of one or more imaging devices.

[0013] The ninth aspect of the present disclosure can further expand the first to eighth aspects of the present disclosure. In the ninth aspect of the present disclosure, performing the inspection plan to capture one or more additional images further includes: using a first imaging device among one or more imaging devices to track the edge of the plastic housing using data from the design file of the plastic housing to capture an image subset of one or more additional images representing the cutting line of the plastic housing.

[0014] The tenth aspect of the present disclosure can further expand the first to ninth aspects of the present disclosure. In the tenth aspect of the present disclosure, determining whether one or more defects are included in the plastic housing based at least in part on one or more additional images further includes obtaining a digital model of the plastic housing from a digital file associated with the plastic housing, determining an approximate first characteristic of the plastic housing from the digital model of the plastic housing, determining a second characteristic of the plastic housing from the first image, and comparing the approximate first characteristic with the second characteristic.

[0015] The eleventh aspect of the present disclosure can further expand the first to tenth aspects of the present disclosure. The eleventh aspect of the present disclosure can include performing an inspection plan to capture one or more additional images of the plastic housing, including configuring the settings of one or more imaging devices to capture one or more images based on at least one of first information or second information. The settings include at least one of the orientation of one or more imaging devices, the position of one or more imaging devices, the zoom of one or more imaging devices, or the depth of focus of one or more imaging devices.

[0016] The twelfth aspect of the present disclosure can further expand the first to eleventh aspects of the present disclosure. In the twelfth aspect of the present disclosure, the first image of the plastic housing includes at least one of a photographic image, an X-ray image, or a digital image, and one or more imaging devices include at least one of a camera, a blue laser scanner, a confocal microscope, a stereo image sensor, an X-ray device, or an ultrasonic device.

[0017] The thirteenth aspect of the present disclosure may include a method that includes: obtaining one or more images of a first shell customized for a patient's dental arch; using the one or more images to identify an identifier on the first shell; and determining a first digital file associated with the first shell from a set of digital files based on the identifier, and determining an approximate first characteristic of the first shell from the first digital file. The approximate first characteristic is based on the manipulation of a digital model of a mold used to create the first shell. The method further includes determining a second characteristic of the first shell from the one or more images, comparing the approximate first characteristic with the second characteristic, and performing quality control on the first shell based on the comparison.

[0018] The fourteenth aspect of the present disclosure may extend the thirteenth aspect of the present disclosure. In the fourteenth aspect of the present disclosure, the digital file includes a digital model of the first shell. Additionally, the method further includes generating a digital model of the first shell by performing the following: expanding the digital model of the mold into an expanded digital model, simulating the process of thermoforming a film on the digital model of the mold, calculating the projection of a cutting line onto the expanded digital model, virtually cutting the expanded digital model along the cutting line to create a cut-expanded digital model, and selecting the outer surface of the cut-expanded digital model.

[0019] The fifteenth aspect of the present disclosure may extend the thirteenth and / or fourteenth aspects of the present disclosure. In the fifteenth aspect of the present disclosure, the digital file includes a digital model of a mold used to create the first shell. Additionally, the method further includes manipulating the digital model of the mold to determine the approximate first characteristic.

[0020] The sixteenth aspect of the present disclosure may extend the thirteenth aspect to the fifteenth aspect of the present disclosure. In the sixteenth aspect of the present disclosure, the approximate first characteristic includes the approximate outer surface of the first shell, and the second characteristic includes the first shape of the first shell.

[0021] The seventeenth aspect of the present disclosure may extend the sixteenth aspect of the present disclosure. In the seventeenth aspect of the present disclosure, the one or more images include a top view image. Additionally, the method further includes: determining a first plane associated with the top view image, calculating a first projection of the approximate outer surface of the first shell onto the first plane, identifying one or more differences between a second shape of the first projection and the first shape based on the comparison, and determining whether the one or more differences exceed a first threshold.

[0022] The eighteenth aspect of the present disclosure may extend the seventeenth aspect of the present disclosure. In the eighteenth aspect of the present disclosure, the method further includes determining whether the first shell is deformed based on whether the difference exceeds the first threshold.

[0023] The nineteenth aspect of the present disclosure can extend the seventeenth aspect to the eighteenth aspect of the present disclosure. In the nineteenth aspect of the present disclosure, identifying one or more differences includes determining one or more regions where the first shape and the second shape do not match, and determining at least one of the thickness of the one or more regions or the area of the one or more regions.

[0024] The twentieth aspect of the present disclosure can extend the seventeenth aspect to the nineteenth aspect of the present disclosure. In the twentieth aspect of the present disclosure, the digital file includes a first digital model of a first housing, the first housing including approximate first characteristics, and the first digital model of the first housing has been generated based on the manipulation of the digital model of the mold. The method further includes: determining a stationary position of the first digital model of the first housing on a flat surface, and calculating a projection of the first digital model with the stationary position onto a first plane of a top view image.

[0025] The twenty - first aspect of the present disclosure can extend the twentieth aspect of the present disclosure. In the twenty - first aspect of the present disclosure, determining the stationary position of the first digital model on a flat surface includes: determining the centroid of the first digital model of the first housing, determining the convex hull of the first digital model, the convex hull including a set of vertices that link the outermost points of the first digital model, and for at least one vertex in the set of vertices, performing the following, including: calculating a line that contains the at least one vertex, calculating a projection of the centroid onto a point on the line, determining whether the point on the line is outside at least one vertex, and in response to determining that the point is not outside at least one vertex, determining that the at least one vertex is the stationary position of the first digital model.

[0026] The twenty - second aspect of the present disclosure can extend the twentieth to the twenty - first aspects of the present disclosure. In the twenty - second aspect of the present disclosure, determining the stationary position of the first digital model on a flat surface includes: determining the centroid of the first digital model of the first housing, determining the convex hull of the first digital model, the convex hull including a set of faces that link the outermost points of the first digital model, and for at least one face in the set of faces, performing the following, including: calculating a plane that contains the at least one face, calculating a projection of the centroid onto a point on the plane, determining whether the point on the plane is outside at least one face, and in response to determining that the point is not outside at least one face, determining that the at least one face is the stationary position of the first housing.

[0027] The twenty-third aspect of the present disclosure can extend the seventeenth to twenty-second aspects of the present disclosure. In the twenty-third aspect of the present disclosure, the method includes: determining that one or more differences do not exceed a first threshold, generating a modified projection of an approximate outer surface of the first housing by deforming a second shape of a first projection such that a first curvature of the deformed second shape approximately matches a second curvature of the first shape, identifying one or more additional differences between the first curvature of the deformed second shape and the second curvature of the first shape, and determining whether the one or more additional differences exceed a second threshold.

[0028] The twenty-fourth aspect of the present disclosure can extend the twenty-third aspect of the present disclosure. In the twenty-fourth aspect of the present disclosure, the method includes: identifying one or more regions where a second curvature and a first curvature do not match. The one or more regions correspond to cutting lines of the first housing. The method further includes determining whether the cutting lines of the first housing will interfere with the fitting of the first housing on a dental arch of a patient.

[0029] In the twenty-fifth aspect of the present disclosure, which can extend the twenty-third or twenty-fourth aspect of the present disclosure, deforming the first shape of the first projection includes: identifying a center line of the first projection; calculating projections of a set of lines that perpendicularly intersect the center line; identifying points on each respective line of the set of lines at intersection points between each respective line of the set of lines and the center line; and moving the points along the set of lines such that a first curvature of the deformed second shape approximately matches a second curvature of the first shape.

[0030] The twenty-sixth aspect of the present disclosure can extend the seventeenth to twenty-fifth aspects of the present disclosure. In the twenty-sixth aspect of the present disclosure, one or more images further include side view images. Additionally, the method further includes: generating a modified projection of an approximate outer surface of the first housing by deforming a second shape of a first projection such that a first curvature of the deformed second shape approximately matches a second curvature of the first shape, determining a second plane associated with the side view image, deforming the approximate outer surface of the first housing according to the deformation of the second shape of the first projection, calculating a second projection of the deformed approximate outer surface of the first housing onto the second plane, identifying one or more additional differences between a third shape of the first housing represented in the side view image and a fourth shape represented in the second projection, and determining whether the one or more additional differences exceed a second threshold.

[0031] The twenty-seventh aspect of the present disclosure may include a non-transitory computer-readable medium storing instructions that, when executed by a processing device, cause the processing device to: obtain one or more images of a first shell customized for a patient's dental arch; identify an identifier on the first shell using the one or more images; determine a first digital file associated with the first shell from a set of digital files based on the identifier, determine an approximate first characteristic of the first shell from the first digital file, wherein the approximate first characteristic is based on manipulation of a digital model of a mold used to create the first shell, determine a second characteristic of the first shell from the one or more images, compare the approximate first characteristic with the second characteristic, and perform quality control on the first shell based on the comparison.

[0032] The twenty-eighth aspect of the present disclosure may further expand the twenty-seventh aspect of the present disclosure. In the twenty-eighth aspect of the present disclosure, the digital file includes a digital model of the first shell. Additionally, the processing device further generates a digital model of the first shell by performing the following operations, including: expanding the digital model of the mold into an expanded digital model, simulating the process of thermoforming a film on the digital model of the mold, calculating the projection of the cutting line onto the expanded digital model, virtually cutting the expanded digital model along the cutting line to create a cut expanded digital model, and selecting the outer surface of the cut expanded digital model.

[0033] The twenty-ninth aspect of the present disclosure may further expand the twenty-eighth aspect of the present disclosure. In the twenty-ninth aspect of the present disclosure, the approximate first characteristic includes the approximate outer surface of the first shell, and the second characteristic includes the shape of the first shell.

[0034] The thirtieth aspect of the present disclosure may include a system that includes a memory storing instructions and a processing device coupled to the memory. Executing the instructions causes the processing device to: obtain one or more images of a first shell customized for a patient's dental arch; identify an identifier on the first shell using the one or more images; determine a first digital file associated with the first shell from a set of digital files based on the identifier, determine an approximate first characteristic of the first shell from the first digital file, wherein the approximate first characteristic is based on manipulation of a digital model of a mold used to create the first shell, determine a second characteristic of the first shell from the one or more images, compare the approximate first characteristic with the second characteristic, and perform quality control on the first shell based on the comparison.

[0035] The thirty - first aspect of the present disclosure can further expand the thirtieth aspect of the present disclosure. In the thirty - first aspect of the present disclosure, one or more images include a top - view image, an approximate first characteristic includes an approximate outer surface of the first housing, and a second characteristic includes a first shape of the first housing. Additionally, the processing device further determines a first plane associated with the top - view image, calculates a first projection of the approximate outer surface of the first housing into the first plane, identifies one or more differences between a second shape of the first projection and the first shape of the first housing based on a comparison, and determines whether the one or more differences exceed a first threshold.

[0036] The thirty - second aspect of the present disclosure can further expand the thirty - first aspect of the present disclosure. In the thirty - second aspect of the present disclosure, the processing device further determines that the one or more differences do not exceed the first threshold, generates a modified projection of the approximate outer surface of the first housing by deforming the second shape of the first projection so that a first curvature of the deformed second shape approximately matches a second curvature of the first shape, identifies one or more additional differences between the first curvature of the deformed second shape and the second curvature of the first shape, and determines whether the one or more additional differences exceed a second threshold.

[0037] The thirty - third aspect of the present disclosure can include a method for inspecting manufacturing defects of a dental appliance. The method includes: obtaining one or more images of the dental appliance, identifying an identifier of the dental appliance, determining a digital file associated with the dental appliance from a set of digital files based on the identified identifier, the digital file associated with the dental appliance including digital models of intermediate components used during the manufacture of the dental appliance, determining expected characteristics of the dental appliance by digitally manipulating the digital models of the intermediate components used during the manufacture of the dental appliance, determining actual characteristics of the dental appliance from the one or more images of the dental appliance, determining whether there are manufacturing defects in the dental appliance by comparing the expected characteristics of the dental appliance with the actual characteristics of the dental appliance, and outputting an output associated with the determination of whether there are manufacturing defects.

[0038] The thirty - fourth aspect of the present disclosure can further expand the thirty - third aspect of the present disclosure. In the thirty - fourth aspect of the present disclosure, the dental appliance includes a customized orthodontic appliance customized for a specific dental arch of a specific patient and a specific stage of orthodontic treatment, and the intermediate components include a positive mold related to the specific dental arch of the patient and the specific stage of orthodontic treatment.

[0039] The thirty - fifth aspect of the present disclosure can further expand the thirty - third or thirty - fourth aspect of the present disclosure. In the thirty - sixth aspect of the present disclosure, the expected characteristics of the customized orthodontic appliance include an expected profile of the customized orthodontic appliance in a plane, and the actual characteristics of the customized orthodontic appliance include an actual profile of the customized orthodontic appliance in the plane captured by one or more images of the customized orthodontic appliance.

[0040] The thirty-sixth aspect of the present disclosure can further expand the thirty-fifth aspect of the present disclosure. The thirty-sixth aspect of the present disclosure can include comparing the expected characteristics of a customized orthodontic appliance with the actual characteristics, including comparing the expected profile with the actual profile and determining whether the difference exceeds a threshold.

[0041] The thirty-seventh aspect of the present disclosure can further expand the thirty-fourth to thirty-sixth aspects of the present disclosure. In the thirty-seventh aspect of the present disclosure, the expected characteristics of the customized orthodontic appliance include the expected cutting line for the customized orthodontic appliance. The actual characteristics of the customized orthodontic appliance include the actual cutting line of the customized orthodontic appliance determined from one or more images.

[0042] The thirty-eighth aspect of the present disclosure can further expand the thirty-fourth to thirty-seventh aspects of the present disclosure. In the thirty-eighth aspect of the present disclosure, an identifier of the customized orthodontic appliance is printed on the customized orthodontic appliance, and the identifier of the customized orthodontic appliance is identified by analyzing one or more images of the customized orthodontic appliance.

[0043] The thirty-ninth aspect of the present disclosure can further expand the thirty-fourth to thirty-eighth aspects of the present disclosure. In the thirty-ninth aspect of the present disclosure, identifying the identifier of the customized orthodontic appliance includes receiving user input of the identifier.

[0044] The fortieth aspect of the present disclosure can further expand the thirty-fourth to thirty-ninth aspects of the present disclosure. In the fortieth aspect of the present disclosure, the method further includes: determining an inspection plan for the customized orthodontic appliance based on one or more images of the obtained customized orthodontic appliance or a digital file associated with the customized orthodontic appliance.

[0045] The forty-first aspect of the present disclosure can further expand the fortieth aspect of the present disclosure. In the forty-first aspect of the present disclosure, the inspection plan specifies one or more additional images to be captured of the customized orthodontic appliance.

[0046] The forty-second aspect of the present disclosure can further expand the fortieth to forty-first aspects of the present disclosure. In the forty-second aspect of the present disclosure, the inspection plan is based on a digital file associated with the customized orthodontic appliance, and the inspection plan is determined by a prediction model that identifies locations of the customized orthodontic appliance that are at a high risk for one or more defects.

[0047] The forty-third aspect of the present disclosure can further expand the thirty-fourth to forty-second aspects of the present disclosure. In the forty-third aspect of the present disclosure, the expected characteristics of the customized orthodontic appliance are determined by digitally manipulating a portion of the surface of a digital model of a mold to approximate the surface of the customized orthodontic appliance.

[0048] The forty-fourth aspect of the present disclosure can further expand the forty-third aspect of the present disclosure. In the forty-fourth aspect of the present disclosure, the surface of a customized orthodontic appliance is approximated by offsetting a portion of the surface of the digital model of the mold by a distance.

[0049] The forty-fifth aspect of the present disclosure can further expand the forty-third to forty-fourth aspects of the present disclosure. In the forty-fifth aspect of the present disclosure, the expected characteristics of a customized orthodontic appliance are determined by virtually projecting a cutting line associated with the customized orthodontic appliance onto the approximated surface of the customized orthodontic appliance.

[0050] The forty-sixth aspect of the present disclosure can further expand the forty-third to forty-fifth aspects of the present disclosure. In the forty-sixth aspect of the present disclosure, the expected characteristics of a customized orthodontic appliance include the expected profile of the customized orthodontic appliance in a plane, and the expected profile of the customized orthodontic appliance is determined by calculating the profile of the approximated surface of the customized orthodontic appliance in the plane.

[0051] The forty-seventh aspect of the present disclosure can further expand the thirty-fourth to forty-sixth aspects of the present disclosure. In the forty-seventh aspect of the present disclosure, the expected characteristics of a customized orthodontic appliance are determined by calculating the profile of the digital model of the mold in a plane.

[0052] The forty-eighth aspect of the present disclosure can further expand the forty-seventh aspect of the present disclosure. In the forty-eighth aspect of the present disclosure, the expected characteristics of a customized orthodontic appliance include the expected profile of the customized orthodontic appliance in a plane, and the expected profile of the customized orthodontic appliance is calculated by offsetting the perimeter of the calculated profile of the digital model of the mold in the plane.

[0053] The forty-ninth aspect of the present disclosure can further expand the thirty-third to forty-eighth aspects of the present disclosure. In the forty-ninth aspect of the present disclosure, the dental appliance includes a mandibular advancement feature.

[0054] The fiftieth aspect of the present disclosure can further expand the thirty-third to forty-ninth aspects of the present disclosure. In the fiftieth aspect of the present disclosure, the output includes determining the presence of a defect, and the output includes recommended digital modifications to the digital model of an intermediate component used during the manufacture of the dental appliance to limit future defects.

[0055] The fifty-first aspect of the present disclosure can further expand the fiftieth aspect of the present disclosure. In the fifty-first aspect of the present disclosure, the recommended digital modifications to the digital model of the intermediate component include at least one of added virtual filling material, correction of the cutting line, and modification of one or more attachments of the intermediate component.

[0056] The fifty-second aspect of the present disclosure may include a method for inspecting manufacturing defects of a dental appliance configured to be applied to a patient's dental arch, the method including: obtaining one or more images of the dental appliance, identifying an identifier of the dental appliance, determining, based on the identified identifier, a digital file associated with the dental appliance from a set of digital files, the digital file associated with the dental appliance including a digital model of the dental appliance, generating the digital model of the dental appliance by digitally manipulating a digital model of the patient's dental arch, determining expected characteristics of the dental appliance from the digital model of the dental appliance, determining actual characteristics of the dental appliance from the one or more images, determining whether there are manufacturing defects in the dental appliance by comparing the expected characteristics of the dental appliance with the actual characteristics of the dental appliance, and outputting an output associated with the determination of whether there are manufacturing defects.

[0057] The fifty-third aspect of the present disclosure may further expand the fifty-second aspect of the present disclosure. In the fifty-third aspect of the present disclosure, the dental appliance includes a customized orthodontic appliance customized for a specific dental arch of a specific patient and a specific stage of orthodontic treatment, and the digital model of the customized orthodontic appliance is generated by manipulating a digital model of the staged dental arch of the patient associated with the specific stage of orthodontic treatment.

[0058] The fifty-fourth aspect of the present disclosure may further expand the fifty-third aspect of the present disclosure. In the fifty-fourth aspect of the present disclosure, the expected characteristics of the customized orthodontic appliance include an expected contour of the customized orthodontic appliance in a plane, and the actual characteristics of the customized orthodontic appliance include an actual contour of the customized orthodontic appliance in the plane captured by one or more images of the customized orthodontic appliance.

[0059] The fifty-fifth aspect of the present disclosure may further expand the fifty-third and / or fifty-fourth aspects of the present disclosure. In the fifty-fifth aspect of the present disclosure, the expected characteristics of the customized orthodontic appliance include an expected cutting line of the customized orthodontic appliance, wherein the actual characteristics of the customized orthodontic appliance include an actual cutting line of the customized orthodontic appliance determined from the one or more images.

[0060] The fifty-sixth aspect of the present disclosure may further expand the fifty-third to fifty-fifth aspects of the present disclosure. In the fifty-sixth aspect of the present disclosure, the method further includes: determining an inspection plan for the customized orthodontic appliance based on the one or more images of the customized orthodontic appliance obtained or the digital file associated with the customized orthodontic appliance.

[0061] The fifty-seventh aspect of the present disclosure can further expand the fifty-third to fifty-sixth aspects of the present disclosure. In the fifty-seventh aspect of the present disclosure, an identifier of a customized orthodontic appliance is printed on the customized orthodontic appliance, and wherein the identifier of the customized orthodontic appliance is identified by analyzing one or more images of the customized orthodontic appliance or by receiving a user input of the identifier.

[0062] The fifty-eighth aspect of the present disclosure can further expand the fifty-second to fifty-seventh aspects of the present disclosure. In the fifty-eighth aspect of the present disclosure, the dental appliance includes a removable palatal expander.

[0063] The fifty-ninth aspect of the present disclosure can further expand the fifty-second to fifty-eighth aspects of the present disclosure. In the fifty-ninth aspect of the present disclosure, the dental appliance includes a removable surgical fixation device.

[0064] The sixtieth aspect of the present disclosure can further expand the fifty-second to fifty-ninth aspects of the present disclosure. In the sixtieth aspect of the present disclosure, the dental appliance includes a removable mandibular protrusion feature.

[0065] The sixty-first aspect of the present disclosure can include a method for inspecting manufacturing defects of a customized orthodontic appliance customized for a specific dental arch and a specific stage of orthodontic treatment of a specific patient, the method including: obtaining one or more images of the customized orthodontic appliance, identifying an identifier of the customized orthodontic appliance, determining expected characteristics of the customized orthodontic appliance based on the identifier of the customized orthodontic appliance, determining expected characteristics of the customized orthodontic appliance by digitally manipulating a digital model of a mold used during the manufacture of the customized orthodontic appliance, determining actual characteristics of the customized orthodontic appliance from the one or more images, determining whether there are manufacturing defects in the customized orthodontic appliance by comparing the expected characteristics of the customized orthodontic appliance with the actual characteristics of the customized orthodontic appliance, and outputting an output associated with the determination of whether there are manufacturing defects.

[0066] The sixty-second aspect of the present disclosure can further expand the sixty-first aspect of the present disclosure. In the sixty-second aspect of the present disclosure, the expected characteristics of the customized orthodontic appliance include an expected profile of the customized orthodontic appliance in a plane, and the actual characteristics of the customized orthodontic appliance include an actual profile of the customized orthodontic appliance in the plane captured by one or more images of the customized orthodontic appliance.

[0067] The sixty-third aspect of the present disclosure can further expand the sixty-first to sixty-second aspects of the present disclosure. In the sixty-third aspect of the present disclosure, the expected characteristics of the customized orthodontic appliance include the expected cutting line of the customized orthodontic appliance, wherein the actual characteristics of the customized orthodontic appliance include the actual cutting line of the customized orthodontic appliance determined from one or more images.

[0068] The sixty-fourth aspect of the present disclosure can further expand the sixty-first to sixty-third aspects of the present disclosure. In the sixty-fourth aspect of the present disclosure, the method further includes: determining an inspection plan for the customized orthodontic appliance based on one or more images of the obtained customized orthodontic appliance or a digital file associated with the customized orthodontic appliance.

[0069] The sixty-fourth aspect of the present disclosure can further expand the sixty-first to sixty-fourth aspects of the present disclosure. In the sixty-fourth aspect of the present disclosure, an identifier of the customized orthodontic appliance is printed on the customized orthodontic appliance, and wherein the identifier of the customized orthodontic appliance is identified by analyzing one or more images of the customized orthodontic appliance or by receiving user input of the identifier. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In the figures of the drawings, the present invention is illustrated by way of example and not limitation.

[0071] Figure 1A-1B A flowchart of a method for performing image-based quality control on a housing using an inspection plan according to one embodiment is shown.

[0072] Figure 2 An example imaging system including a top view camera and a side view camera according to one embodiment is shown.

[0073] Figure 3A-3B An example top view image according to one embodiment and an example movement control and screen path generated based on the top view image are shown.

[0074] Figure 4A-4C An example side view composite image of the side of a housing according to one embodiment, an edge detected using the side view composite image, and a comparison of the edge with a second edge obtained from a digital model of the housing are shown.

[0075] Figure 5 A flowchart of a method for determining an inspection plan based on characteristics of a plastic housing according to one embodiment is shown.

[0076] Figure 6 A flowchart of a method for determining one or more additional images to be generated based on an output from a model according to one embodiment is shown.

[0077] Figure 7 The flowchart of a method for determining an inspection plan using a rule engine according to an embodiment is shown.

[0078] Figure 8A-8B The flowchart of a method for performing image-based quality control on a housing according to an embodiment is shown.

[0079] Figure 9A-9B The flowchart of a method for determining whether the shape of a housing is deformed according to an embodiment is shown.

[0080] Figure 10 The digital model of an appliance projected onto an image of the appliance according to an embodiment is shown.

[0081] Figure 11 The flowchart of a method for determining the shape difference between the digital model of a housing and the image of the housing according to an embodiment is shown.

[0082] Figure 12 The user interface for image-based quality control of a housing according to an embodiment is shown.

[0083] Figure 13A-13B An example comparison of the contour of the digital model of an appliance and the contour of the image of the appliance to detect deformation according to an embodiment is shown.

[0084] Figure 14A-14C The flowchart of a method for deforming the digital model contour to more closely match the image contour of the appliance to detect changes in the cutting line according to an embodiment is shown.

[0085] Figure 15A-15C An example of deforming the digital model contour to more closely match the image contour of the appliance to detect changes in the cutting line according to an embodiment is shown.

[0086] Figure 16 The flowchart of a method for generating a digital model of a housing according to an embodiment is shown.

[0087] Figure 17 The flowchart of a general method for determining the static position of the digital model of a housing on a flat surface according to an embodiment is shown.

[0088] Figure 18 The flowchart of a method for using a two-dimensional digital model to determine the static position of the digital model of a housing on a flat surface according to an embodiment is shown.

[0089] Figure 19A-19C An example image for determining the static position of an appliance on a flat surface according to an embodiment is shown.

[0090] Figure 20 A flowchart of a method for determining a stationary position of a digital model of a housing on a flat surface using a three-dimensional digital model according to one embodiment is shown.

[0091] Figure 21 A block diagram of an example computing device according to one embodiment is shown.

[0092] Figure 22A An example side view image captured without a back screen and without structured light illumination is shown.

[0093] Figure 22B An example side view image captured using structured light illumination (e.g., focused light) without a back screen according to one embodiment is shown.

[0094] Figure 23 An example of crack detection in a contour of an image of an appliance captured using focused light according to one embodiment is shown. DETAILED DESCRIPTION

[0095] Embodiments of systems, methods, and / or computer-readable media for image-based quality control (IBQC) of customized manufactured products are described. The customized manufactured products can be customized medical devices. For example, in some embodiments, the image-based quality control system and method can be implemented in the inspection of orthodontic appliances after manufacture. Quality control of customized manufactured products is particularly difficult, especially in the manufacture of orthodontic appliances, where the orthodontic appliances must be customized individually for each patient. Additionally, each appliance in a series of appliances for treating a single patient is unique compared to other appliances in the same series, as each appliance is specific to a different treatment stage. Compounding the problem is that for each treatment stage, each patient receives a pair of appliances, one unique appliance for treating the upper dental arch and one unique appliance for treating the lower dental arch. In some cases, a single treatment can include 50 - 60 stages for treating complex cases, which means 100 - 120 appliances uniquely manufactured for a single patient. When manufacturing appliances for patients worldwide, hundreds of thousands of completely unique and customized appliances may need to be manufactured each day. Thus, quality control of customized manufactured products can be a particularly daunting task. Quality control of the manufactured appliances can be performed to ensure that the appliances are defect-free or that the defects are within a tolerable threshold. The goal of the quality control process can be to detect one or more of the following quality issues: dental arch changes, bending, cut line changes, debris, webbing, trimmed attachments, missing attachments, etc. Typically, technicians manually perform the quality control process to inspect the appliances. However, this manual quality control process can be very time-consuming and error-prone due to the inherent subjectivity of the technicians. Thus, embodiments of the present invention can provide a more scalable, automated, and / or objective appliance quality control process.

[0096] It should be noted that "appliance" and "housing" may be used interchangeably herein. As described above, embodiments can detect various quality issues for a given set of appliances. Quality issues can include one or more of the following: arch changes, deformation, bending (compression or expansion) of the appliance, cut line changes, debris, sidebands, trimmed attachments, missing attachments, burrs, flaring, power ridge issues, material fractures, short hooks, air bubbles, etc. Detecting quality issues can enable the repair of the appliance to eliminate the quality issues, prevent the delivery of malformed or non-conforming appliances, and / or remanufacture malformed appliances prior to delivery. In some embodiments, the identification of appliance quality issues can be based on an image of the appliance compared to a digitally generated model of each appliance. In some embodiments, the digital model of each appliance can be included in a digital file associated with the appliance. Optionally, the digital file associated with the manufactured appliance can provide characteristics of a digital approximation of the manufactured appliance (e.g., the outer surface of the appliance, two-dimensional projections of the outer surface onto a plane, etc.). The digitally generated model of the appliance and / or the characteristics of the digital approximation of the appliance can be based on the manipulation of a digital model of the mold used to create the appliance. In some embodiments, the identification of appliance quality issues can be based on comparing an approximate first characteristic of the appliance (such as determined based on the manipulation of a digital model of the mold used to manufacture the appliance) and a determined second characteristic of the appliance (such as determined from one or more images of the appliance). Some advantages of the disclosed embodiments can include the automatic detection of various appliance quality issues and the automatic collection of data for statistical analysis. The embodiments can also improve the detection results by eliminating human errors (e.g., false positives and false negatives). Additionally, the embodiments can reduce the amount of time spent performing quality control, thereby reducing the lead-time of the appliance, which can enable the timely distribution of the appliance to customers.

[0097] Various software and / or hardware components can be used to implement the disclosed embodiments. For example, the software components can include computer instructions stored in a tangible, non-transitory computer-readable medium that are executed by one or more processing devices to perform image-based quality control on a custom-manufactured housing (e.g., an appliance). The software can set up and calibrate a camera included in the hardware components, use the camera to capture images of the appliance from various angles, generate a digital model of the appliance, perform an analysis that compares the digital model of the appliance with the images of the appliance to detect one or more quality issues (e.g., deformation, cut line changes, etc.), and classify the appliance based on the analysis results.

[0098] In some embodiments, a digital file associated with a custom-made shell for a patient's dental arch and being inspected can be received. In some examples, the dental appliance includes identification information, such as a custom barcode or a part identification number. One or more imaging devices (e.g., a camera, a blue laser scanner, a confocal microscope, a stereoscopic image sensor, an X-ray device, an ultrasound device, etc.) can be used to generate a first image (e.g., a photographic image, an X-ray image, a digital image) of the plastic shell. The dental appliance identification information can be captured in the first image and interpreted by the inspection system. Optionally, a technician can also manually enter this information at the inspection site so that the inspection system can obtain the digital file. In additional examples, a dental appliance sorting system can sort a series of dental appliances in a known order. The inspection system can obtain the dental appliance order from the dental appliance sorting system to know which dental appliances are currently being inspected and the order in which they arrived at the site. Optionally, the dental appliances can arrive at the inspection system in trays carrying dental appliance identification information (e.g., RFID tags, barcodes, serial numbers, etc.) that can be read by the inspection system. Thereafter, the inspection system can obtain the digital file associated with the dental appliance based on the dental appliance identification information.

[0099] An inspection recipe for the plastic shell can be determined based on at least one of first information associated with the first image of the plastic shell or second information associated with the digital file. The inspection recipe can specify one or more additional images (if any) to be generated of the plastic shell and can specify settings (e.g., zoom, orientation, focus, etc.) of one or more imaging devices. The first information and / or the second information can indicate one or more of the following: the shape of the appliance, the size of the appliance, one or more features of the appliance, areas of higher risk of defects, one or more defects (e.g., distortion, cracks, etc.) of the appliance, etc., which can be used to determine the additional images to be captured for the inspection recipe. Various defects of the plastic shell can be determined based on the first image and / or one or more additional images. For example, when the first image is a top view image of a plastic aligner, distortion of the plastic shell can be detected, as discussed further below. If the additional image is a side view of the plastic shell, air bubbles present on the surface of the plastic shell can be detected and / or inaccurate cut lines can be detected.

[0100] The inspection scheme can be determined dynamically based on the first information and / or the second information, or an inspection scheme can be predetermined for the plastic housing and retrieved from a memory location. For example, determining the inspection scheme can include using at least one of the first information or the second information to determine one or more characteristics of the plastic housing (e.g., precise cut lines, presence of attachment wells, angle of the cut lines, crowding between teeth, crowding between attachment wells, etc.), and determining one or more additional images to be generated based on the one or more characteristics. Additionally, the dimensions and / or shape of the plastic housing can be determined from the first information and / or the second information, and the settings of the imaging device (e.g., orientation, zoom, focus) for generating the additional images can be determined based on one or more characteristics, dimensions, and / or shape of the plastic housing.

[0101] In some embodiments, to determine the inspection scheme, a digital file of the plastic housing can be applied as an input to a model (e.g., a predictive model, a machine learning model, etc.) or a rule engine. The model or rule engine can generate an output that identifies one or more locations of the plastic housing that are identified as high-risk areas for one or more defects. In one example, the model can be a predictive model that performs a numerical simulation on the digital file of the plastic appliance by applying one or more forces to the plastic appliance to simulate the removal process of the plastic appliance from the patient's dental arch. The predictive model can calculate strain values and the force values applied to cause the strain values. If the strain value or the force value exceeds a threshold at a certain location on the plastic appliance, that location can be determined as a high-risk area for a defect. In another example, the model can be a machine learning model that is trained to identify the locations of high-risk areas. The machine learning model can be applied to the first image of the appliance and / or the digital file of the appliance, and can generate an output indicating one or more high-risk areas of the appliance where there are defects at the locations. When one or more characteristics are included at one or more locations, the rule engine can use one or more rules that specify that one or more locations are high-risk areas for one or more defects. One or more additional images for the inspection scheme can be determined based on the one or more locations identified as high-risk areas by the output. Additional details of the method and system for identifying specific appliance areas for inspection can be found in co-pending U.S. Provisional Application No. 62 / 737,458, filed on September 27, 2018, the entire contents of which are incorporated herein by reference.

[0102] Further determining the inspection scheme may include determining settings for one or more imaging devices for capturing one or more additional images based on at least one of first information associated with a first image or second information associated with a digital file. In one example, the first information and / or the second information may include the dimensions and / or shape of the plastic housing, which may be used to determine the zoom settings and / or focus settings of one or more imaging devices. In another example, the first information and / or the second information may include features of the plastic housing (e.g., the angle of a cut line), which may be used to determine the orientation of one or more imaging devices (e.g., the angle at which to generate an image of the feature). Additionally, if it is determined that a side view image is to be generated, the settings may include determining the number of images to be generated to be synthesized together to form a composite two-dimensional (2D) image that is a side panoramic view of the plastic appliance or an image of the side of the plastic housing in three dimensions (3D). The settings of the one or more imaging devices may be different to allow the imaging devices to capture different images, which may enable the detection of different defects. For example, to detect air bubbles in the plastic housing, the zoom setting may be configured to a specific micron setting (e.g., 10 microns to 30 microns), while another zoom setting may enable the detection of another type of defect (e.g., burrs).

[0103] The inspection scheme may be performed using the determined settings to capture one or more additional images of the plastic housing. Based on at least one of the first image and / or the one or more additional images, it may be determined whether one or more defects are included in the plastic housing. Quality control may be performed on the plastic housing in response to determining that one or more defects are included in the plastic housing.

[0104] In some embodiments, detecting a defect can include comparing aspects of a digital file of a mold associated with an orthodontic appliance with aspects of a first image and / or additional images of the orthodontic appliance. For example, an approximate first characteristic of the orthodontic appliance (e.g., an approximate outer surface of the orthodontic appliance) can be determined from a digital file of a mold associated with the orthodontic appliance. The approximate first characteristic can be determined based on manipulation of a digital model of the mold used to create the orthodontic appliance. Also, a second characteristic of the orthodontic appliance (e.g., the shape of the orthodontic appliance in a captured image) can be determined from the captured image. The approximate first characteristic and the second characteristic can be compared. The comparison can include calculating a projection of the approximate outer surface of the orthodontic appliance (e.g., the digital model) into a plane having the same shape as the orthodontic appliance, and identifying an area between the contour of the approximate outer surface and the shape of the orthodontic appliance in the image. If the size of the area (e.g., thickness or area) exceeds a threshold, the orthodontic appliance can be determined to be deformed. If the size is within the threshold, another comparison can be performed that deforms the curvature of the dental arch in the digital model or approximate projection towards the curvature of the dental arch of the manufactured orthodontic appliance in the orthodontic appliance image. Once deformed, other contours of the digital model or other approximate characteristics can be compared with the orthodontic appliance image or other images of the manufactured orthodontic appliance to determine if the cutting line or other characteristics of the imaged orthodontic appliance match within the threshold. If there is no match, the cutting line or other characteristics can be determined to be defective. Other orthodontic appliance characteristics that can be analyzed can include debris, sidebands, trimmed attachments, and missing attachments, etc. The software can also determine how the manufactured orthodontic appliance is placed on a two-dimensional plane and accordingly adjust the projection of the digital model or approximate characteristics and the image of the manufactured orthodontic appliance for comparative analysis.

[0105] In some embodiments, a digital model of an orthodontic appliance can be generated as part of the manufacturing process of the orthodontic appliance, and the digital model can be received as an input. Additionally, a user interface can be provided that displays the image-based quality control process (e.g., digital model of the orthodontic appliance, image of the orthodontic appliance, comparison of the digital model with the image) and the results (e.g., measurements, classifications).

[0106] The hardware components may include a platform with a fixed or rotating table for orthodontic appliance positioning and image capture, a camera setup (e.g., positioning), and / or a lighting system for uniform exposure and image capture using uniform environmental parameters. Additionally, the hardware components may also enable automatic feeding of components (e.g., orthodontic appliances) into the station where image-based quality control is being performed and be capable of sorting at the exit of the station. Images of the orthodontic appliances can be obtained from one or several projections. Thus, the hardware components may include a fixed or rotating table and one or more cameras with adjustable positioning. Adjusting the configuration of the hardware components can enable obtaining orthodontic appliance images from different angles (e.g., top view, side view, diagonal view, etc.). In one embodiment, the first camera may be positioned at an angle such that a top view image of the orthodontic appliance being analyzed can be captured, and the second camera may be configured to capture one or more side view images or diagonal view images based on the settings. The hardware components may also include a blue laser scanner that includes a camera, a background screen, and a lighting device to obtain an image of the housing. Specific information (e.g., a second characteristic) can be extracted from the image. The blue laser can be applied at an angle to the surface of the orthodontic appliance, and a blue light beam (e.g., having a wavelength of approximately 440 - 490 nm) can be generated, which is received by the camera to generate an image of the plastic housing. Depth information can be extracted from the image to obtain the desired information (e.g., a second characteristic). The hardware may also include a robot-guided camera that uses a design file to guide the camera to generate an image of the orthodontic appliance being analyzed. The image can be an image of the edge / cut line of the orthodontic appliance. The hardware may also include an ultrasonic device that emits sound waves to measure the thickness of the orthodontic appliance. Measuring the thickness of the orthodontic appliance can enable forming a quality trend analysis and detecting thickness-related defects. The hardware may also include a stereoscopic image sensor to obtain a three-dimensional (3D) image of the plastic housing. The hardware may also include a confocal microscope for obtaining orthodontic appliance images at various depths of focus. The hardware may also include an X-ray device for scanning a transparent plastic orthodontic appliance at various cross-sections and obtaining an image of the transparent plastic orthodontic appliance. Specific current and voltage settings can be used to scan the cross-section of the transparent plastic orthodontic appliance and obtain an image of the transparent plastic orthodontic appliance.

[0107] Some embodiments are discussed with reference to an orthodontic appliance (also simply referred to as an appliance). However, the embodiments also extend to other types of shells formed on a mold, such as orthodontic retainers, orthodontic splints, oral insertable sleep appliances (e.g., for minimizing snoring, sleep apnea, etc.) and / or shells for non-dental applications. Other applications can be found when examining 3D printed palatal expanders, removable mandibular repositioning devices, and removable surgical fixation devices. Thus, it should be understood that the embodiments involving appliances herein are also applicable to other types of dental appliances. For example, the principles, features, and methods discussed can be applied to any application or process in which image-based quality control is useful for any suitable type of customized device (such as spectacle frames, contact lenses or glass lenses, hearing aids or plugs, artificial knees, prosthetics and devices, orthopedic inserts), as well as protective devices (such as knee pads, sports cups, or elbow, chin, and shin guards and other similar sports / protective devices).

[0108] In some embodiments, a mold of a patient's dental arch can be fabricated, and a shell can be formed on the mold. The fabrication of the mold can be performed by the processing logic of a computing device such as Figure 21 a computing device in. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations can be performed by a processing device executing a computer-aided drafting (CAD) program or module.

[0109] To fabricate the mold, the shape of the patient's dental arch at a treatment stage is determined based on a treatment plan. In an orthodontics example, a treatment plan can be generated based on an intraoral scan of the dental arch to be modeled. An intraoral scan of the patient's dental arch can be performed to generate a three-dimensional (3D) virtual model (mold) of the patient's dental arch. For example, a full scan of the patient's mandibular and / or maxillary arch can be performed to generate its 3D virtual model. The intraoral scan can be performed by creating multiple overlapping intraoral images from different scan stations and then stitching the intraoral images together to provide a composite 3D virtual model. In other applications, a virtual 3D model can also be generated based on a scan of the object to be modeled or based on the use of computer-aided drafting techniques (e.g., for designing a virtual 3D mold). Alternatively, an initial negative mold (e.g., a dental impression, etc.) can be generated from the actual object to be modeled. The negative mold can then be scanned to determine the shape of the positive mold to be produced.

[0110] Once a virtual 3D model of a patient's dental arch is generated, a dentist can determine a desired treatment outcome that includes the final position and orientation of the patient's teeth. Then, processing logic can determine a plurality of treatment phases to progress the teeth from the starting position and orientation to the target final position and orientation. The shape of the final virtual 3D model and each intermediate virtual 3D model can be determined by calculating the progression of tooth movement throughout the orthodontic treatment from the initial tooth placement and orientation to the final corrected tooth placement and orientation. For each treatment phase, a separate virtual 3D model of the patient's dental arch can be generated. The shape of each virtual 3D model will be different. The original virtual 3D model, the final virtual 3D model, and each intermediate virtual 3D model are unique and customized for the patient.

[0111] Thus, multiple different virtual 3D models can be generated for a single patient. The first virtual 3D model can be a unique model of the patient's dental arch and / or teeth as currently presented, and the final virtual 3D model can be a model of the patient's dental arch and / or teeth after correcting one or more teeth and / or jaws. Multiple intermediate virtual 3D models can be modeled, each of which can be incrementally different from the previous virtual 3D model.

[0112] Each virtual 3D model of the patient's dental arch can be used to generate a unique customized physical mold of the dental arch at a particular stage of treatment. The shape of the mold can be at least partially based on the shape of the virtual 3D model at that treatment stage. The virtual 3D model can be represented in a file such as a computer-aided drafting (CAD) file or a 3D printable file such as a stereolithography (STL) file. The virtual 3D model of the mold can be sent to a third party (e.g., a clinician's office, a laboratory, a manufacturing facility, or other entity). The virtual 3D model can include instructions that will control a manufacturing system or device to produce a mold with a specified geometry.

[0113] A clinician's office, a laboratory, a manufacturing facility, or other entity can receive the virtual 3D model of the mold, i.e., the digital model that has been created as described above. The entity can input the digital model into a rapid prototyper. Then, the rapid prototyper uses the digital model to fabricate the mold. An example of a rapid prototyping machine is a 3D printer. 3D printing includes any layer-based additive manufacturing process. 3D printing can be achieved using an additive process, in which successive layers of material are formed in a prescribed shape. 3D printing can be performed using extrusion deposition, granular material binding, lamination, photopolymerization, continuous liquid interface production (CLIP), or other techniques. 3D printing can also be achieved using a subtractive process (e.g., milling).

[0114] In some cases, stereolithography (SLA), also known as optical fabrication solid imaging, is used to fabricate SLA molds. In SLA, a mold is fabricated by sequentially printing thin layers of a photocurable material (e.g., a polymeric resin) one on top of another. A platform is positioned in a bath of a liquid photopolymer or resin, just below the surface of the bath. A light source (e.g., an ultraviolet laser) traces a pattern on the platform, curing the photopolymer that the light source is directed at to form the first layer of the mold. The platform is lowered incrementally, and the light source traces a new pattern on the platform to form another layer of the mold at each increment. This process is repeated until the mold is fully fabricated. Once all the layers of the mold are formed, the mold can be cleaned and cured.

[0115] Materials such as polyesters, copolyesters, polycarbonates, polycarbonates, thermoplastic polyurethanes, polypropylenes, polyethylenes, polypropylene and polyethylene copolymers, acrylics, cyclic block copolymers, polyetheretherketones, polyamides, polyethylene terephthalate, polybutylene terephthalate, polyetherimides, polysulfones, polytrimethylene terephthalate, styrene block copolymers (SBCs), silicone rubbers, elastomer alloys, thermoplastic elastomers (TPEs), thermoplastic vulcanizates (TPVs) elastomers, polyurethane elastomers, block copolymer elastomers, polyolefin blend elastomers, thermoplastic copolyester elastomers, thermoplastic polyamide elastomers, or combinations thereof can be used to directly form the mold. The material used to fabricate the mold can be provided in an uncured form (e.g., as a liquid, resin, powder, etc.) and can be cured (e.g., by photopolymerization, photocuring, gas curing, laser curing, crosslinking, etc.). The properties of the material before curing may be different from the properties of the material after curing.

[0116] An appliance can be formed from each mold and, when applied to a patient's teeth, can provide a force to move the patient's teeth as prescribed by a treatment plan. The shape of each appliance is unique and customized for a specific patient and a specific treatment stage. In one example, the appliance can be pressure formed or thermoformed on the mold. Each mold can be used to fabricate an appliance that will apply a force to a patient's teeth at a specific stage of orthodontic treatment. Each appliance has a tooth receiving cavity that houses the teeth and elastically repositions the teeth according to the specific treatment stage.

[0117] In one embodiment, a sheet of material is pressed or thermoformed on the mold. The sheet can be, for example, a plastic sheet (e.g., an elastomeric thermoplastic, a polymeric material sheet, etc.). To thermoform a shell on the mold, the sheet of material can be heated to a temperature at which the sheet becomes pliable. Pressure can be applied to the sheet simultaneously to form the now pliable sheet around the mold. Once the sheet cools, it will have the shape that conforms to the mold. In one embodiment, a release agent (e.g., a non-stick material) is applied to the mold before forming the shell. This can facilitate subsequent removal of the mold from the shell.

[0118] Additional information can be added to the appliance. The additional information can be any information related to the appliance. Examples of such additional information include component number identifiers, patient names, patient identifiers, case numbers, sequence identifiers (e.g., indicating which appliance in the treatment sequence a particular liner is), manufacturing dates, clinician names, logos, etc. For example, after the appliance is thermoformed, the appliance can be laser marked with a component number identifier (e.g., serial number, barcode, etc.). In some embodiments, the system can be configured to read (e.g., optically, magnetically, etc.) an identifier of the mold (barcode, serial number, electronic tag, etc.) to determine the component number identifier associated with the appliance formed thereon. After determining the component number identifier, the system can then use the unique component number identifier to mark the appliance. The component number identifier can be computer-readable and can associate the appliance with a specific patient, a specific stage in the treatment sequence, whether it is an upper or lower shell, a digital model of the mold from which the appliance is manufactured, and / or a digital file including the virtual generated digital model of the appliance or its approximated characteristics (e.g., generated by approximating the outer surface of the appliance by manipulating the digital model of the mold, inflating or scaling the projections of the mold in different planes). In some embodiments, the virtual generated digital model of the appliance or its approximated characteristics can be compared with the characteristics of the manufactured appliance (e.g., the shape of the appliance) determined from an image of the manufactured appliance for image-based quality control, as described in more detail below with reference to FIGS. 1 and Figure 8A and Figure 8B described in more detail.

[0119] After the appliance is formed on the mold for a treatment stage, the appliance is subsequently trimmed along a cutting line (also referred to as a trimming line) and the appliance can be removed from the mold. The processing logic determines the cutting line of the appliance. The determination of the cutting line can be based on a virtual 3D model of the dental arch for a specific treatment stage, a virtual 3D model of the appliance to be formed on the dental arch, or a combination of the virtual 3D model of the dental arch and the virtual 3D model of the appliance. The position and shape of the cutting line are important for the function of the appliance (e.g., the ability of the appliance to apply the desired force to the patient's teeth) as well as the fit and comfort of the appliance. For shells such as orthodontic appliances, orthodontic retainers, and orthodontic splints, the trimming of the shell plays an important role in the efficacy of the intended purpose of the shell (e.g., aligning, retaining, or positioning one or more teeth of the patient) and the fit of the shell on the patient's dental arch. For example, if the shell is trimmed too much, the shell may lose rigidity and the ability of the shell to apply force to the patient's teeth may be impaired.

[0120] On the other hand, if too little trimming of the housing is done, some parts of the housing may impact the patient's gums and cause discomfort, swelling, and / or other dental problems. Additionally, if too little trimming of the housing is done at one location, the housing may be too rigid at that location. In some embodiments, the cutting line can be a straight line passing through the appliance at the gum line, below the gum line, or above the gum line. In some embodiments, the cutting line can be a gingival cutting line, which represents the interface between the appliance and the patient's gums. In such embodiments, the cutting line controls the distance between the edge of the appliance and the patient's gingival line or gum surface.

[0121] Each patient has a unique dental arch with unique gums. Thus, the shape and location of the cutting line can be unique and customized for each patient and each treatment stage. For example, the cutting line is customized to follow the gum line (also known as the gingival line). In some embodiments, the cutting line can deviate from the gum line in some areas and be on the gum line in other areas. For example, in some cases, it may be desirable for the cutting line to deviate from the gum line (e.g., not touch the gums), where in the interproximal areas between teeth, the housing will contact the teeth and be on the gum line (e.g., touch the gums). Therefore, it is important to trim the housing along a predetermined cutting line.

[0122] In some embodiments, the housing can have multiple cutting lines. The first or main cutting line can control the distance between the edge of the housing and the patient's gum line. Additional cutting lines can be used to cut slots, holes, or other shapes in the housing. For example, additional cutting lines can be used to remove the occlusal surface of the housing, an additional surface of the housing, or a portion of the housing, which when removed, results in the formation of a hook that can be used with an elastic.

[0123] In some embodiments, the gingival cutting line is determined by first defining an initial gingival curve along a line around a tooth (LAT) of the patient's dental arch from a virtual 3D model (also known as a digital model) of the patient's dental arch at the treatment stage. The gingival curve can include the interproximal areas between adjacent teeth of the patient and the interface area between the teeth and the gums. The initially defined gingival curve can be replaced with a modified dynamic curve representing the cutting line.

[0124] Defining an initial gingival curve along a tooth perimeter line (LAT) can be appropriately performed by various conventional procedures. For example, such generation of the gingival curve can include any traditional computational orthodontic method or process for identifying the gingival curve. In one example, an initial gingival curve can be generated by using a Hermite-Spline process. Generally, the Hermite form of a cubic polynomial curve segment is determined by constraints at endpoints P1 and P4 and tangent vectors at endpoints R1 and R4. The Hermite curve can be written in the following form:

[0125] Q(s) = (2s 3 - 3s 2 + 1)P 1 + (-2s 3 + 3s 2 )P 4 + (s 3 - 2s 2 + s)R 1 + (s 3 - s 2 )R 4 ; s ∈ [0, 1] (1)

[0126] Equation (1) can be rewritten as:

[0127] Q(s) = F 1 (s)P 1 + F 2 (s)P 4 + F 3 (s)R 1 + F 4 (s)R 4 ; (2)

[0128] where Equation (2) is the geometric form of the Hermite spline curve, vectors P 1 , P 4 , R 1 , R 4 are geometric coefficients, and the F terms are Hermite basis functions.

[0129] The gingival surface is defined by the gingival curves and a baseline on all teeth, where the baseline is obtained from a digital model of the patient's dental arch. Thus, using multiple gingival curves and the baseline, a Hermite surface patch representing the gingival surface can be generated.

[0130] Rather than having cut lines that create sharp points or other constricted areas (which may cause weakening of the appliance material during use) in the interdental regions between teeth, the initial gingival curve can be replaced with a cut line that has been modified from the initial gingival curve. By initially obtaining a plurality of sample points from a pair of gingival curve portions located on each side of the interdental region, a cut line can be generated to replace the initial gingival curve. The sample points are then converted into a list of points with associated geometric information (e.g., converted into the Amsterdam Dental Function (ADF) format or other similar data formats). The sample points can be appropriately selected near the internal region between two teeth, but at a sufficient distance from the location where the two teeth meet or reach a point within the interdental region between the two teeth (or where the space between the two teeth narrows).

[0131] The set of sample points provides a plurality of points in space (not in the same plane), which can be used to generate an average plane and a vector perpendicular to the average plane. The sample points associated with the gingival curve portions can then be projected onto the average plane to generate two new curves. To minimize weakening of the area of the appliance material within the interdental region, the modified dynamic curve can be configured with an offset adjustment that includes a minimum radius setting in the interdental region to prevent breakage of the appliance material during use. The offset adjustment is also configured to ensure that the resulting cut line has a sufficient radius in the interdental region to contribute sufficient resistance to the teeth to cause effective movement, but not a radius that is too small and prone to breakage. For example, a cusp or other constricted portion of the material may generate stress regions that are prone to cracking during use and should therefore be avoided. Thus, rather than having the cut line include a cusp or other constricted area, a plurality of intersection points and tangent points can be used to generate a cut line in the interdental region between adjacent teeth that maintains the structural strength of the appliance and prevents cusp points and / or constricted portions that may break. In one embodiment, the cut line is spaced from the gingival surface at the area where the appliance will contact the teeth and is designed to at least partially contact the patient's gingival surface in one or more interdental regions between the teeth.

[0132] After determining the cut line, the appliance can then be cut along the cut line (or cut lines) using markings and / or elements printed in the appliance. In some embodiments, the appliance can be manually cut by a technician using scissors, a drill, a cutting wheel, a scalpel, or any other cutting tool. In another embodiment, the appliance is cut along the cut line by a computer-controlled milling machine (e.g., a CNC machine or a laser milling machine). The computer-controlled milling machine can include a camera capable of identifying the cut line in the appliance. The computer-controlled milling machine can use an image from the camera to determine the position of the cut line from the markings in the appliance and can control the angle and position of the cutting tool of the milling machine to trim the appliance along the identified cut line using the identified markings.

[0133] Additionally or alternatively, the appliance may include coordinate system reference markers that can be used to reference the coordinate system of the milling machine to a predetermined coordinate system of the appliance. The milling machine may receive a digital file having milling instructions (e.g., that indicate the position and angle of the milling machine's laser or cutting tool to cause the milling machine to mill the appliance along a cutting line). By aligning the coordinate system of the milling machine with the appliance, the accuracy of computer-controlled milling of the appliance at the cutting line can be improved. The coordinate system reference markers may include markers sufficient to identify an origin and x, y, and z axes.

[0134] Prior to milling the appliance, a technician may apply a dye, colored filler, or other material to the appliance to fill in slight indentations left by one or more elements imprinted in the appliance. The dye, colored filler, etc. may color the slight indentations without coloring the remainder of the appliance. This can increase the contrast between the cutting line and the remainder of the appliance. After removing the appliance from the mold and milling, additional polishing (e.g., polishing of edges) and / or removal of undesired artifacts may be performed. After milling, the appliance may be removed from the mold.

[0135] In embodiments disclosed herein, each manufactured appliance or other dental device (e.g., removable surgical fixation device, removable mandibular repositioning appliance, removable palatal expander) may be sent to an image-based quality control (IBQC) site that detects one or more quality issues (e.g., deformation) that the appliance has. Alternatively, appliances marked for quality inspection may be sent to the IBQC site. For example, a digital file of the appliance may be input into a machine learning model, numerical simulation, rule engine, and / or other module to determine if any of these appliances will have an increased chance of being defective. The machine learning model, numerical simulation, rule engine, and / or other module may identify a subset of appliances to be inspected using the IBQC system. Optionally, the IBQC system and method may classify the inspected appliances as deformed, possibly deformed, or not deformed, and may also provide recommendations (e.g., further inspection needed, remanufacturing needed, approval, etc.) including the results of its analysis.

[0136] Turning now to the drawings, Figure 1A A flowchart of a method 100 for performing image-based quality control on a shell (e.g., orthodontic appliance) according to one embodiment is shown. One or more operations of method 100 are performed by processing logic of a computing device. The processing logic may include hardware (e.g., circuitry, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 100 may be performed by executing Figure 21The processing device of the image-based quality control module 2150 performs. It should be noted that method 100 can be performed for each unique appliance or subset of unique appliances manufactured for a patient's treatment plan.

[0137] At block 102, the processing logic can receive a digital file associated with a plastic shell customized for a patient's dental arch. Dental appliance identification information can be captured by a camera and interpreted by the inspection system. Optionally, a technician can also manually input this information so that the inspection system can obtain the digital file. In additional embodiments, a dental appliance collation system can collate a series of dental appliances in a known order. The inspection system can obtain the dental appliance order from the dental appliance collation system to know which dental appliances are currently being inspected and the order in which they arrived at the station. Optionally, the dental appliance can arrive at the inspection system in a tray carrying dental appliance identification information (e.g., RFID tag, barcode, serial number, etc.), which is read by the inspection system. Thereafter, the inspection system can obtain the digital file associated with the dental appliance based on the dental appliance identification information received by the system.

[0138] In some embodiments, the digital file can include a digital model of the plastic shell. In some embodiments, the digital file associated with the plastic shell can include a digital model of the mold used to manufacture the plastic shell. The digital model of the plastic shell can be obtained by manipulating the digital model of the mold and approximating a first characteristic of the plastic shell. For example, in some embodiments, the surface of the mold can be enlarged, inflated, or otherwise offset to approximate the surface (inner surface and / or outer surface) of the plastic shell. In some cases, the (one or more) surfaces of the mold associated with the patient's teeth and / or attachments and / or virtual fillings are enlarged, inflated, or offset. Optionally, the inner surface or outer surface of the plastic shell is determined, and the other surfaces are approximated based on the thickness of the material used to form the plastic shell. In some cases where the shell is to be formed by thermoforming a sheet of material over a physical mold, the approximated surfaces can take into account the stretching and thinning of the material on certain parts of the mold.

[0139] At block 104, the processing logic can use one or more imaging devices to generate a first image of the plastic shell. The first image can be a top view image, a side view image, or a diagonal image, and can include at least one of a photographic image, an X-ray image, or other digital images (e.g., ultrasound image). The one or more imaging devices can include at least one of a camera, a blue laser scanner, a confocal microscope, a stereoscopic image sensor, an X-ray device, and / or an ultrasound device.

[0140] In some embodiments, two cameras can be used in conjunction with specific lighting, a backing screen, mirrors, and / or an X-Y stage to capture a first image and / or additional images, as further described below with reference to Figure 2 For example, the first image can be captured by a first top-view camera and can be used to determine an inspection scenario, as described below. A top-view image (such as Figure 3A shown) can be used to determine movement control and / or a screen path of the backing screen between the front and back sides of the plastic housing identified in the first image (such as Figure 3B shown). The movement control and screen path can be used to position the backing screen when capturing additional images (e.g., side-view images) specified in the inspection scenario. In some embodiments, a digital file can be used to guide one or more imaging devices (e.g., robot-guided image acquisition) to capture a first image by tracing the edges and / or cut lines of the plastic housing based on a digital model of the plastic appliance included in the digital file. Additionally, in some embodiments, processing logic can analyze the digital file associated with the plastic housing to determine one or more features included in the plastic housing and configure the settings of one or more imaging devices to capture a first image at locations associated with the one or more features.

[0141] At block 106, the processing logic can determine an inspection scenario for the plastic housing based on at least one of first information associated with a first image of the plastic housing or second information associated with the digital file. The inspection scenario can specify one or more additional images to be generated of the plastic housing. For example, the first information associated with the first image of the plastic housing or the second information associated with the digital file can indicate the size and / or shape of the plastic housing, and the processing logic can determine to capture one or more additional images using specific zoom settings, focus settings, and / or orientation (e.g., angle, position, etc.) settings. If the first information indicates that the plastic housing is small in size, the processing logic can determine to capture additional images with magnified image device settings. Additionally, in some embodiments, the first information associated with the first image of the plastic housing or the second information associated with the digital file can indicate the presence of certain features (e.g., precise cut lines of the plastic appliance, cavities of the plastic appliance associated with attachments, angles of the cut lines, distances between cavities of the plastic appliance associated with teeth, or distances between cavities of the plastic appliance associated with attachments). The processing logic can determine an inspection scenario based on the identified features to include one or more additional images of the plastic housing, as further discussed with reference to Figure 5 below.

[0142] The processing logic can also determine settings of the imaging device for capturing one or more additional images. The settings can be based on the dimensions, shapes, and / or features identified in the plastic housing. For example, the settings can include at least one of one or more positions of one or more imaging devices for generating one or more additional images, one or more orientations of one or more imaging devices for capturing one or more additional images, one or more depths of focus of one or more imaging devices for capturing one or more additional images, or the number of one or more additional images of one or more imaging devices.

[0143] In addition, in some embodiments, the processing logic can determine an inspection plan by applying a digital file associated with the plastic housing as an input to a model (e.g., a trained machine learning model or a numerical simulation), and the model can output one or more positions of the plastic housing that are identified as high-risk areas for defects, as further described below with reference to Figure 6 Further described. Additionally, in some embodiments, the processing logic can determine an inspection plan by applying a digital file associated with the plastic housing to a rule engine that uses one or more rules that specify capturing one or more additional images at one or more positions when one or more features are present at one or more positions of the plastic housing, as further described below with reference to Figure 7 Further described.

[0144] In some embodiments, determining the inspection plan can include obtaining the inspection plan from a Figure 21 memory location of the computer system shown. The settings of one or more imaging devices for generating one or more additional images can be preset in the inspection plan obtained from the memory location. The settings include at least one of one or more positions of one or more imaging devices for generating one or more additional images, one or more orientations of one or more imaging devices for capturing one or more additional images, one or more depths of focus of one or more imaging devices for capturing one or more additional images, or the number of one or more additional images of one or more imaging devices.

[0145] At block 108, the processing logic may execute an inspection scenario to capture one or more additional images of the plastic housing. In some embodiments, executing an inspection scenario to capture one or more additional images of the plastic housing may include configuring settings of one or more imaging devices to capture one or more images based on at least one of first information associated with the first image or second information associated with the digital file. The settings may include at least one of the orientation of one or more imaging devices, the position of one or more imaging devices, the zoom of one or more imaging devices, or the depth of focus of one or more imaging devices. In some embodiments, executing the inspection scenario may include: using the imaging device, using data from the design file of the plastic housing or following the edge of the plastic housing according to the movement control and screen path determined from the first image, to capture a subset of images of one or more additional images representing the cut lines of the plastic housing.

[0146] At block 110, the processing logic may determine whether one or more defects are included in the plastic housing based on the first image and / or one or more additional images. If no defects are included, the method may end. In some embodiments, determining whether there are defects may include: obtaining a digital file associated with the plastic housing; determining an approximate first characteristic (or expected characteristic) of the plastic housing from the digital file; determining a second characteristic (or actual characteristic) of the manufactured plastic housing from the first image and / or one or more additional images; and comparing the approximate first / expected characteristic with the second / actual characteristic. If the approximate first characteristic and the second characteristic differ by a threshold amount, it may be determined that there are defects in the plastic housing. In some applications, the digital file associated with the plastic housing includes a digital model of the mold and / or a digital model of the housing. Optionally, determining the expected characteristics of the plastic housing includes manipulating the digital model of the mold, examples of which are provided throughout the text. For example, the expected characteristics may be determined by manipulating the surface of the digital model of the mold to approximate the outer surface of the manufactured dental appliance. In some cases, the expected characteristics of the dental appliance may be the projection or profile of the expected outer surface of the dental appliance in a plane. In some embodiments, the approximate first characteristic may be the virtual cut line of the digital model of the aligner, and the second characteristic may be the actual cut line from the first image of the plastic housing. It should be noted that "cut line" and "edge" may be used interchangeably in this document. In Figure 4A-4C Examples of determining the second characteristic from the additional images of the captured inspection scenario and comparing the second characteristic with the approximate first characteristic to determine whether any defects exist are shown.

[0147] At block 110, if it is determined that one or more defects are included in the plastic housing, then at block 112, the processing logic may perform quality control on the plastic housing. For example, the processing logic may classify the plastic housing as defective and specify one or more remedial measures (e.g., adding filler material, smoothing cutting lines, modifying one or more attachments on the mold or attachment cavities of the dental appliance, remanufacturing, etc.) to attempt to eliminate one or more defects. Examples of quality control operations are described in more detail below.

[0148] Figure 1B A flowchart of a method 120 for performing image-based quality control on a housing (e.g., an orthodontic appliance) according to one embodiment is shown. One or more operations of method 120 are performed by the processing logic of a computing device. The processing logic may include hardware (e.g., circuitry, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 120 may be performed by a processing device executing Figure 21 the image-based quality control module 2150. It should be noted that method 120 may be performed for each unique appliance or subset of unique appliances manufactured for each patient's treatment plan.

[0149] At block 122, the processing logic may use one or more imaging devices to generate a first image of the plastic housing. In some embodiments, the first image may be a top view of the plastic housing. The one or more imaging devices may include cameras, X-ray devices, blue laser scanners, etc.

[0150] At block 124, the processing logic may determine whether one or more defects are detected in the plastic housing based on the first image. If no defects are detected, method 120 may end. One or more defects may be detected by comparing the first image of the plastic housing with a digital file including a digital model of the plastic housing. For example, the processing logic may compare the top view image of the first housing with the top view of the digital model of the plastic housing to determine whether the shape of the plastic housing is deformed. If one or more differences between the shape of the plastic housing from the top view image and the shape of the plastic housing from the digital model of the housing exceed a threshold, it may be determined that the plastic housing includes a defect. The first image may be applied to a trained machine learning model or a rule engine that is trained to identify high-risk areas of defects, and the rule engine includes rules that certain features at specified locations indicate high-risk areas of defects.

[0151] If one or more defects are detected in the plastic appliance or a possible defect (a high-risk area for a defect) is detected, then at block 126, the processing logic can determine an inspection scenario for the plastic housing based at least on the one or more defects. The processing logic can determine that one or more additional images are to be generated for the inspection scenario and settings for one or more imaging devices for capturing the one or more additional images. For example, if it is determined based on a top-view image that the plastic housing is deformed, the processing logic can determine to capture one or more side-view images of the plastic housing at the deformed location for further analysis (e.g., to verify and / or detect other defects) of the plastic housing. The determined settings can include magnifying the detected defect at certain locations, adjusting the depth of focus to verify the defect and / or detect other defects, etc. Once the inspection scenario is determined, the processing logic can execute the inspection scenario by capturing the one or more additional images using the determined settings.

[0152] At block 128, the processing logic can analyze the one or more additional images to verify the detected defects and / or possible defects in the first image, and / or detect one or more new defects and / or possible defects in the plastic housing. In some embodiments, the processing logic can compare the additional images with a similar representation of the plastic appliance in the digital model (e.g., similar zoom settings, focus settings, etc.) to determine any differences and / or verify the defects detected based on the first image. In some embodiments, the processing logic can detect new defects (e.g., cracks) and / or possible defects based on the additional images. As described above, the additional images can be applied to a machine learning model and / or a rule engine.

[0153] At block 130, the processing logic can determine based on the one or more additional images whether the defects and / or possible defects detected in the first image are verified, and / or whether any new defects and / or possible defects are detected. If one or more of the defects and / or possible defects detected in the first image are not verified, and / or no new defects and / or possible defects are detected in the one or more additional images, then method 120 can end. If the defects and / or possible defects detected in the first image are verified based on the one or more additional images and / or new defects and / or possible defects are detected based on the one or more additional images, then at block 132, the processing logic can perform quality control on the plastic housing.

[0154] Figure 2FIG. 200 shows an example imaging system 200 including a top view camera 202 and a side view camera 204 according to one embodiment. The imaging system 200 can be used to extract the cutting line of the plastic housing 206 to be analyzed to determine whether there are defects by comparing the cutting line of the plastic housing with the virtual cutting line obtained from the digital model of the plastic housing 206. The plastic housing 206 can be fixed in a fixed position by the platform assembly holder 208. According to an embodiment, the top view camera 202 can be configured to obtain a top view image 300 of the transparent plastic housing 206 using a specific lighting setting so that the transparent plastic appliance 206 can be visible in the top view image 300, as Figure 3A shown. According to an embodiment, the processing logic can obtain the outline 302 of the projection or profile of the plastic housing, as Figure 3B shown. The side view camera 204 can be used to obtain the front view and the rear view of the plastic appliance by rotating around the plastic appliance 206 while the platform assembly holder 208 holds the plastic appliance 206, or by rotating the plastic appliance 206 while the side view camera 204 remains stationary. In some embodiments, the plastic appliance 206 may not be fixed by the platform assembly holder 208, and the plastic appliance 206 may be stationary on the platform while the side view camera 204 takes multiple images around the side of the plastic appliance 206. In some embodiments, the cameras 202 and 204 can be stationary and placed away from the conveyor path. In some embodiments, the imaged plastic housing 206 can be placed on an x-y-z-θ (4 axes of movement control) platform or stage.

[0155] In an embodiment, the imaging system 200 can obtain separate front and rear view images by using a backing screen 210 without stray light interference from the currently unexamined side. The backing screen 210 can be inserted into the gap between the front (buccal) and rear (lingual) sides of the plastic appliance. The movement control and the screen path 304 of the plastic housing 206 are determined by identifying points between the front and rear sides of the plastic appliance that enable the generation of a screen path such that the backing screen 210 does not contact the plastic appliance throughout the screen path. The processing logic can detect the center 306 of the plastic appliance and adjust the movement control and screen path parameters accordingly. Additionally, the movement control speed can be high enough to achieve an inspection cycle for both the front and rear sides of the plastic appliance within a target time period (e.g., 10 - 20 seconds). In an embodiment, the mirror 212 can be used as a deflector to capture an image from the rear or front side of the plastic appliance when the plastic appliance is held in the platform assembly holder 208. The mirror 212 can be angled at a specific angle (e.g., 45°, 50°, 55°, etc.) and can be used in combination with a light source to enable the capture of an image depicting the profile of the cutting line of the front (buccal) side and the cutting line of the back (lingual) side of the plastic appliance.

[0156] In some embodiments, the imaging system 200 can prevent light interference from the front row of teeth from interfering with the imaging of the back row of teeth without using a backplate. In some embodiments, the imaging system 200 can use focused light to obtain the cutting line of the plastic housing 206. For example, the focused light can be used to irradiate only the currently examined cutting line (e.g., buccal or lingual) without stray light interference from other cutting lines that are simultaneously present in the field of view of the camera (e.g., preventing interference between the buccal and lingual cutting lines). In such an embodiment, the top view camera 202 can capture a top view image of the plastic housing 206 and extract the top view profile. In some embodiments, the plastic housing 206 can be placed within the field of view of the top view camera 202, and the imaging system 200 can align the plastic housing 206 to capture the top view image.

[0157] The top view image can be used to determine an inspection scenario that includes one or more side view images. Using the top view contour, the contour x-y points can be sent to the side view camera 204. In some embodiments, the characteristics of the camera, such as zoom and / or depth of focus, can be determined for the inspection scenario based on which side of the plastic housing 206 is being captured. For example, when the buccal side is farther away, the focus area of the side view camera 204 can be adjusted to accurately capture the lingual side of the plastic housing 206 without interference. Additionally, the top view image can be used to rotate the plastic housing 206 to an appropriate orientation such that the cut line being inspected faces the side view camera 204. In some embodiments, the rotational movement required to rotate the plastic housing 206 can occur simultaneously with the x-y movement of the side view camera 204 and can not affect the inspection time.

[0158] Once the inspection scenario is determined, imaging of the cut line can begin. The contour x-y points can be used to move a small cylindrical light beam (local structured illumination or SLI) along the contour such that only the cut line above the light path is illuminated. An x-y-rotation stage motion control system or a multi-axis robotic arm can be used to obtain the image. In some embodiments, the plastic housing 206 can be placed on a glass platform and the light beam can be illuminated from below the glass platform to avoid total internal reflection from guiding light to other rows in the field of view.

[0159] Figure 22A An example side view image 2200 captured without a backing screen and without structured light illumination is shown. As depicted, the side view camera captured the cut line 2202 on the first side (e.g., the buccal side) of the appliance without structured light, which caused the cut line 2204 on the opposite side (e.g., the lingual side) of the appliance to be illuminated and interfere with correct cut line detection.

[0160] Figure 22B An example side view image 2210 captured using structured light illumination (e.g., focused light) without a backing screen according to one embodiment is shown. As depicted, the side view camera captured the cut line 2212 on the first side (e.g., the buccal side) of the appliance using directed structured light illumination (LSI), which can suppress the cut line on the opposite side (e.g., the lingual side) and can allow for reliable cut line detection. In cases where the channel between the sides of the appliance is too narrow and / or too angled to insert a backing screen, LSI enables reliable cut line detection.

[0161] Figure 23An example of crack detection in the contour of an image 2300 of an orthodontic appliance captured using focused light is shown. As shown, the focused light uses directed structured light illumination to capture a side view image 2300, which produces a high-contrast delineation of the cut line 2302. The high contrast can be produced by light reaching the cut line surface at a vertical or near-vertical angle. The captured cut line 2302 can be analyzed using the disclosed techniques and cracks 2304 can be identified. In some embodiments, the depth of focus (e.g., 20 microns to 80 microns) can be configured to provide a desired image.

[0162] In embodiments using the backing screen 210 or using focused light without a backing screen, a number of images of the cut line can be captured and "stitched" or registered together to form a composite image 400. In some embodiments, the composite image 400 can be a panorama of each of the front and back of the plastic housing. The side view images of the front and back can be stitched together to show the cut line in a single plane, which can be similar to the unfolding or opening of the cut line of a plastic orthodontic appliance. For example, Figure 4A A side view composite image 400 of the back of a housing according to one embodiment is depicted. The side view composite image 400 includes three images 402, 404, and 406 stitched together in a linear manner. In the example shown, each of the images 402, 404, and 406 was captured at a pixel resolution of 20 microns. Any suitable number of images (e.g., 20) can be taken to produce the front and back composite images.

[0163] Figure 4B An example of using Figure 4A in the side view composite image 400 of the back of the housing is shown for detected example edges 408 (e.g., second feature). The edges 408 can be detected by obtaining a line representing the edges 408 by tracing the cut line in the side view composite image 400. Figure 4C An example comparison of an edge 410 detected from a side view composite image with a virtual or expected edge (e.g., approximate first feature) 412 of a plastic housing from a digital model of the plastic housing is shown. In some embodiments, the expected edge 412 is determined by unfolding the virtual cut line for the dental appliance into a plane for comparison with the edge 410. The edge 410 and the virtual edge 412 can be superimposed and the difference between the two edges can be determined. If the difference exceeds a threshold, it can be determined that a defect is included in the plastic orthodontic appliance and quality control can be performed.

[0164] Figure 5FIG. 500 is a flow chart of a method for determining an inspection plan based on characteristics of a plastic shell according to an embodiment. One or more operations of method 500 are performed by the processing logic of a computing device. The processing logic may include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 500 may be performed by a processing device of an image-based quality control module 2150 that executes Figure 21 Method 500 should be noted that it can be executed for each unique orthodontic appliance manufactured for each patient's treatment plan.

[0165] At block 502, the processing logic may determine an inspection plan based on at least one of first information associated with a first image of the plastic shell or second information associated with a digital file associated with the plastic shell. In some embodiments, the first image may be a top view image, a side view image, or a diagonal image of the plastic shell. The first information associated with the first image and / or the second information associated with the digital file may include the dimensions of the plastic shell, the shape of the plastic shell, and / or one or more features of the plastic shell. The one or more features may include at least one of the following: the precise cutting line of the plastic orthodontic appliance, the cavity of the plastic orthodontic appliance associated with the attachment, the angle of the cutting line of the plastic orthodontic appliance, the distance between the cavities of the plastic orthodontic appliance associated with the teeth (e.g., tooth crowding), or the distance between the cavities of the plastic orthodontic appliance associated with the attachments (e.g., attachment crowding), the distance between the front and back of the plastic shell, or the thickness of the plastic shell. Block 502 may include performing the operations of blocks 504, 506, 508, and 510.

[0166] At block 504, the processing logic may use at least one of the first information or the second information to determine one or more features of the plastic shell. At block 506, the processing logic may determine one or more additional images to be generated for the inspection plan based on the one or more features. As described further below, the determination of generating additional images for the inspection plan may be pre-determined for the plastic shell or performed dynamically using a model and / or a rule engine. In one example, if the plastic orthodontic appliance includes a precise cutting line as a feature, the processing logic may determine to generate an additional image at a location associated with the precise cutting line. If the angle of the cutting line is higher than a specific threshold angle, the processing logic may determine to generate an additional image at a location associated with the cutting line having an excessive angle. If the distance between the teeth is greater than a threshold, the processing logic may determine to generate an additional image at the location of the plastic shell associated with those teeth because the plastic may be thinner and more prone to cracking at that location.

[0167] At block 508, the processing logic can determine the dimensions of the plastic housing from at least one of the first information or the second information. In some embodiments, the processing logic can also determine the shape of the plastic housing from at least one of the first information or the second information. At block 510, the processing logic can determine settings for generating one or more additional images based on at least one of one or more features, the dimensions of the plastic housing, or the shape of the plastic housing. The settings can include at least one of the orientation of one or more imaging devices, the zoom of one or more imaging devices, or the focus of one or more imaging devices. For example, a smaller-sized plastic housing may result in a greater zoom setting and a greater depth of focus. If there is a cavity associated with an attachment identified in the first information or the second information, the orientation (e.g., positioning and angle) of the camera can be configured to capture a suitable image of the cavity.

[0168] In some embodiments, the settings can be predetermined for a first set of additional images that can be generated for each appliance by default and / or the settings can be dynamically configured for a second set of additional images that can be determined dynamically for an inspection protocol. For example, some defects such as air bubbles on the surface of the plastic housing may be detected at a particular depth of focus (e.g., 20 microns), and thus, a predetermined set of images can be configured to be captured on the front and back of the plastic housing to determine if there are air bubbles.

[0169] At block 512, the processing logic can perform an inspection protocol by capturing one or more additional images using the settings. The one or more additional images can be analyzed to determine if one or more defects are included in the plastic housing. If so, quality control can be performed on the plastic housing.

[0170] Figure 6 A flowchart of a method 600 for determining one or more additional images to be generated based on an output from a model according to one embodiment is shown. One or more operations of method 600 are performed by the processing logic of a computing device. The processing logic can include hardware (e.g., circuitry, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 600 can be performed by a processing device of an image-based quality control module 2150 that executes Figure 21 It should be noted that method 600 can be performed for each unique appliance manufactured for a treatment plan for each patient.

[0171] At block 602, the processing logic can apply a digital file as input to a model. The model can allow for targeted inspections by identifying high-risk areas of defects in a plastic housing. For example, at block 604, the processing logic can generate an output by the model that identifies one or more locations of the plastic housing that are identified as high-risk areas for one or more defects. At block 606, the processing logic can determine one or more additional images to generate for an inspection scenario based on the one or more locations identified as high-risk areas by the output.

[0172] In some embodiments, the model can be a machine learning model that is trained to identify one or more high-risk areas of one or more defects at one or more locations of a plastic housing. The processing logic can train the machine learning model to generate a trained machine learning model. A machine learning model can refer to a model artifact created by a training engine using training data (e.g., training inputs and corresponding target outputs). Training can be performed using a set of training data that includes at least one of the following: a) digital files of a first set of plastic appliances with labels that indicate whether each of the first set of plastic appliances has experienced one or more defects, or b) digital files of a second set of plastic appliances with labels that indicate whether each of the second set of plastic appliances includes one or more possible defects. The actual defects of the appliances can be reported by manufacturing technicians, by an automated manufacturing system, and / or by patients. This historical data regarding the actual defects on the physical appliances can then be added as labels or metadata to the associated digital files of the appliances and / or images of the appliances. The possible defects of the digital files of the appliances can be determined by processing the digital design of the model using numerical simulation, as further described below. For example, numerical simulation can be used to process the digital files of the appliances to determine possible defects. Digital files of appliances with associated defects (as provided by real-world data) and digital files of appliances with associated possible defects (as provided by the output of numerical simulation) can be used together to generate a robust machine learning model that can predict the possible defects of new appliances from those digital files of the appliances.

[0173] A machine learning model can consist of a single level of linear or non-linear operations (e.g., a support vector machine (SVM) or a single-level neural network), or can be a deep neural network consisting of multiple levels of non-linear operations. Examples of deep networks and neural networks include convolutional neural networks and / or recurrent neural networks with one or more hidden layers. Some neural networks can consist of interconnected nodes, where each node receives an input from a previous node, performs one or more operations, and sends the resulting output to one or more other connected nodes for further processing.

[0174] As mentioned, information related to whether a plastic appliance has experienced a defect can be obtained from a patient's historical feedback. For example, a patient can provide a report indicating a defect in the plastic appliance, and the location of the defect can be determined (e.g., from the report, from scanning the appliance, etc.). Also, the patient can specify which appliance (e.g., top or bottom) failed at a particular stage of the treatment plan. In some cases, the patient can return the defective appliance to a site, and the defective appliance can be scanned at that site to obtain an image of a digital model of the plastic appliance including the location of the defect. Thus, images of defective appliances can be collected for an image corpus (a set of image corpora, which can include a large set of images) and used as part of the training data. Information provided by the patient about the defective appliance or information determined from the scanned images can be correlated to determine the ID of the appliance, which can then be used to obtain the digital file of that particular appliance. The location of the defect can be placed in the digital file of the plastic appliance with a label indicating that a defect exists at that location.

[0175] Digital files can be applied as inputs to a predictive model that uses numerical simulation. Numerical simulation can be performed on a digital file of an orthodontic appliance to simulate one or more forces on the orthodontic appliance. In some embodiments, the force simulation removes the appliance from the teeth or the mold. The numerical simulation can determine when the amount of force required to remove the appliance from the mold or dental arch reaches a stress or strain level that exceeds a threshold at any point on the orthodontic appliance, which can indicate that a particular point will break. In some embodiments, a strain or stress threshold can be used during the numerical simulation to determine when a point on the digital design of the appliance is likely to fail. In this way, the numerical simulation can be used as a predictive model that predicts possible defects on the digital file of the appliance by identifying one or more high-risk areas of the defect. This simulation can be run dynamically on the digital file of the orthodontic appliance to identify high-risk areas of the defect to allow targeted inspection at those locations. Additionally, these simulations can be run multiple times on multiple digital files of the orthodontic appliance and can include, with the digital file, a label indicating whether the digital file includes one or more possible failure points. Digital files that include a label indicating whether the digital file includes one or more possible defects can be used as inputs to train a machine learning model.

[0176] The numerical simulation can include a finite element method, a finite difference method, a finite volume method, a meshless method, a smoothed particle Galerkin method, a combination of these methods, and the like. The finite element method (also known as finite element analysis) can refer to a numerical method for solving structural problems related to an appliance by generating approximate values of unknowns at a discrete number of points over a domain using a series of partial differential equations. The finite difference method can refer to a numerical method for solving a differential equation by approximating the differential equation with a difference equation and calculating the approximate values at discrete points. The finite volume method can refer to a method of representing and evaluating a partial differential equation in the form of an algebraic equation. The finite volume method can also calculate values (e.g., strain, force) at discrete locations on the mesh geometry of the digital design of the appliance. "Finite volume" can refer to a small volume surrounding each point on the grid. The meshless method can refer to a method based on the interaction of nodes or points with all adjacent nodes or points. In other words, the meshless method does not require connections between the nodes of the simulation domain. The smoothed particle Galerkin method can be a form of the meshless method.

[0177] Figure 7 A flowchart of a method 700 for determining an inspection plan using a rule engine according to one embodiment is shown. One or more operations of method 700 are performed by the processing logic of a computing device. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 700 can be performed by executingFigure 21 is performed by the processing device of the image-based quality control module 2150. It should be noted that method 700 can be performed for each unique orthotic device manufactured for a patient's treatment plan.

[0178] At block 702, the processing logic can generate one or more rules for the rule engine. The rules can be generated based on at least one of the following: a) historical data (e.g., images, reports, etc. of defective orthotic devices), the historical data including a set of reported defects of plastic shells and the locations of the reported defects on the set of plastic shells, b) a digital file of a set of plastic shells with labels indicating whether each of the set of plastic shells has undergone a defect, or c) a digital file of a set of plastic shells with labels indicating whether each of the set of plastic shells includes a likelihood of a defect existing in the plastic shell. The rules can be determined based on observations, outputs of numerical simulations, etc. For example, a customer can provide a report that describes an orthotic device that broke or included another defect during removal, a manufacturing technician can observe damage to the orthotic device during removal from the mold, etc. Hundreds or thousands of observations of orthotic devices that include defects can be used to determine the patterns or combinations of features that may cause defects in defective orthotic devices. A rule can be determined that specifies that a defect may exist when these patterns or combinations of features are present in subsequent designs. Additionally, numerical simulations can be performed and possible defects can be identified as outputs. The outputs from hundreds or thousands of numerical simulations can be aggregated, and patterns or combinations of features associated with possible defects can be identified. A rule can be determined that specifies that a defect may exist when these patterns or combinations of features are present in subsequent designs.

[0179] At block 704, the processing logic can determine an inspection scenario based on at least one of first information associated with a first image of the plastic housing or second information associated with a digital file. Block 706 can include performing the operations of blocks 706, 708, and 710. At block 706, the processing logic can apply the digital file to a rule engine using one or more rules. At block 708, the processing logic can determine whether one or more features, one or more defects, and / or one or more possible defects are included at one or more locations in the plastic housing. For example, the processing logic can determine whether one or more defects or one or more possible defects (e.g., high-risk areas of defects) are detected based on the first information associated with the first image of the plastic housing and / or based on the second information associated with the digital file. The rules can specify that a defect or a possible defect associated with a feature exists at certain locations. Additionally, if a defect is detected in the plastic housing, the rules can specify capturing certain images. If no defect or possible defect is included at one or more locations, method 700 can end. If it is determined that one or more defects or one or more possible defects exist at one or more locations, at block 710, the processing logic can specify generating one or more additional images of the plastic housing according to one or more rules.

[0180] The rules can include rules associated with a group of features (e.g., multiple features within a threshold proximity of each other) and / or with individual parameters. The processing logic can determine the features of the plastic appliance based on the first information associated with the first image or the second information associated with the digital file of the plastic appliance. The features can include at least one of the following: the angle of a cut line at a location of the plastic appliance associated with an interproximal region of the patient's dental arch, the curvature of the plastic appliance, the thickness of the plastic appliance, the undercut height associated with an attachment of a tooth of the patient's dental arch, a precise cut line, the distance between cavities of the plastic appliance associated with an attachment of a tooth of the patient's dental arch, and / or the number of cavities of the plastic appliance. Any one or combination of these features can indicate a high-risk area where a defect may exist in the plastic appliance. Thus, the processing logic can determine to generate one or more additional images in the inspection scenario at the high-risk areas of possible defects in the plastic appliance.

[0181] Figure 8A A flowchart of a method 800 for performing image-based quality control on a housing according to one embodiment is shown. One or more operations of method 800 are performed by the processing logic of a computing device. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 800 can be performed by executing Figure 21The processing apparatus of the image-based quality control module 2150 performs. It should be noted that method 800 can be executed for each unique appliance manufactured for a patient's treatment plan.

[0182] At block 802, the processing logic can obtain one or more images of a first shell (e.g., an appliance). As described above, the first shell may have been manufactured for a patient's dental arch. The first shell can be received by an automatic supply mechanism at an image-based quality control (IBQC) station or placed in the IBQC station by a user. The IBQC station can include one or more cameras and a stationary or rotating table on which the received shell is placed. The IBQC station can also include an illumination system configured to provide consistent exposure and image capture with consistent environmental parameters. In one embodiment, the processing logic can configure the position of the cameras so that one camera obtains a top view image of the shell while another camera obtains a side view image and / or a diagonal view image of the shell. The rotating worktable can enable the shell being inspected at the IBQC station to be rotated so that images can be obtained from different sides of the shell.

[0183] The obtained images can include a first image that includes a component number identifier of the first shell. In some embodiments, an image of the component number identifier with a light (e.g., white) background can facilitate reading a laser mark that identifies the component number identifier (e.g., a barcode, a serial number, etc.). The image of the appliance for quality control analysis can have a dark (e.g., black) background and the appliance is evenly illuminated. Illuminating the appliance with a dark background can help distinguish the edges and shape of the appliance for quality control analysis.

[0184] At block 804, a technician or the processing logic can use the image to identify an identifier (laser mark) on the first shell. For example, in some embodiments, the processing logic can use optical character recognition to read a serial number or other text to identify the component number identifier of the imaged appliance. Optionally, a technician can identify and enter the component number identifier by visually inspecting the appliance. As described above, the identifier can represent a component number and can be laser marked on the appliance. The processing logic can use the first image with a light background to identify the identifier. The identifier can be associated with a digital model of the appliance generated by the processing logic. In particular, before receiving an image of the shell, the processing logic can receive a file that includes a digital model of the mold used to create the specific appliance being inspected.

[0185] At block 806, the processing logic can determine a first digital model of a first shell from a set of digital models of shells based on an identifier. Each digital model in the set of digital models is for a specific shell customized for a specific patient at a specific stage of a patient treatment plan. The digital model of the shell can be generated based on the digital model of a mold at each corresponding stage of the patient's treatment plan, as discussed in detail in method 1600 of reference Figure 16 .

[0186] At block 808, the processing logic can compare an image of the first shell with a projection of the first digital model. In one embodiment, the processing logic can compare a top view image of the first shell with the top view of the first digital model to determine whether the shape of the first shell is deformed. If one or more differences between a first shape of a first projection of the digital model of the shell and a second shape of the first shell exceed a first threshold, the processing logic can determine that the shape of the first shell is deformed, as further described in methods 900 and 1100 of reference Figure 9A and Figure 11 and as shown in the examples of Figure 10 , Figure 12 and Figure 13A-13B .

[0187] However, if one or more differences do not exceed the first threshold, the processing logic can perform additional comparisons. For example, the processing logic can generate a modified projection of the digital model by deforming a first shape of a first projection of the digital model of the first shell towards the second shape of the first shell to approximately match the second shape of the first shell. The processing logic can determine whether one or more remaining differences between a third shape of the modified projection and the second shape of the first shell exceed a second threshold. If so, the processing logic can determine that the cutting line (or other feature) of the first shell is deformed, as further described in method 1400 of reference Figure 14A and as shown in the examples of Figure 15A-15C .

[0188] At block 810, the processing logic can perform quality control on the first shell based on the comparison. The results of the IBQC analysis can be compiled, and a classification can be assigned to the appliance being inspected. If any comparison indicates a manufacturing defect, the processing logic can classify the appliance as defective. If each comparison does not indicate a defect, the processing logic can classify the appliance as non - defective. Optionally, when the analysis is inconclusive, the system can indicate that the appliance needs to be further inspected by a technician. The results can be presented to the user in a user interface.

[0189] Figure 8BFIG. 820 is a flow chart showing another method 820 for performing image-based quality control on a housing. One or more operations of method 820 are performed by the processing logic of a computing device. The processing logic may include hardware (e.g., circuitry, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 820 may be performed by a processing device executing Figure 21 the image-based quality control module 2150. It should be noted that method 820 may be performed for each unique orthodontic appliance manufactured for a treatment plan for each patient.

[0190] At block 822, the processing logic may obtain one or more images of a first housing (e.g., an orthodontic appliance). As described above, the first housing may have been manufactured for a patient's dental arch. The first housing may be received by an automated supply mechanism at an image-based quality control (IBQC) station or placed by a user in the IBQC station. The IBQC station may include one or more cameras and a stationary or rotating table on which the received housing is placed. The IBQC station may also include a lighting system configured to provide consistent exposure and image capture with consistent environmental parameters. In one embodiment, the processing logic may configure the position of the cameras such that one camera obtains a top view image of the housing and another camera obtains a side view image and / or a diagonal view image of the housing. The rotating worktable may enable the housing being inspected at the IBQC station to be rotated so that images from different sides of the housing can be obtained.

[0191] The obtained images may include a first image that includes a component number identifier of the first housing. In some embodiments, an image of the component number identifier with a light (e.g., white) background may facilitate reading a laser mark that identifies the component number identifier (e.g., a barcode, serial number, etc.). The image of the orthodontic appliance for quality control analysis may have a dark (e.g., black) background and the orthodontic appliance is evenly illuminated. Illuminating the orthodontic appliance with a dark background may facilitate distinguishing the edges and shape of the orthodontic appliance for quality control analysis.

[0192] At block 824, a technician or the processing logic may use the image to identify an identifier (laser mark) on the first housing. For example, in some embodiments, the processing logic may use optical character recognition to read a serial number or other text to identify the component number identifier of the imaged orthodontic appliance. Optionally, a technician may visually inspect the orthodontic appliance to identify and enter the component number identifier. As described above, the identifier may represent a component number and may be laser marked on the orthodontic appliance. The processing logic may use the first image with a light background to identify the identifier. The identifier may be associated with a digital file.

[0193] At block 826, the processing logic can determine a first digital file associated with the first shell from a set of digital files based on an identifier. Each digital file in the set of digital files includes a digital model of at least one of a shell (e.g., an appliance) or a mold for manufacturing an appliance. Each digital file is specific to a particular shell customized for a particular patient at a specific stage of a patient treatment plan.

[0194] In one embodiment, the digital file associated with the identifier includes a digital model of the first shell (e.g., an appliance), which is dynamically generated by the processing logic or received from another source. The digital model of the first shell can be dynamically generated by manipulating the digital model of the mold for manufacturing the first shell. The digital model of the first shell can be generated by simulating the process of thermoforming a film on the digital model of the mold by expanding the digital model of the mold into an expanded digital model (e.g., by scaling or inflating the surface of the digital model). Additionally, the generation of the digital model of the first shell can include calculating the projection of a cut line onto the expanded digital model, virtually cutting the expanded digital model along the cut line to create a cut expanded digital model, and selecting the outer surface of the cut expanded digital model. In one embodiment, the digital model of the first shell includes the outer surface of the first shell, but does not necessarily have a thickness and / or does not include the inner surface of the first shell, although in other embodiments it can include a thickness or an inner surface.

[0195] In one embodiment, the digital file includes a mold for manufacturing the first shell. In one embodiment, the digital file can include multiple files associated with the first shell, where the multiple files include a first digital file and a second digital file, the first digital file includes a digital model of the mold, and the second digital file includes a digital model of the first shell. Alternatively, a single digital file can include both a digital model of the mold and a digital model of the first shell.

[0196] At block 828, the processing logic determines an approximate first characteristic of the first shell from the first digital file. In one embodiment, the approximate first characteristic is based on the projection of the digital model of the first shell onto a plane defined by an image of the first shell. In one embodiment, the approximate first characteristic is based on the manipulation of the digital model of the mold used to create the first shell. For example, in some embodiments, the approximate first characteristic can be based on the projection of the digital model of the mold onto a plane defined by an image of the first shell. In this case, the projection of the mold can be scaled or otherwise inflated to approximate the projection of the orthodontic appliance thermoformed on the mold. In another embodiment, the approximate first characteristic is based on the manipulation of the digital model, wherein the manipulation causes the outer surface of the digital model to have an approximate shape of the first shell and is also based on the projection of the outer surface of the digital model onto a surface defined by an image of the first shell. In some embodiments, the approximate first characteristic can include the approximate outer surface of the first shell. The approximate outer surface of the first shell can be referred to as the digital model of the first shell. In some embodiments, the approximate first characteristic can include a first shape of the projection of the approximate outer surface of the first shell onto a plane defined by an image of the first shell.

[0197] At block 830, the processing logic determines a second characteristic of the first shell from one or more images. The image of the first shell can define a plane. The second characteristic can include a second shape of the first shell or its projection. The second characteristic can be determined directly from one or more images (e.g., top view, side view, etc.). In one embodiment, the contour of the second shape is drawn from the image.

[0198] At block 832, the processing logic can compare the approximate first characteristic with the second characteristic. If one or more differences between the approximate first characteristic and the second characteristic exceed a first threshold, the processing logic can determine that the first shell is deformed, as further described in method 920 below with reference to Figure 9B and as shown in the example in Figure 10 . For example, if the first shape does not approximately match the second shape, it can be determined that the first shell is deformed.

[0199] However, if one or more differences do not exceed the first threshold, the processing logic can perform an additional comparison. For example, the processing logic can generate a modified projection of the approximate outer surface of the first shell by deforming the curvature of the projected first shape towards the second curvature of the second shape of the first shell such that the curvature of the deformed projected first shell approximately matches the second curvature of the second shape of the first shell. The processing logic can determine whether one or more additional differences between the deformed first shape and the second shape exceed a second threshold. If so, the processing logic can determine that the cutting line (or other characteristic) of the first shell is inaccurate, as further described below with reference to Figure 14B and14C as described in method 1420 and method 1440 in

[0200] At block 834, the processing logic can perform quality control on the first housing based on a comparison. The results of the IBQC analysis can be compiled, and a category can be assigned to the housing being inspected. If any comparison indicates a manufacturing defect, the processing logic can classify the housing as defective. If each comparison does not indicate a defect, the processing logic can classify the housing as non-defective. Optionally, when the analysis is inconclusive, the system can indicate that the housing needs to be further inspected by a technician. The results can be presented to the user in a user interface.

[0201] Figure 9A FIG. shows a flowchart of a method 900 for determining whether the shape of a housing is deformed according to one embodiment. One or more operations of method 900 are performed by the processing logic of a computing device. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 900 can be performed by a processing device of an image-based quality control module 2150 that executes Figure 21 . Method 900 can be executed to determine whether the shape of the housing is deformed. It should be noted that in some embodiments, before executing method 900, the processing logic may have executed blocks 802, 804, and 806 of method 800 (e.g., obtained a top view image of the first housing and determined a first digital model for the first housing based on an identifier).

[0202] At block 902, the processing logic can determine a plane associated with the top view image of the first housing. The top view image of the first housing can include a two-dimensional object or a three-dimensional object that includes pixels representing an image of the first housing located in the image plane. At block 904, the processing logic can project a first digital model of the first mold (or a manipulated digital model of the mold used to manufacture the first housing) into the determined plane to generate a first projection. For example, as Figure 10 shown, a first projection 1000 of the digital model is projected onto an image 1002 of an orthodontic appliance. In some embodiments, the first projection 1000 is projected into the same plane as the image 1002 of the first housing such that the first projection 1000 covers the image 1002. Method 900 will be discussed with reference to the first digital model of the first orthodontic appliance. However, it should be understood that the described operations work equally well using a manipulated digital model of the mold of the first housing.

[0203] At block 906, the processing logic may identify one or more differences between the first shape of the first projection 1000 and the second shape of the first housing based on the comparison performed at block 808 of method 800. In some embodiments, the processing logic may identify one or more differences by determining (block 908) one or more regions where the first shape of the first projection 1000 does not match the second shape of the first housing. The processing logic may also determine (block 910) differences in the regions (e.g., at least one of the thickness of one or more regions or the area of one or more regions).

[0204] At block 912, the processing logic may determine whether one or more differences (e.g., thickness, area, etc.) between the first shape of the first projection 1000 and the second shape of the first housing exceed a first threshold. The first threshold may be any suitable configurable amount (e.g., a thickness greater than three millimeters (mm), 5 mm, 10 mm, a region having an area greater than one hundred square millimeters, etc.). At block 914, the processing logic may determine whether the first housing is deformed based on whether the one or more differences exceed the first threshold. When the differences exceed the first threshold, the processing logic may classify the appliance as deformed. When the one or more differences do not exceed the first threshold, the processing logic may determine that the shape of the first housing is not deformed and may continue with additional quality control (e.g., cut line deformation detection).

[0205] Figure 9B A flowchart of another method 920 for determining whether the shape of a housing is deformed according to one embodiment is shown. One or more operations of method 920 are performed by the processing logic of a computing device. The processing logic may include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 820 may be performed by a processing device of the image-based quality control module 2150 that Figure 21 is executed. Method 920 may be executed to determine whether the shape of the housing is deformed. It should be noted that in some embodiments, before executing method 920, the processing logic may have executed blocks 822, 824, 826, and 828 of method 820 (e.g., obtained a top view image of the first housing, identified an identifier of the first housing, determined a first digital model for the first housing based on the identifier, and determined approximate first characteristics of the first housing from the first digital file).

[0206] At block 922, the processing logic may determine a plane associated with the top view image of the first housing. The top view image of the first housing may include pixels representing an image of the first housing located in the image plane. At block 924, the processing logic may calculate a projection of the approximate outer surface of the first housing into the first plane. For example, as Figure 10As shown, a first projection 1000 of an approximate outer surface of the first shell is projected onto an image 1002 of the appliance. In some embodiments, the first projection 1000 is projected into the same plane as the image 1002 of the first shell such that the first projection 1000 covers the image 1002. In some embodiments, the digital model of the mold of the first shell is manipulated by inflating or expanding the size of the digital model, wherein the amount of inflation or expansion is based on the thickness of the first shell. The inflated or expanded digital model of the mold can then be cut along a cutting line to calculate the approximate outer surface of the first shell. Then, at block 924, the approximate outer surface of the first shell can be projected onto the plane.

[0207] At block 926, based on the comparison made at block 832 of method 820, the processing logic can identify one or more differences between a first shape of the first projection 1000 and a second shape of the first shell. In some embodiments, the processing logic can identify one or more differences by determining (block 928) one or more regions where the first shape of the first projection 1000 does not match the second shape of the first shell. The processing logic can also determine (block 930) differences in the regions (e.g., at least one of the thickness of one or more regions or the area of one or more regions).

[0208] In block 932, the processing logic can determine whether one or more differences (e.g., thickness, area, etc.) between the second shape of the first shell and the first shape of the first projection 1000 exceed a first threshold. The first threshold can be any suitable configurable amount (e.g., a thickness greater than three millimeters (mm), 5 mm, 10 mm, a region having an area greater than one hundred square millimeters, etc.). At block 934, the processing logic can determine whether the first shell is deformed based on whether one or more differences exceed the first threshold. When the differences exceed the first threshold, the processing logic can classify the appliance as deformed. When one or more differences do not exceed the first threshold, the processing logic can determine that the shape of the first shell is not deformed and can continue with additional quality control (e.g., cutting line deformation detection).

[0209] Figure 11 A flowchart of a method 1100 for determining a shape difference between a digital model of a shell (e.g., an approximate outer surface of a shell without thickness) and an image of the shell according to one embodiment is shown. One or more operations of method 1100 are performed by the processing logic of a computing device. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 1100 can be performed by a processing device of an image-based quality control module 2150 that executes Figure 21 the same.

[0210] In some embodiments, method 1100 may be performed to identify one or more differences between the shape of the first projection 1000 and the shape of the first housing (block 906 of method 900, block 926 of method 920). Figure 12 An example user interface 1200 is shown that depicts the first projection 1000 overlaid on a top view image 1202 of the first housing. Image-based quality control may include coloring an obtained orthodontic appliance image having a dark / black background such that the area occupied by the orthodontic appliance stands out in the plane of the image. The user interface 1200 may be implemented with computer instructions stored on one or more storage devices and executable by one or more processing devices of a computing device (e.g., Figure 21 the computing device 2100 in). The user interface 1200 may be displayed on a display of the computing device.

[0211] At block 1102, the processing logic may generate a contour of the first housing from the top view image 1202. As shown, the first projection 1000 has a first color (e.g., red), and the contour of the first housing has a second color (e.g., blue). The processing logic may set (block 1104) the region 1204 where the contour and the projection 1000 overlap to a third color (e.g., white). The first color, the second color, and the third color may be different from each other. Additionally, the processing logic may determine one or more differences by identifying any regions having the first color or the second color. Based on the first and second colors bordering the third color of the region 1204, the non-overlapping portions of the projection can be easily identified using the user interface 1200. For example, the Figure 12 visible regions in may be identified and certain measurements may be made, such as the thickness, area, and / or perimeter of the region. If the measurement results exceed a first threshold, the processing logic may determine that the orthodontic appliance is deformed.

[0212] The user interface 1200 displays a pop-up message providing various options and results. These options may include connecting to a camera, loading an image from a camera, loading an image from a file, and performing an IBQC process. The results may include a classification (e.g., deformed in this case) and the actual results of the inward and outward parameters. The defective area of the inward parameter is 96.9, and the defective thickness of the inward parameter is 2.13. The defective area of the outward parameter is 60.9, and the defective thickness of the outward parameter is 1.85. Since one or more of these measurements exceed the desired threshold, the processing logic may determine that the orthodontic appliance being inspected is deformed. A technician may read this result and may create another orthodontic appliance, repair the deformation, etc.

[0213] Figure 13A-13BAn additional example comparison of the contour of a digital model of an appliance (e.g., the approximate outer surface of the appliance without thickness) and the contour of an image of the appliance is shown to detect deformation. Figure 13A A top view image 1202 of an appliance having a first line 1300 and a second line 1302 is depicted. The first line 1300 represents the contour edge of a first projection 1000 of the digital model of the appliance, and the second line 1302 represents the contour edge of the image 1202 of the appliance. In some cases, the first line 1300 may be set to a first color (e.g., red), while the second line 1302 may be set to a second color different from the first color (e.g., blue). A defective area of the appliance can be determined to be between the first line 1300 and the second line 1302. As described above, one or more measurements can be obtained from the defective area. For example, the thickness of the defective area, the area of the defective area, etc. If the measurement exceeds a threshold, it can be determined that the appliance is deformed.

[0214] Figure 13B A close-up view of a portion of the image 1202 is shown. As shown, the first line 1300 representing the contour edge of the digital model of the projection of the appliance does not match the second line 1302 representing the contour edge of the image 1202 of the appliance. Thus, the appliance appears wider than expected based on the first projection of the digital model of the appliance. In this case, the processing logic can measure the thickness, area, or perimeter of the area between the line 1300 and the line 1302 and determine that the appliance is deformed.

[0215] Figure 14A An example of deforming the digital model contour to more closely match the contour of the image of the appliance to detect other manufacturing defects (e.g., cut line variations, debris, sidebands, trimmed attachments, and missing attachments, etc.) according to one embodiment is shown. One or more operations of method 1400 are performed by the processing logic of a computing device. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 1400 can be performed by a processing device of an image-based quality control module 2150 that executes Figure 21 Method 1400 can include operations to detect other manufacturing defects of the housing when the processing logic determines that the difference between the shape of the first projection and the shape of the image of the first housing is within a first threshold. Although method 1400 may be described below as specifically detecting cut line deviations, it should be understood that method 1400 is equally applicable to detecting other manufacturing defects in the appliance.

[0216] For example, at block 1402, the processing logic may determine that one or more differences (e.g., thickness, area, perimeter, etc.) do not exceed a first threshold. Accordingly, the processing logic may perform additional comparisons to identify other deformations. In some embodiments, at block 1404, the processing logic may generate a modified projection of the first digital model of the appliance by deforming the curvature of the first shape of the first projection 1000 to approximate the curvature of the second shape of the first shell. After the curvature deformation, the modified projection may have a new third shape. To generate the modified projection, the processing logic may perform operations at blocks 1406-1414.

[0217] For clarity, Figure 15A-15C is discussed in conjunction with method 1400 because Figure 15A-15C illustrates an example of generating a modified projection to detect a cut line change by deforming a digital model profile to more closely match the profile of an image of an appliance, according to one embodiment. Figure 15A shows a top view of the profile edges of the profile of an image of an appliance and the profile edges of the first projection 1000 of the digital model of the appliance. Figure 15B shows points on an intermediate line that are pulled or offset on an intersection line such that the entire profile of the digital model of the appliance is offset outward to more closely match the profile of the image of the appliance. Figure 15C shows a modified projection that has been deformed such that its curvature approximately matches the curvature of the profile of the second shape of the shell.

[0218] At block 1406 of method 1400, the processing logic may identify an intermediate line 1500 of the first projection 1000 of the digital model of the appliance, as Figure 15A shown. In one example, the intermediate line may be approximated based on the thickness between the contoured edges of the digital model. Multiple pairs of points may be added to the profile edges of the profiles for both the image and the digital model of the appliance. At block 1410, the processing logic may project multiple intersection lines 1502 that perpendicularly intersect the intermediate line 1500, which is also shown in Figure 15A . Pairs of points on the first projection 1000 of the digital model and pairs of points on the image of the appliance may be matched, and the intersection lines 1502 perpendicular to the intermediate line 1500 may intersect the matched pairs and the intermediate line 1500. The intermediate line 1500 may be projected onto the first projection 1000. The intersection lines 1502 may also intersect at least two points on the profile of the first shape of the first projection 1000 and at least two points on the profile of the second shape of the shell. At block 1412, the processing logic may identify the points on the intersection lines 1502 at the intersections between each respective line and the intermediate line 1500.

[0219] At block 1414, the processing logic may move points along the intersection line 1502 to approximate a match between the contour of the first shape of the first projection 1000 and the contour of the second shape of the shell, as Figure 15B shown. The processing logic may offset points on the centerline 1500 of the first projection 1000 of the digital model to move the centerline 1500 of the digital model to a position where it is centered within the image of the appliance. Once the processing logic determines that the centerline 1500 of the first projection 1000 of the digital model is close enough to the center of the image of the appliance, the processing logic may stop offsetting the points. Moving the points may generate a modified projection 1504 having a third shape. The third shape (e.g., the contour of the modified projection 1504) may more closely match the contour of the image 1202 of the appliance, as Figure 15C shown.

[0220] When the contour of the first projection 1000 of the digital model more closely matches the image 1202 of the appliance, the processing logic may compare the image of the cut line of the appliance with the cut line of the modified projection 1504 of the digital model of the appliance to identify (block 1416) remaining differences between the second shape of the first shell and the third shape of the modified projection 1504. For example, the processing logic may identify one or more additional regions where the second shape of the first shell does not match the third shape of the modified projection 1504. The one or more additional regions may correspond to the cut line of the first shell. At block 1418, the processing logic may determine whether the remaining differences exceed a second threshold. For example, up to 1 millimeter of variation between the contour of the modified digital model and the contour of the image may be tolerated. Any variation greater than or equal to 1 millimeter may be determined to be a cut line defect. It should be noted that any suitable measurement threshold may be used. If the measured region exceeds the second threshold, the processing logic may determine that there is a cut line distortion. That is, the processing logic may determine that the cut line of the first shell may interfere with the fit of the first shell on the patient's dental arch. In some embodiments, when cut line distortion is determined, a side view image of the appliance may be used to compare with the side view of the digital model.

[0221] Figure 14B FIG. shows a flowchart of another method 1420 of deforming the contour of an approximate outer surface (e.g., a digital model of an appliance without thickness) of an appliance to more closely match the contour of an image of the appliance to detect cut line variations. One or more operations of method 1420 are performed by the processing logic of a computing device. The processing logic may include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 1420 may be performed by executing Figure 21The processing device of the image-based quality control module 2150 performs. Method 1420 may include performing operations to determine the cutting line deformation of the housing when the processing logic determines that the difference between the shape of the first projection and the shape of the image of the first housing is within a first threshold.

[0222] For example, at block 1422, the processing logic may determine that one or more differences (e.g., thickness, area, perimeter, etc.) between the approximate characteristics of the first housing and the measured characteristics of the first housing do not exceed the first threshold. Thus, the processing logic may perform additional comparisons to identify other deformations. In some embodiments, at block 1424, the processing logic may generate a modified projection of the approximate outer surface of the first housing by deforming the first shape of the first projection 1000 such that the first curvature of the deformed first shape approximately matches the second curvature of the second shape of the first housing. After deformation, the modified projection may have a new third shape. To generate the modified projection, the processing logic may perform the operations at blocks 1426-1424.

[0223] For clarity, Figure 15A-15C is discussed together with method 1420 because Figure 15A-15C illustrates an example of generating a modified projection to detect cutting line variations by deforming an approximate outer surface profile to more closely match the profile of an image of an appliance. Figure 15A shows a top view of the profile edges of the profile of an image of an appliance and the profile edges of the first projection 300 of the approximate outer surface of the appliance. Figure 15B shows points on the intermediate line that are pulled or offset on the intersection line such that the entire profile of the approximate outer surface of the appliance is offset outward to more closely match the profile of the image of the appliance. Figure 15C shows the modified projection that has been deformed such that its shape approximately matches the profile of the shape of the housing.

[0224] At block 1426 of method 1420, the processing logic may identify the intermediate line 1500 of the first projection 1000 of the approximate outer surface of the appliance, as Figure 15A shown. In one example, the intermediate line may be approximated based on the width between the profiled edges of the approximate outer surface. Multiple pairs of points may be added to the profile edges of the profiles of both the image of the appliance and the approximate outer surface. At block 1428, the processing logic may calculate the projections of multiple intersection lines 1502 that perpendicularly intersect the intermediate line 1500, which is also shown in Figure 15AAs shown. Pairs of points on the first projection 1000 of the approximate outer surface and pairs of points on the image of the appliance can be matched, and the intersecting line 1502 perpendicular to the center line 1500 can intersect the matched pairs and the center line 1500. The center line 1500 can be projected onto the first projection 1000. The intersecting line 1502 can also intersect at least two points on the contour of the first shape of the first projection 1000 and at least two points on the contour of the shape of the housing. At block 1430, the processing logic can identify the points at the intersections on the intersecting line 1502 between each respective line and the center line 1500.

[0225] At block 1432, the processing logic can move these points along the intersecting line 1502 to approximate the contour of the first shape of the first projection 1000 to match the contour of the second shape of the housing, as Figure 15B shown. The processing logic can offset the points on the center line 1500 of the first projection 1000 of the approximate outer surface of the first housing to move the center line 1500 of the approximate outer surface to a position where it is centered within the image of the appliance. Once the processing logic determines that the center line 1500 of the first projection 1000 of the approximate outer surface is close enough to the center of the image of the appliance, the processing logic can stop offsetting these points. Moving these points can generate a modified projection 1504 with a deformed first shape that includes a curvature that approximately matches the curvature of the second shape.

[0226] When the contour of the first projection 1000 of the approximate outer surface more closely matches the contour of the image 1202 of the appliance, the processing logic can compare the image of the cutting line of the appliance with the image of the cutting line of the modified projection 1504 of the approximate outer surface of the appliance to identify (block 1434) additional differences between the first curvature of the deformed second shape and the second curvature of the first shape of the first housing. For example, the processing logic can identify one or more additional regions where the second curvature and the first curvature do not match. This comparison can be performed based on the remaining differences between the third shape of the modified projection measured from the top view image and the second shape of the first housing. Alternatively or additionally, a new projection onto a second plane defined by a side view image can be calculated based on the modified projection in the first plane defined by the top view image. One or more additional regions can correspond to the cutting line of the first housing. At block 1436, the processing logic can determine whether the additional differences exceed a second threshold. For example, variations of up to 1 millimeter between the contour of the modified approximate outer surface and the contour of the image can be tolerated. Any variation greater than or equal to 1 millimeter can be determined as a cutting line defect. It should be noted that any suitable measurement threshold can be used. If the measured region exceeds the second threshold, the processing logic can determine that there is a cutting line deformation. That is, the processing logic can determine that the cutting line of the first housing may interfere with the fit of the first housing on the patient's dental arch.

[0227] Figure 14C Shows a flowchart of another method 1440 for deforming the contour of an approximate outer surface of an appliance (e.g., a digital model of an appliance without thickness) to more closely match the contour of an image of the appliance to detect changes in a cut line. One or more operations of method 1440 are performed by processing logic of a computing device. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 1440 can be performed by a processing device of an image-based quality control module 2150 that executes Figure 21 Method 1440 can include an operation of determining a cut line deformation of the housing when the processing logic determines that a difference between the shape of a first projection and the shape of an image of a first housing is within a first threshold.

[0228] At block 1442, the processing logic can generate a modified projection of an approximate outer surface of the first housing by deforming a first shape of the first projection such that a first curvature of the deformed first shape approximately matches a second curvature of a second shape of the first housing. The deformation of the shape of the first projection can be performed as described above.

[0229] At block 1444, the processing logic can determine a second plane associated with a side view image. At block 1446, the processing logic can deform the approximate outer surface of the first housing based on the deformation of the second shape of the first projection. At block 1448, the processing logic can calculate a second projection of the deformed approximate outer surface of the first housing onto the second plane. At block 1450, the processing logic can determine an additional difference between a third shape of the first housing represented in the side view image and an approximate fourth shape of the first housing represented in the second projection.

[0230] At block 1452, the processing logic can determine whether one or more additional differences exceed a second threshold. For example, a change of up to 1 millimeter between the contour of the second projection of the deformed approximate outer surface and the contour of the second image can be tolerated. Any change greater than or equal to 1 millimeter can be determined as a cut line defect. It should be noted that any suitable measurement threshold can be used. If the measured area exceeds the second threshold, the processing logic can determine that there is a cut line deformation. That is, the processing logic can determine that the cut line of the first housing may interfere with the fit of the first housing on the patient's dental arch.

[0231] Figure 16FIG. 1600 is a flow chart of a method 1600 for generating a digital model of a shell (e.g., an approximate outer surface of an orthodontic appliance) that can be included in a digital file. Alternatively or additionally, a digital model of a mold of the shell can be used to perform method 1600 to approximate the characteristics of the shell, which can be compared with the measured characteristics of the shell that can be determined from an image of the shell. Thus, method 1600 can be performed without generating a digital model of the shell. One or more operations of method 1600 are performed by the processing logic of a computing device. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 1600 can be performed by a processing device of an Figure 21 image-based quality control module 2150 that executes. Each digital model of a specific shell at a specific stage of a patient treatment plan can be generated based on manipulating a digital model of a mold of the patient's dental arch at that specific stage of the patient treatment plan. The mold is manufactured using the digital model of the mold (e.g., by 3D printing), and the shell is manufactured using the mold with a thermoforming process. The digital model of the shell may not be used to manufacture the shell.

[0232] At block 1602, the processing logic can simulate the process of thermoforming a film on the digital model of the mold by expanding the digital model of the mold into an expanded digital model. The expansion performed can take into account the thickness of the shell. The inflation or expansion can scale the surface of the digital model of the mold by a predetermined factor. The processing logic can determine a cutting line of the expanded digital model of the shell. In one example, the processing logic can determine a cutting line of the surface of the expanded digital model of the shell by finding the intersection line between the gum and the tooth and modifying the intersection line by raising it so that it does not contact the gum. At block 1604, the processing logic can project the cutting line onto the expanded digital model. In some embodiments, the cutting line can be displayed to the user as a line (e.g., yellow) superimposed on the expanded digital model of the surface of the orthodontic appliance.

[0233] At block 1606, the processing logic may virtually cut the enlarged digital model along a cutting line to create a cut enlarged digital model. That is, the determined cutting line may be used during the simulation process to remove the redundant enlarged surfaces to generate a virtual surface of the appliance by cutting along the determined virtual cutting line. At block 1608, the processing logic may select the outer surface of the cut enlarged digital model as the digital model of the housing. The digital model of the housing may be obtained by simulated trimming, and the digital model may represent the outer surface of the appliance. The digital model of the appliance may be three-dimensional, and various two-dimensional views (e.g., top view, side view) or three-dimensional views may be obtained using the digital model of the appliance. Additionally, the digital model of the appliance may be associated with a component number identifier of the corresponding appliance such that the digital model of the appliance may be retrieved when the processing logic identifies the component number identifier using an image of the manufactured appliance. In an alternative embodiment, a technician may read the component number identifier and input it into the IBQC system. Alternatively, the IBQC system may read other computer-readable tags associated with the appliance for component number identification. The system may receive the component number identifier information and then retrieve a previously generated digital model of the appliance or a digital model of a mold associated with the appliance from a database to generate a digital model of the appliance. Thus, a separate digital model of the appliance may be generated for each appliance at each stage of a patient treatment plan before or during the quality control process. The processing logic may use the identified component number identifier to perform IBQC for each uniquely manufactured appliance.

[0234] Figure 17 FIG. 1700 is a flow diagram of a general method for determining a static position of a digital model of a housing on a flat surface in accordance with one embodiment. The digital model may be included in a digital file, and the digital model may include approximate first characteristics (e.g., the outer surface of the housing). The digital model of the housing may be generated based on manipulation of a digital model of a mold. One or more operations of method 1700 are performed by processing logic of a computing device. The processing logic may include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or combinations thereof. For example, one or more operations of method 1700 may be performed by executing Figure 21The processing device of the image-based quality control module 2150 performs. Method 1700 can be executed to project a first projection of the digital model of the housing onto the same plane associated with the top view of the housing. Method 1700 can also be used when projecting the digital model of the housing onto other planes associated with images of the housing taken from other angles. Method 1700 can be executed, for example, in blocks 904 and / or 924 of methods 900 and / or 920 respectively to correctly calculate the projection of the digital model onto the plane associated with the top view image of the appliance.

[0235] At block 1702, the processing logic can determine the stationary position of the first digital model of the first housing on a flat surface. Figure 18 Method 1800 describes using a two-dimensional digital model to determine the stationary position of the digital model of the housing on a flat surface, Figure 19A-19C which depicts an example of using a two-dimensional digital model to determine the stationary position. Figure 20 Method 2000 describes using a three-dimensional digital model to determine the stationary position of the digital model of the housing on a flat surface. Once the stationary position is determined, the processing logic can calculate the projection of the first digital model with the stationary position onto the plane of the top view image (block 1704).

[0236] Figure 18 FIG. shows a flowchart of method 1800 for using a two-dimensional digital model to determine the stationary position of the digital model of the housing (e.g., the approximate outer surface of the housing) on a flat surface according to one embodiment. One or more operations of method 1800 are performed by the processing logic of a computing device. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 1440 can be performed by the processing device of the Figure 21 image-based quality control module 2150. For clarity, method 1800 and Figure 19A-19C .

[0237] Figure 19A FIG. shows a two-dimensional profile of an arbitrary object 1900. Although the arbitrary object 1900 is depicted, it should be understood that method 1800 can be applied to the profile of the digital model of the housing. At block 1802, the processing logic can determine the centroid 1902 of the object 1900. At block 1804, the processing logic can determine the convex hull 1904 of the object 1900 (e.g., a polygon). The convex hull can include a plurality of vertices that link the outermost points of the object 1900. In particular, three vertices F1, F2, and F3 are identified on the convex hull 1904.

[0238] For each vertex of the convex hull 1904, the processing logic can calculate (block 1806) a line 1906 that contains that vertex. At block 1808, the processing logic can calculate the projection of the centroid 1902 onto that line 1906, as shown by the projection point 1908 in Figure 19B In block 1810, the processing logic can determine whether the projection point 1908 is outside the vertex. For vertex F1, the projection point 1908 is outside vertex F1, so the object 1900 may not rest on side F1 because the object 1900 would roll to the left. The same is true for vertex F2. If the projection point 1908 is outside the vertex, the processing logic can return to repeat blocks 1806, 1808, and 1810 until a rest position is found.

[0239] If the projection point 1908 is inside the vertex, the processing logic can determine that that particular vertex is the rest position of the first digital model of the first shell. Figure 19C A line 1910 that contains vertex F3 is shown. The centroid 1902 is projected onto that line 1910, and as shown, the projection point 1912 is inside vertex F3 that defines the line 1910. Thus, the object 1900 can rest on this side without rolling. Using this method, the processing logic can determine the rest position and identify the appropriate projection of the digital model of the appliance for quality control analysis depending on the image view (e.g., top view, side view, etc.).

[0240] Figure 20 A flowchart of a method 2000 for determining the rest position of a digital model of a shell on a flat surface using a three-dimensional digital model according to one embodiment is shown. One or more operations of method 2000 are performed by the processing logic of a computing device. The processing logic can include hardware (e.g., circuits, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions executed by a processing device), firmware, or a combination thereof. For example, one or more operations of method 2000 can be performed by a processing device of an image-based quality control module 2150 that executes Figure 21 of.

[0241] At block 2002, processing logic may determine the centroid of the first digital model of the first housing. At block 2004, processing logic may determine the convex hull of the digital model (e.g., a polyhedron for a three-dimensional digital model). The convex hull may include a number of faces that link the outermost points of the first digital model. For each face of the convex hull, processing logic may calculate (block 2006) the plane containing that face. At block 2008, processing logic may calculate the projection of the centroid onto that plane. At block 2010, processing logic may determine whether the projected point on the plane lies outside that face. If the projected point lies outside the face, processing logic may return to repeat blocks 2006, 2008, and 2010 until a stationary position is found. If the projected point does not lie outside the face, processing logic may determine (block 2012) that the face is the stationary position of the first digital model.

[0242] Figure 21 A graphical representation of a machine in the example form of a computing device 2100 is shown, where a set of instructions causes the machine to perform any one or more of the methods discussed herein. In some embodiments, the machine may be part of an IBQC site or communicatively coupled to an IBQC site. In alternative embodiments, the machine may be connected (e.g., networked) to other machines in a local area network (LAN), intranet, extranet, or the Internet. For example, the machine may be networked to an IBQC site and / or a rapid prototyping device, such as a 3D printer or an SLA device. The machine may operate in a client-server network environment with the capabilities of a server or client machine, or as a peer machine in a peer-to-peer (or distributed) network environment. The machine may be a personal computer (PC), tablet computer, set-top box (STB), personal digital assistant (PDA), cellular phone, network appliance, server, network router, switch, or bridge, or any machine capable of executing a set of instructions (sequentially or otherwise) that specify actions to be taken by that machine. Further, although only a single machine is shown, the term "machine" shall also be taken to include any collection of machines (e.g., computers) that individually or jointly execute a set (or multiple sets) of instructions to perform any one or more of the methods discussed herein.

[0243] Example computing device 2100 includes a processing device 2102, a main memory 2104 (e.g., read only memory (ROM), flash memory, dynamic random access memory (DRAM) such as synchronous DRAM (SDRAM), etc.), a static memory 2106 (e.g., flash memory, static random access memory (SRAM), etc.), and auxiliary memory (e.g., data storage device 2128), which communicate with each other via a bus 2108.

[0244] The processing device 2102 represents one or more general-purpose processors such as a microprocessor, a central processing unit, etc. More specifically, the processing device 2102 can be a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a processor implementing other instruction sets, or a processor implementing a combination of instruction sets. The processing device 2102 can also be one or more dedicated processing devices such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), a network processor, etc. The processing device 2102 is configured to execute processing logic (instructions 2126) for performing the operations and steps discussed herein.

[0245] The computing device 2100 may also include a network interface device 2122 for communicating with the network 2164. The computing device 2100 may also include a video display unit 2110 (e.g., a liquid crystal display (LCD) or a cathode ray tube (CRT)), an alphanumeric input device 2112 (e.g., a keyboard), a cursor control device 2114 (e.g., a mouse), and a signal generation device 2120 (e.g., a speaker).

[0246] The data storage device 2128 may include a machine-readable storage medium (or more specifically, a non-transitory computer-readable storage medium) 2124, on which a set or multiple sets of instructions 2126 are stored that embody any one or more of the methods or functions described herein. A non-transitory storage medium refers to a storage medium other than a carrier wave. During the execution of the instructions by the computer device 2100, the instructions 2126 may also reside completely or at least partially within the main memory 2104 and / or within the processing device 2102, and the main memory 2104 and the processing device 2102 also constitute computer-readable storage media.

[0247] The computer-readable storage medium 2124 can also be used to store one or more virtual 3D models (also referred to as electronic models) and / or the IBQC module 2150, which can perform one or more operations of the methods described herein. The computer-readable storage medium 2124 can also store a software library containing methods that call the IBQC module 2150. Although the computer-readable storage medium 2124 is shown as a single medium in the exemplary embodiment, the term "computer-readable storage medium" should be regarded as including a single medium or multiple media (e.g., a centralized or distributed database, and / or an associated cache and server) that store a set or multiple sets of instructions. The term "computer-readable storage medium" should also be regarded as including any medium that can store or encode a set of instructions for a machine to execute and cause the machine to perform any one or more of the methods of the present invention. Therefore, the term "computer-readable storage medium" should be regarded as including, but not limited to, solid-state memory, as well as optical and magnetic media.

[0248] It should be understood that the above description is intended to be illustrative and not restrictive. After reading and understanding the above description, many other embodiments will be apparent. Although embodiments of the present invention have been described with reference to specific example embodiments, it should be recognized that the present invention is not limited to the described embodiments, but may be practiced with modifications and alterations within the spirit and scope of the appended claims. Accordingly, the specification and drawings are to be regarded as illustrative rather than restrictive. Thus, the scope of the present invention should be determined with reference to the appended claims and the full scope of equivalents of those claims.

Claims

1. A method for inspecting a shell customized for a patient's dental arch, comprising: obtaining one or more images of the shell; identifying an identifier of the shell; determining a digital file associated with the shell from a plurality of digital files based on the identifier; comparing at least one of the one or more images of the shell with data from the digital file; and performing quality control on the shell based on the result of the comparison.

2. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 1.

3. A system for performing the method of claim 1, the system comprising: a platform that supports the shell during imaging; one or more cameras that capture the one or more images; a lighting system for illuminating the shell during capture of the one or more images; and a processing device configured to perform one or more operations of the method of claim 1.

4. A method for analyzing the quality of an orthodontic appliance, the method comprising: receiving, by a processor, a digital representation of a manufactured orthodontic appliance, the digital representation being generated based on imaging of the manufactured orthodontic appliance; determining a first digital file and a second digital file associated with the manufactured orthodontic appliance, wherein the first digital file includes a digital model of a mold used during manufacture of the manufactured orthodontic appliance, and wherein the second digital file includes trimming information for trimming the manufactured orthodontic appliance along a cutting line; analyzing, by the processor, the digital representation of the manufactured orthodontic appliance based on information from the first digital file and the second digital file to identify quality-related characteristics of the manufactured orthodontic appliance, the analysis including comparing the digital representation of the manufactured orthodontic appliance with the trimming information from the second digital file; determining, based on the quality-related characteristics, that the manufactured orthodontic appliance includes a manufacturing defect; and classifying, based on determining that the manufactured orthodontic appliance includes a manufacturing defect, the manufactured orthodontic appliance as requiring further inspection by a technician.

5. A computer-readable medium comprising instructions that, when executed by a processing device, cause the processing device to perform the method of claim 4.

6. A system for performing the method of claim 4, the system comprising: one or more cameras for generating the digital representation; and a processing device for processing the one or more images.

7. A system for analyzing the quality of an orthodontic appliance, the system comprising: a processor; and a memory coupled to the processor and storing instructions that, when executed by the processor, cause the processor to: receive a digital representation of a manufactured orthodontic appliance, the digital representation having been generated based on imaging of the manufactured orthodontic appliance; determine a first digital file and a second digital file associated with the manufactured orthodontic appliance, wherein the first digital file includes a digital model of a mold used during manufacture of the manufactured orthodontic appliance, and wherein the second digital file includes trimming information for trimming the manufactured orthodontic appliance along a cutting line; Analyze a digital representation of a manufactured orthodontic appliance based on information from a first digital file and a second digital file to identify quality-related characteristics of the manufactured orthodontic appliance, the analysis including comparing the digital representation of the manufactured orthodontic appliance with trimming information from the second digital file; Determine that the manufactured orthodontic appliance includes a manufacturing defect based on the quality-related characteristics; and Classify the manufactured orthodontic appliance as defective based on determining that the manufactured orthodontic appliance includes a manufacturing defect.

8. A quality control system comprising: An image capture device configured to generate a digital representation of a manufactured orthodontic appliance; A light source configured to illuminate the manufactured orthodontic appliance in a manner that enhances the image quality of the digital representation of the manufactured orthodontic appliance; and A computing device for: Determining a first digital file and a second digital file associated with the manufactured orthodontic appliance, wherein the first digital file associated with the manufactured orthodontic appliance includes a digital model of a mold used during the manufacture of the manufactured orthodontic appliance, and wherein the second digital file includes trimming information for trimming the manufactured orthodontic appliance along a cutting line; Analyze a digital representation of a manufactured orthodontic appliance based on information from the first digital file and the second digital file to identify quality-related characteristics of the manufactured orthodontic appliance, the analysis including comparing the digital representation of the manufactured orthodontic appliance with trimming information from the second digital file; Determine that the manufactured orthodontic appliance includes a manufacturing defect based on the quality-related characteristics; and Classify the manufactured orthodontic appliance as defective based on determining that the manufactured orthodontic appliance includes a manufacturing defect.

9. A non-transitory computer-readable medium comprising instructions that, when executed by a processor, cause the processor to perform operations, the operations comprising: Receiving, by the processor, a digital representation of a manufactured orthodontic appliance, the digital representation having been generated based on imaging of the manufactured orthodontic appliance; Determining a first digital file and a second digital file associated with the manufactured orthodontic appliance, wherein the first digital file includes a digital model of a male mold used during the manufacture of the manufactured orthodontic appliance, and wherein the second digital file includes trimming information for trimming the manufactured orthodontic appliance along a cutting line; Analyzing, by the processor, the digital representation of the manufactured orthodontic appliance based on information from the male mold of the first digital file and the second digital file to identify quality-related characteristics of the manufactured orthodontic appliance, the analysis including comparing the digital representation of the manufactured orthodontic appliance with trimming information from the second digital file; Determine that the manufactured orthodontic appliance includes a manufacturing defect based on the quality-related characteristics; and Based on determining that the manufactured orthodontic appliance includes a manufacturing defect, classify, by the processor, the manufactured orthodontic appliance as requiring further inspection by a technician.

10. A method of manufacturing an orthodontic appliance, the method comprising: Manufacturing an orthodontic appliance, wherein manufacturing the orthodontic appliance includes: Printing the mold based on a digital model of the mold associated with the patient's dental arch; Forming an orthodontic appliance on the mold; and Trimming the orthodontic appliance; and Evaluating the quality of an orthodontic appliance, wherein evaluating the quality of the orthodontic appliance includes: Receiving, by a processor, a digital representation of the orthodontic appliance, the digital representation having been generated based on imaging of the orthodontic appliance; Analyzing, by the processor, the digital representation of the orthodontic appliance to identify quality-related characteristics of the orthodontic appliance, the analysis including: Comparing the digital representation of the orthodontic appliance with data generated based on a digital model of the mold on which the orthodontic appliance is formed; and Calculating one or more differences between features determined from the data and features of the digital representation of the orthodontic appliance; Determining, based on the quality-related characteristics, that the orthodontic appliance includes a manufacturing defect; and Based on determining that the orthodontic appliance includes a manufacturing defect, classifying, by the processor, the orthodontic appliance as requiring further inspection by a technician.

11. A system for performing the method of claim 10, the system comprising: A rapid prototyping machine for printing a mold; A thermoforming machine for forming an orthodontic appliance on the mold; A finishing machine for finishing the orthodontic appliance; and An image-based quality control (IBQC) station for evaluating the quality of the orthodontic appliance.

12. A system for manufacturing an orthodontic appliance, the system comprising: A three-dimensional (3D) printer for printing the mold based on a digital model of the mold associated with a patient's dental arch; Thermoforming equipment for thermoforming the orthodontic appliance on the mold; Finishing equipment for finishing the orthodontic appliance; An image capture device for generating a digital representation of the orthodontic appliance; and A processor for: Receiving the digital representation of the orthodontic appliance; Analyzing the digital representation of the orthodontic appliance to identify quality-related characteristics of the orthodontic appliance, wherein analyzing the digital representation of the orthodontic appliance includes: Comparing the digital representation of the orthodontic appliance with data generated based on the digital model of the mold on which the orthodontic appliance is formed; and Calculating one or more differences between features determined from the data and features of the digital representation of the orthodontic appliance; Determining, based on the quality-related characteristics, that the orthodontic appliance includes a manufacturing defect; and Based on determining that the orthodontic appliance includes a manufacturing defect, classifying the orthodontic appliance as defective.

13. A non-transitory computer-readable medium comprising instructions that, when executed by a manufacturing system for an orthodontic appliance, cause the manufacturing system to perform operations, the operations comprising: Manufacturing an orthodontic appliance, wherein manufacturing the orthodontic appliance includes: Printing the mold based on a digital model of the mold associated with a patient's dental arch; Forming the orthodontic appliance on the mold; and Finishing the orthodontic appliance; and Evaluating the quality of the orthodontic appliance, wherein evaluating the quality of the orthodontic appliance includes: Receiving the digital representation of the orthodontic appliance, the digital representation having been generated based on imaging of the orthodontic appliance; Analyzing, by a processor, the digital representation of the orthodontic appliance to identify quality-related characteristics of the orthodontic appliance, the analysis including: Comparing the digital representation of the orthodontic appliance with data generated based on the digital model of the mold on which the orthodontic appliance is formed; and Calculate one or more differences between features determined from the data and features of a digital representation of an orthodontic appliance; Determine that the orthodontic appliance includes a manufacturing defect based on the quality-related feature; and Classify the orthodontic appliance as requiring further inspection by a technician based on determining that the orthodontic appliance includes a manufacturing defect.

14. A method of manufacturing an orthodontic appliance, comprising: Manufacturing an orthodontic appliance, wherein manufacturing the orthodontic appliance includes: Printing the mold based on a digital model of a mold associated with a patient's dental arch; Forming an orthodontic appliance on the mold; and Trimming the orthodontic appliance, wherein after trimming, the orthodontic appliance includes a cut line; and evaluating the quality of the orthodontic appliance, wherein evaluating the quality of the orthodontic appliance includes: Imaging the orthodontic appliance to generate a first digital representation of the orthodontic appliance; Comparing the first digital representation of the orthodontic appliance with a digital file associated with the orthodontic appliance by a processor; Based on the comparison, determining whether a cut line change is detected between the orthodontic appliance and the digital file; and Determining whether there is a manufacturing defect in the orthodontic appliance based on whether the tangent change exceeds a tangent change threshold.

15. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 14.

16. A system for performing the method of claim 14, the system comprising: A rapid prototyping machine for printing a mold; A thermoforming machine for forming an orthodontic appliance on the mold; A trimming machine for trimming the orthodontic appliance; and An image-based quality control (IBQC) station for evaluating the quality of the orthodontic appliance.

17. A method of manufacturing an orthodontic appliance, comprising: Manufacturing an orthodontic appliance, wherein manufacturing the orthodontic appliance includes: Printing the mold based on a digital model of a mold associated with a patient's dental arch; Forming an orthodontic appliance on the mold; and Trimming the orthodontic appliance; and Evaluating the fit of the orthodontic appliance on the patient's dental arch, wherein the fit of the orthodontic appliance on the patient's dental arch includes: Receiving a first digital representation of the orthodontic appliance, the first digital representation having been generated based on imaging of the orthodontic appliance; Analyzing the first digital representation of the orthodontic appliance to identify quality-related features of the orthodontic appliance, wherein the quality-related features are identified at least in part based on whether a cut line change between the first digital representation of the orthodontic appliance and a digital file associated with the orthodontic appliance exceeds a cut line change threshold; and Determining the fit of the orthodontic appliance on the patient's dental arch based on the quality-related features of the orthodontic appliance.

18. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 17.

19. A system for performing the method of claim 17, the system comprising: A rapid prototyping machine for printing a mold; A thermoforming machine for forming an orthodontic appliance on the mold; A trimming machine for trimming the orthodontic appliance; and An image-based quality control (IBQC) station for evaluating the quality of the orthodontic appliance.

20. A method of manufacturing an orthodontic appliance, the method Comprising: Manufacturing an orthodontic appliance; And Evaluating the quality of the orthodontic appliance, wherein evaluating the quality of the orthodontic appliance includes: Receiving, by a processor, a digital representation of the orthodontic appliance, the digital representation having been generated based on imaging of the orthodontic appliance; Analyzing, by the processor, the digital representation of the orthodontic appliance to identify quality-related characteristics of the orthodontic appliance, the analysis including: Comparing the digital representation of the orthodontic appliance with data based on a digital model associated with the orthodontic appliance; and Calculating one or more differences between features determined from the data and features of the digital representation of the orthodontic appliance; Determining, based on the quality-related characteristics, that the orthodontic appliance includes possible manufacturing defects; and Based on determining that the orthodontic appliance includes possible manufacturing defects, classifying, by the processor, the orthodontic appliance as requiring further inspection by a technician.

21. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 20.

22. A system for performing the method of claim 20, the system Comprising: A manufacturing system for forming an orthodontic appliance; And An image-based quality control (IBQC) station for evaluating the quality of the orthodontic appliance.

23. A method for inspecting manufacturing defects of a customized orthodontic appliance, the customized orthodontic appliance being customized for a specific dental arch of a specific patient and a specific stage of orthodontic treatment, the method Comprising: Manufacturing the customized orthodontic appliance; And Evaluating the quality of the customized orthodontic appliance, wherein evaluating the quality of the customized orthodontic appliance includes: Obtaining one or more images of the customized orthodontic appliance; Determining, from a plurality of digital files, a digital file associated with the customized orthodontic appliance, the digital file associated with the customized orthodontic appliance including a digital model of at least one of the customized orthodontic appliance or a mold used during the manufacture of the customized orthodontic appliance; Determining the expected characteristics of the customized orthodontic appliance based on the digital file; Determining the actual characteristics of the customized orthodontic appliance from the one or more images of the customized orthodontic appliance; Determining whether there are possible manufacturing defects in the customized orthodontic appliance by comparing the expected characteristics of the customized orthodontic appliance with the actual characteristics of the customized orthodontic appliance; and Outputting an output associated with determining whether there are possible manufacturing defects.

24. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 23.

25. A system for performing the method of claim 23, the system Comprising: A manufacturing system for forming an orthodontic appliance; And An image-based quality control (IBQC) station for evaluating the quality of the orthodontic appliance.

26. A method for inspecting manufacturing defects of a customized dental device, Wherein, The customized dental device is associated with a dental application and is customized for a specific patient, the method comprising: Obtaining one or more images of the customized dental device; Identify one or more digital files associated with a customized dental device, the one or more digital files including at least one of a first digital model of the customized dental device or a second digital model of a mold used during the manufacture of the customized dental device; Determine the expected characteristics of the customized dental device based on at least one of the first digital model or the second digital model; Determine the actual characteristics of the customized dental device from the one or more images of the customized dental device; Determine whether at least one of a gingival cut line change, an arch change, or a bend of the customized dental device is detected between the customized dental device and at least one of the first digital model or the second digital model by comparing the expected characteristics of the customized dental device with the actual characteristics of the customized dental device; Determine whether there is a possible manufacturing defect in the customized dental device based on whether at least one of the gingival cut line change exceeds a gingival cut line change threshold, the arch change exceeds an arch change threshold, or the bend exceeds a bend threshold; and Output an output associated with determining whether there is a possible defect in the customized dental device.

27. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 26.

28. A system for performing the method of claim 26, the system comprising: A manufacturing system for forming an orthodontic appliance; and An image-based quality control (IBQC) station for evaluating the quality of the orthodontic appliance.

29. A method of manufacturing an orthodontic appliance, comprising: Manufacturing an orthodontic appliance; and Evaluating the quality of the manufactured orthodontic appliance, wherein evaluating the quality of the manufactured orthodontic appliance comprises: Imaging the manufactured orthodontic appliance to generate a first digital representation of the manufactured orthodontic appliance; Comparing, by a processor, the first digital representation of the manufactured orthodontic appliance with a digital file associated with the manufactured orthodontic appliance; Determining, based on the comparison, whether at least one of the following exists: a) A cut line change is detected between the manufactured orthodontic appliance and the digital file, b) An arch change is detected between the manufactured orthodontic appliance and the digital file, c) A bend of the manufactured orthodontic appliance is detected between the manufactured orthodontic appliance and the digital file; and Determining whether there is a possible defect in the manufactured orthodontic appliance based on at least one of the following: a) Whether the cut line change exceeds a cut line change threshold, b) Whether the arch change exceeds an arch change threshold, or c) Whether the bend of the manufactured orthodontic appliance exceeds a bend threshold.

30. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 29.

31. A system for performing the method of claim 29, the system comprising: A manufacturing system for forming an orthodontic appliance; and An image-based quality control (IBQC) station for evaluating the quality of the orthodontic appliance.

32. A method of manufacturing an orthodontic appliance for a patient, comprising: Manufacturing an orthodontic appliance for the patient's dental arch; and Evaluating the fit of an orthodontic appliance on a patient's dental arch, wherein evaluating the fit of the orthodontic appliance on the patient's dental arch includes: Receiving a first digital representation of the orthodontic appliance, the first digital representation having been generated based on imaging of the orthodontic appliance; Analyzing the first digital representation of the orthodontic appliance to identify quality-related characteristics of the orthodontic appliance, wherein the quality-related characteristics are identified based at least in part on at least one of the following: a) Whether the change in the cutting line between the first digital representation of the orthodontic appliance and a digital file associated with the orthodontic appliance exceeds a cutting line change threshold, b) Whether the change in the dental arch between the first digital representation of the orthodontic appliance and a digital file associated with the orthodontic appliance exceeds a dental arch change threshold, or c) Whether the bend of the manufactured orthodontic appliance determined based on a comparison between the first digital representation of the orthodontic appliance and a digital file associated with the orthodontic appliance exceeds a bend threshold; and Determining the fit of the orthodontic appliance on the patient's dental arch based on the quality-related characteristics of the orthodontic appliance.

33. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 32.

34. A system for performing the method of claim 32, the system comprising: A manufacturing system for forming an orthodontic appliance; and An image-based quality control (IBQC) station for evaluating the quality of the orthodontic appliance.

35. A method, comprising: Receiving a digital file associated with a plastic shell customized for a patient's dental arch; Generating a first image of the plastic shell using one or more imaging devices; Determining an inspection protocol for the plastic shell based on at least one of first information associated with the first image of the plastic shell or second information associated with the digital file; and Executing the inspection protocol to capture one or more additional images of the plastic shell.

36. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 35.

37. A system for performing the method of claim 35, the system comprising: A manufacturing system for forming the plastic shell; and An image-based quality control (IBQC) station for evaluating the quality of the plastic shell.

38. A method, comprising: Using one or more imaging devices to generate a first image of a plastic shell, wherein the plastic shell is customized for a patient's dental arch; Analyzing the first image to determine whether the plastic shell may include a defect; Responsive to determining that the plastic shell may include a defect, determining an inspection protocol for the plastic shell based at least in part on the defect; and Executing the inspection protocol to capture one or more additional images of the plastic shell.

39. A computer-readable medium comprising instructions that, when executed by a processing device, cause a manufacturing system to perform the method of claim 38.

40. A system for performing the method of claim 38, the system comprising: A platform for supporting the plastic shell during imaging; One or more cameras for capturing the first image and the one or more additional images; ​ A lighting system for illuminating a plastic housing during capture of the first image and the one or more additional images; and A processing device configured to perform one or more operations of the method recited in claim 38.

Images