Generating dental renders from model data

Abstract

Methods and systems for generating a panoramic two-dimensional (2D) image of a dental part using a three-dimensional (3D) model of the dental part are described. In one example, a method includes receiving a 3D model of a dental site generated at least in part from an intraoral scan; generating a surface shaped to substantially surround a dental arch represented by the dental site; calculating a surface projection by projecting the dental site onto the surface; and generating a panoramic 2D image according to the surface projection.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of dentistry, and in particular to generating two-dimensional (2D) dental arch renderings using three-dimensional (3D) models from intraoral scans. Background Art

[0002] Ionizing radiation has historically been used to image teeth, with X-ray bitewing radiographs being a common technique for providing non-quantitative images of a patient's dentition. However, in addition to the risks of ionizing radiation, such images are often limited in their ability to depict specific features and can involve time-consuming and expensive procedures. Other techniques, such as cone-beam computed tomography (CBCT), can provide tomographic images but still require ionizing radiation.

[0003] Specialized 3D scanning tools have also been used to image teeth. Scans from 3D scanning tools provide topographic data of the patient's dentition, which can be used to generate a 3D dental mesh model of the patient's teeth. For restorative dental work such as crowns and bridges, an intraoral scanner can be used to generate one or more intraoral scans of the prepared teeth and / or surrounding teeth on the patient's dental arch. Surface representations of 3D surfaces of teeth have proven extremely useful in designing and making dental prostheses (e.g., crowns or bridges) and treatment planning.

[0004] Two-dimensional (2D) renderings can be easily generated from such a 3D model. Conventional rendering methods typically focus on a localized portion of the patient's jaw, but fail to provide a comprehensive picture of the entire dental arch. Therefore, at least seven images (i.e., right buccal view, right lingual view, anterior buccal view, anterior lingual view, left buccal view, left lingual view, and occlusal view) are typically required to have a more complete picture of the jaw. While techniques exist for stitching multiple partial dental arch renderings into a single image, these methods often exhibit unintended / undesirable distortions. Therefore, there is a need for methods that can minimize or reduce geometric distortions during the rendering process. Summary of the Invention

[0005] A number of example implementations have been summarized. Many other implementations are also contemplated.

[0006] In a first embodiment, a method includes receiving a three-dimensional (3D) model of a dental site generated from one or more intraoral scans; generating a projection target shaped to substantially surround a dental arch represented by the dental site; calculating a surface projection by projecting the 3D model of the dental site onto one or more surfaces of the projection target; and generating at least one panoramic 2D image of the dental site based on the surface projections.

[0007] In a second embodiment, a method includes: receiving a three-dimensional (3D) model of a dental part generated based on one or more intraoral scans; generating a plurality of vertices along a dental arch represented by the dental part; calculating a projection target, the projection target including a plurality of surface segments connected in series to each other at the locations of the vertices; scaling the projection target relative to a dental arch center located within a central region of the dental arch so that the projection target substantially surrounds the dental arch; calculating a surface projection by projecting the 3D model of the dental part onto each of the surface segments of the projection target; and generating at least one panoramic 2D image of the dental part based on the surface projections.

[0008] In a third embodiment, a method includes receiving a 3D model of a dental site generated based on one or more intraoral scans; generating a projection target shaped to substantially surround a dental arch represented by the dental site; calculating a first surface projection by projecting the 3D model of the dental site onto one or more surfaces of the projection target along a buccal direction; calculating a second surface projection by projecting the 3D model of the dental site onto one or more surfaces of the projection target along a lingual direction; and generating at least one panoramic 2D image by combining the first surface projection and the second surface projection.

[0009] In a fourth embodiment, a non-transitory computer-readable medium includes instructions that, when executed by a processing device, cause the processing device to perform the method of any preceding embodiment.

[0010] In a fifth embodiment, an intraoral scanning system includes an intraoral scanner and a computing device operatively connected to the intraoral scanner, wherein the computing device is configured to perform the method of any preceding embodiment.

[0011] In a sixth embodiment, a system includes a memory and a processing device, wherein the processing device is configured to execute instructions from the memory to perform the method of any preceding embodiment. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In the accompanying drawings, embodiments of the disclosure are illustrated by way of example and not limitation.

[0013] Figure 1 An exemplary system for performing intraoral scanning and / or generating panoramic 2D images of a dental site in accordance with at least one embodiment is shown.

[0014] Figure 2 A cylindrical modeling method for generating a 2D projection of a 3D dentition is shown in accordance with at least one embodiment.

[0015] Figure 3 Projecting 3D dentition onto a cylindrical projection surface is shown in accordance with at least one embodiment.

[0016] Figure 4 FIG2 is a flowchart illustrating a workflow for generating an X-ray panoramic simulation image according to at least one embodiment.

[0017] Figure 5 is a comparison of an actual X-ray image and an X-ray panoramic simulation image according to at least one embodiment.

[0018] Figure 6A A dental arch curve following modeling method for generating a 2D projection of a 3D dentition is shown in accordance with at least one embodiment.

[0019] Figure 6B The present invention illustrates a workflow for generating a panoramic projection from a 3D dentition based on a dental arch curve following modeling method according to at least one embodiment.

[0020] Figure 7 A graphical user interface is shown displaying various renderings of 3D dentition in accordance with at least one embodiment.

[0021] Figure 8A Another dental arch curve following modeling method for generating a 2D projection of a 3D dentition is shown in accordance with at least one embodiment.

[0022] Figure 8B Shown is a 2D buccal rendering of a 3D dentition using an arch curve following modeling approach, according to at least one embodiment.

[0023] Figure 8C Shown is a 2D lingual rendering of a 3D dentition using an arch curve following modeling approach, according to at least one embodiment.

[0024] Figure 9A A polynomial curve modeling method for generating 2D projections of 3D dentition is shown in accordance with at least one embodiment.

[0025] Figure 9B Shown is a 2D panoramic rendering overlaid onto an X-ray image in accordance with at least one embodiment.

[0026] Figure 10 A flow chart of a method for generating a panoramic 2D image according to at least one embodiment is shown.

[0027] Figure 11 A flow chart of a method for generating a panoramic 2D image based on a multi-surface projection target according to at least one embodiment is shown.

[0028] Figure 12A A flow chart of a method for generating an X-ray panoramic simulation image according to at least one embodiment is shown.

[0029] Figure 12BA flow chart illustrating a method of generating projected segmentation / classification information from a panoramic 2D image onto a 3D model of a dental site in accordance with at least one embodiment.

[0030] Figure 13 A block diagram of an example computing device is shown, in accordance with an embodiment of the present disclosure. DETAILED DESCRIPTION

[0031] Described herein are methods and systems for generating panoramic 2D images of a patient's dental part (e.g., the dentition) using a 3D model of the dental part. For example, the 2D image can be used to examine and assess the shape, position, and orientation of teeth, as well as to identify and label dental features. Examples of dental features that can be identified and / or labeled include cracks, chips, gum lines, worn areas of teeth, cavities (also known as caries), eruption contours (e.g., the gum-to-tooth interface), implant gum lines, implant margins, scan volume margins / curves, margin lines of prepared teeth, and the like. Also described herein are methods and systems for rendering simulated X-ray images (referred to herein as "X-ray panoramic simulated images") based on panoramic renderings of the 3D model. Also described herein are methods and systems for labeling dental features in a panoramic 2D image and assigning labels to corresponding dental features in the 3D model from which the panoramic 2D image is derived. Certain embodiments described herein parameterize the rendering process by projecting the 3D model onto various types of projection targets to reduce or minimize geometric distortion. Certain embodiments also relate to projection targets that closely track the contours of the patient's dental arch. Such an embodiment may provide more accurate panoramic rendering with minimal distortion, thereby further facilitating visual oral diagnosis and patient education for dentists.

[0032] The embodiments described herein provide a framework for panoramic dental arch renderings (both buccal and lingual views). When combined with occlusal views of the jaws, dental personnel can have a comprehensive overview of the patient's jaws, thereby facilitating diagnosis and patient education. Unlike traditional rendering methods that typically require at least seven images (i.e., right buccal view, right lingual view, anterior buccal view, anterior lingual view, left buccal view, left lingual view, and occlusal view), the embodiments described herein can reduce the number of renderings used to fully visualize the patient's dentition to three, i.e., buccal panoramic rendering, lingual panoramic rendering, and occlusal rendering. In addition, panoramic dental arch rendering provides easier image labeling for various image-based oral diagnostic modeling processes. Such labeling in 2D can be projected back onto the original 3D model for machine learning purposes, visualization, etc. Panoramic dental arch rendering also provides a method for simulating panoramic radiographs, which can potentially reduce or eliminate the need to take actual panoramic radiographs during or after a patient's orthodontic treatment. For example, using the patient's initial panoramic X-ray image and the original 3D scan, the radiograph simulation process can be calibrated to render new radiograph-like images even after the patient's teeth have moved due to treatment. This can potentially reduce / eliminate the need for follow-up radiographs during or after the patient's orthodontic treatment, especially when combined with 3D root reconstruction.

[0033] Advantages of embodiments of the present disclosure include, but are not limited to: (1) providing a method for rendering panoramic images of a dental arch directly from a 3D scan of a patient's dentition to provide a comprehensive picture of the patient's jaw that facilitates easier oral diagnosis and patient education; (2) facilitating the labeling of various dental features from a panoramic rendering and implementing various image-based machine learning methods; (3) simulating panoramic X-ray images to potentially reduce or eliminate follow-up X-rays during or after a patient's orthodontic treatment; and (4) utilizing parametric methods to allow easy control of various aspects of the final rendering (e.g., the amount of posterior molar angulation in a panoramic rendering).

[0034] Various embodiments are described herein. It should be understood that these different embodiments can be implemented as stand-alone solutions and / or can be combined. Therefore, references to an embodiment or an embodiment can refer to the same embodiment and / or different embodiments. Some embodiments are discussed herein with reference to intraoral scans and intraoral images. However, it should be understood that the embodiments described with reference to intraoral scans also apply to laboratory scans or model / impression scans. A laboratory scan or model / impression scan can include one or more images of a dental part or a model or impression of a dental part, which images may or may not include a height map and may or may not include a color image.

[0035] Figure 1An exemplary system 100 for performing intraoral scanning and / or generating panoramic 2D images of dental sites according to at least one embodiment is shown. In at least one embodiment, one or more components of the system 100 perform the following with reference to Figure 10 to Figure 1 2 describes one or more operations.

[0036] System 100 includes a dental office 108 and a dental lab 110. Dental office 108 and dental lab 110 each include computing devices 105, 106, where computing devices 105, 106 can be connected to each other via a network 180. Network 180 can be a local area network (LAN), a public wide area network (WAN) (e.g., the Internet), a private WAN (e.g., an intranet), or a combination thereof.

[0037] The computing device 105 can be coupled to an intraoral scanner 150 (also referred to as a scanner) and / or a data storage device 125. The computing device 106 can also be connected to a data storage device (not shown). The data storage device can be a local data storage device and / or a remote data storage device. The computing device 105 and the computing device 106 can each include one or more processing devices, memory, secondary storage, one or more input devices (e.g., a keyboard, a mouse, a tablet computer, etc.), one or more output devices (e.g., a display, a printer, etc.), and / or other hardware components.

[0038] In at least one embodiment, the scanner 150 is wirelessly connected to the computing device 105 via a direct wireless connection. In at least one embodiment, the scanner 150 is wirelessly connected to the computing device 105 via a wireless network. In at least one embodiment, the wireless network is a Wi-Fi network. In at least one embodiment, the wireless network is a Bluetooth network, a Zigbee network, or some other wireless network. In at least one embodiment, the wireless network is a wireless mesh network, examples of which include a Wi-Fi mesh network, a Zigbee mesh network, and the like. In an example, the computing device 105 can be physically connected to one or more wireless access points and / or wireless routers (e.g., a Wi-Fi access point / router). The intraoral scanner 150 can include a wireless module, such as a Wi-Fi module, and via the wireless module can join a wireless network via a wireless access point / router.

[0039] In an embodiment, the scanner 150 includes an inertial measurement unit (IMU). The IMU may include an accelerometer, a gyroscope, a magnetometer, a pressure sensor, and / or other sensors. For example, the scanner 150 may include one or more microelectromechanical systems (MEMS) IMUs. The IMU may generate inertial measurement data (referred to herein as movement data or motion data), including acceleration data, rotation data, and the like.

[0040] The intraoral scanner 150 may include a probe (e.g., a handheld probe) for optically capturing three-dimensional structures. The intraoral scanner 150 may be used to perform an intraoral scan of a patient's oral cavity, generating a plurality of intraoral scans (also referred to as intraoral images). An intraoral scanning application 115 running on the computing device 105 may communicate with the scanner 150 to implement the intraoral scanning process. The result of the intraoral scan may be intraoral scan data 135A, 135B to 135N, which may include one or more sets of intraoral scans or intraoral images. Each intraoral scan or image may include a two-dimensional (2D) image and / or may include a 3D point cloud, the 2D image including depth information (e.g., a height map through a portion of a dental site). In either case, each intraoral scan includes x, y, and z information. Some intraoral scans (e.g., intraoral scans generated by a confocal scanner) include a 2D height map. In at least one embodiment, the intraoral scanner 150 generates a plurality of discrete (i.e., separate) intraoral scans. The discrete intraoral scan sets can be combined into smaller hybrid intraoral scan sets, where each hybrid intraoral scan is a combination of multiple discrete intraoral scans. Optionally, the intraoral scan data 135A-N may include one or more color images (e.g., color 2D images) and / or images generated under specific lighting conditions (e.g., 2D ultraviolet (UV) images, 2D infrared (IR) images, 2D near-IR images, 2D fluorescence images, etc.).

[0041] The scanner 150 may transmit the intraoral scan data 135A, 135B through 135N to the computing device 105. The computing device 105 may store the intraoral scan data 135A-135N in the data store 125.

[0042] According to an example, a user (e.g., a practitioner) can perform an intraoral scan of a patient. In doing so, the user can apply the scanner 150 to one or more locations within the patient's mouth. The scan can be divided into one or more segments. As an example, the segments can include an upper dental arch segment, a lower dental arch segment, an occlusal segment, and optionally one or more prepared tooth segments. As another example, the segments can include the patient's lower cheek area, the patient's lower tongue area, the patient's upper cheek area, the patient's upper tongue area, one or more prepared teeth of the patient (e.g., teeth of the patient to which a dental device such as a crown or other dental restoration will be applied), one or more teeth in contact with the prepared teeth (e.g., teeth that are not themselves subjected to a dental device but are located near one or more such teeth or engage with one or more such teeth when the mouth is closed), and / or the patient's occlusion (e.g., a scan performed when the patient's mouth is closed, scanning toward the joining area of ​​the patient's upper and lower teeth). Via such scanner application, the scanner 150 can provide intraoral scan data 135A-135N to the computing device 105. The intraoral scan data 135A-135N can be provided in the form of intraoral scan datasets, each of which can include a 3D point cloud, a 2D scan / image, and / or a 3D scan / image of a region of a specific tooth and / or intraoral location. In at least one embodiment, separate datasets are created for the maxillary arch, the mandibular arch, the patient's bite, and each prepared tooth. Alternatively, a single large dataset (e.g., for the mandibular arch and / or the maxillary arch) can be generated. Such scans can be provided to the computing device 105 from the scanner 150 in the form of one or more points (e.g., one or more point clouds).

[0043] The manner in which the patient's mouth is scanned may depend on the procedure being performed. For example, if an upper or lower denture is being fabricated, the entire mandibular or maxillary edentulous dental arch may be scanned. In contrast, if a dental bridge is being fabricated, only a portion of the entire dental arch may be scanned, including the edentulous area, adjacent prepared teeth (e.g., adjacent teeth), and the opposing dental arch and dentition. Furthermore, the manner in which the mouth is scanned may depend on the physician's scanning preferences and / or the patient's condition.

[0044] As a non-limiting example, dental procedures can be broadly divided into restorative (restoration) and orthodontic procedures, which are then further subdivided into specific forms of these procedures. In addition, dental procedures can include the identification and treatment of gum disease, sleep apnea, and intraoral conditions. The term restorative procedure refers in particular to any process involving the oral cavity and for the design, manufacture, or installation of a dental restoration at a dental site (intraoral location) within the oral cavity, or a real or virtual model thereof, or for the design and preparation of an intraoral location to receive such a restoration. A restoration may include any restoration, such as a crown, veneer, inlay, onlay, implant, and bridge, as well as any other artificial partial or complete denture. The term orthodontic procedure refers in particular to any process involving the oral cavity and for the design, manufacture, or installation of an orthodontic element at an intraoral location within the oral cavity, or a real or virtual model thereof, or for the design and preparation of an intraoral location to receive such an orthodontic element. These elements can be appliances including, but not limited to, brackets and wires, retainers, clear aligners, or functional appliances.

[0045] During an intraoral scan, the intraoral scanning application 115 can register and stitch together two or more intraoral scans (e.g., intraoral scan data 135A and intraoral scan data 135B) generated so far from the intraoral scanning session. In at least one embodiment, performing the registration includes capturing 3D data of various points of a surface in multiple scans and registering the scans by calculating a transformation between the scans. One or more 3D surfaces can be generated based on the intraoral scans that were registered and stitched together during the intraoral scan. The one or more 3D surfaces can be output to a display so that a physician or technician can view their scanning progress so far.

[0046] As each new intraoral scan is captured and registered with the previous intraoral scan and / or 3D surface, the one or more 3D surfaces can be updated, and the updated 3D surfaces can be output to a display. In an embodiment, the intraoral scan and / or 3D surface is segmented to segment points and / or patches on the intraoral scan and / or 3D surface into one or more categories. In at least one embodiment, the intraoral scanning application 115 classifies the points as hard tissue or soft tissue. The 3D surface can then be displayed using the classification information. For example, hard tissue can be displayed using a first visualization (e.g., an opaque visualization) and soft tissue can be displayed using a second visualization (e.g., a transparent or semi-transparent visualization).

[0047] In an embodiment, separate 3D surfaces are generated for the upper and lower jaws.This process can be performed in real time or near real time to provide an updated view of the captured 3D surface during the intraoral scanning process.

[0048] When a scanning session or a portion of a scanning session associated with a particular scanning role or segment (e.g., a maxillary role, a mandibular role, an occlusal role, etc.) is completed (e.g., all scans of the intraoral locations or dental parts have been captured), the intraoral scanning application 115 can automatically generate a virtual 3D model of one or more scanned dental parts (e.g., the maxillary and mandibular). The final 3D model can be a set of 3D points and their connections to each other (i.e., a mesh). In at least one embodiment, the final 3D model is a volumetric 3D model having both surface features and internal features. In an embodiment, the 3D model is a volumetric model generated as described in international patent application publication number WO2019 / 147984A1, entitled “Diagnostic Intraoral Scanning and Tracking,” the entire contents of which are incorporated herein by reference.

[0049] To generate the virtual 3D model, the intraoral scanning application 115 may register and stitch together the intraoral scans generated from the intraoral scanning session that are associated with a particular scan character or segment. The registration performed at this stage may be more accurate than the registration performed during the capture of the intraoral scans and may take more time to complete than the registration performed during the capture of the intraoral scans. In at least one embodiment, performing scan registration includes capturing 3D data for various points of the surface in multiple scans and registering the scans by calculating a transformation between the scans. The 3D data can be projected into the 3D space of the 3D model to form a portion of the 3D model. By applying the appropriate transformation to the points of each registered scan and projecting each scan into the 3D space, the intraoral scans can be integrated into a common reference frame.

[0050] In at least one embodiment, registration is performed on adjacent or overlapping intraoral scans (e.g., each consecutive frame of an intraoral video). In at least one embodiment, registration is performed using hybrid scans. A registration algorithm is performed to register two adjacent or overlapping intraoral scans (e.g., two adjacent hybrid intraoral scans) and / or to register an intraoral scan with a 3D model, which essentially involves determining a transformation that aligns one scan with the other scan and / or the 3D model. Registration can involve identifying multiple points in each scan (e.g., a point cloud) of a scan pair (or a scan and a 3D model), performing a surface fit on the points, and matching points of the two scans (or scan and 3D model) using a local search around the points. For example, the intraoral scanning application 115 can match points of one scan with the nearest point interpolated on the surface of the other scan and iteratively minimize the distance between the matched points. Other registration techniques can also be used.

[0051] The intraoral scanning application 115 can repeat the registration process for all intraoral scans in the sequence of intraoral scans to obtain a transformation for each intraoral scan to align each intraoral scan with the previous intraoral scan(s) and / or a common reference frame (e.g., with the 3D model). The intraoral scanning application 115 can integrate the intraoral scans into a single virtual 3D model by applying an appropriately determined transformation to each intraoral scan. Each transformation can include rotations about one to three axes and translations in one to three planes.

[0052] In many cases, data from one or more intraoral scans does not perfectly correspond to data from one or more other intraoral scans. Therefore, in an embodiment, the intraoral scanning application 115 can process the intraoral scans (e.g., which can be a mixed intraoral scan) to determine which intraoral scans (or which portions of the intraoral scans) to use for portions of the 3D model (e.g., for representing portions of a particular dental part). The intraoral scanning application 115 can use data such as geometric data represented in the scans and / or timestamps associated with the intraoral scans to select the best intraoral scan to use for depicting the dental part or portion of a dental part. In at least one embodiment, the images are input into a machine learning model that has been trained to select scans of dental parts and / or to grade scans of dental parts. In at least one embodiment, each scan is assigned one or more scores, where each score can be associated with a particular dental part and indicates the quality of the representation of that dental part in the intraoral scan.

[0053] Additionally or alternatively, weights can be assigned to the intraoral scans based on the scores assigned to them (e.g., based on temporal proximity to the timestamps of one or more selected 2D images). The assigned weights can be associated with different dental sites. In at least one embodiment, a weight can be assigned to each scan (e.g., each hybrid scan) of a dental site (or multiple dental sites). During model generation, conflicting data from multiple intraoral scans can be combined using a weighted average to depict the dental site. The weights applied can be those assigned based on the quality scores of the dental site. For example, processing logic can determine that the data quality of a particular overlapping region from a first set of intraoral scans is better than the data quality of a particular overlapping region from a second set of intraoral scans. Then, when averaging the differences between the intraoral scan datasets, the weight of the first intraoral scan dataset can be greater than the weight of the second intraoral scan dataset. For example, the first intraoral scan, which was assigned a higher rating, can be assigned a weight of 70%, and the second intraoral scan can be assigned a weight of 30%. Thus, when the data is averaged, the combined result will more closely resemble the depiction from the first intraoral scan dataset and less closely resemble the depiction from the second intraoral scan dataset.

[0054] In at least one embodiment, the images and / or intraoral scans are input into a machine learning model that has been trained to select images and / or intraoral scans of dental sites and / or to grade images and / or intraoral scans of dental sites. In at least one embodiment, one or more scores are assigned to each image and / or intraoral scan, wherein each score may be associated with a particular dental site and indicates the quality of the representation of that dental site in the 2D image and / or intraoral scan. Once a set of images are selected for use in generating a portion of a 3D model / surface representing a particular dental site (or a portion of a particular dental site), those images / scans and / or portions of those images / scans can be locked. The locked images or portions of locked images selected for a dental site can be used specifically for creating a particular area of ​​the 3D model (e.g., for creating an associated tooth in the 3D model).

[0055] The intraoral scanning application 115 can generate one or more 3D surfaces and / or 3D models based on the intraoral scan, and can display the 3D surfaces and / or 3D models to a user (e.g., a physician) via a user interface. The physician can then visually inspect the 3D surfaces and / or 3D models. The physician can virtually manipulate the 3D surfaces and / or 3D models (i.e., translation and / or rotation relative to one or more of three mutually orthogonal axes) via the user interface using appropriate user controls (hardware and / or virtual) to enable viewing of the 3D models from any desired direction. The physician can view (e.g., visually inspect) the generated 3D surfaces and / or 3D models of the intraoral location and determine whether the 3D surfaces and / or 3D models are acceptable.

[0056] Once a 3D model of a dental part (e.g., a dental arch or a portion of a dental arch including a prepared tooth) is generated, it can be sent to the dental modeling logic 116 for review, analysis, and / or updating. Additionally or alternatively, the intraoral scanning application 115 can perform one or more operations associated with reviewing, analyzing, and / or updating the 3D model.

[0057] The intraoral scanning application 115 and / or the dental modeling logic 116 may include modeling logic 118 and / or panoramic 2D image processing logic 119. The modeling logic 118 may include logic for generating projection targets onto which the 3D model can be projected. The modeling logic 118 may import 3D model data to identify various parameters for generating projection targets. These parameters include, but are not limited to, the dental arch center (which can be used as the projection center for performing the projection transformation), 3D coordinate axes, tooth positions / centers, and dental arch dimensions. Based on these parameters, the modeling logic 118 can determine the position, size, and orientation of various projection targets to be positioned around the dental arch represented by the 3D model.

[0058] In at least one embodiment, the panoramic 2D image processing logic 119 (or "image processing logic") can generate / derive a panoramic 2D image from a 3D model of the dental site using one or more models (i.e., projection targets) generated by the modeling logic 118. For example, the image processing logic 119 can generate a 2D panoramic image from the 3D model based on a projection center. For example, projecting radially outward onto the projection target can produce a panoramic lingual view of the dentition, while projecting radially inward onto the projection target can produce a panoramic buccal view of the dentition. The image processing logic 119 can also be used to generate an X-ray panoramic simulation image based on, for example, a lingual 2D panoramic projection and a buccal 2D panoramic projection. In at least one embodiment, the result of such a projective transformation can include not only the original image data, but also retain other information related to the 3D model. For example, each pixel of the 2D panoramic image can have associated depth information (e.g., radial distance from the projection center), density information, 3D surface coordinates, and / or other data. In at least one embodiment, this data can be used to transform the 2D panoramic image back into a 3D image. In at least one embodiment, these data can be used in identifying tooth overlaps detectable from buccal and lingual projections.

[0059] In at least one embodiment, the visualization component 120 of the intraoral scanning application 115 can be used to visualize the panoramic 2D image for review, labeling, patient education, or any other purpose. In at least one embodiment, the visualization component 120 can be used to compare panoramic 2D images generated from the intraoral scan at various stages of treatment planning. Such embodiments allow for visualization of tooth movement and displacement. In at least one embodiment, a machine learning model can be trained to detect and automatically label tooth movement and displacement using panoramic 2D images, panoramic X-ray images, and / or intraoral scan data as input.

[0060] Now about Figure 2 to Figure 1 2 describes the functionality of the dental modeling logic 116 in more detail. Figure 2 , Figure 2A cylindrical modeling method for generating a 2D projection of a dental part (e.g., a 3D model of a patient's dentition, or "3D dentition" 210) is shown in accordance with at least one embodiment. A top-down view of the 3D dentition 210 is shown, wherein a projection center 230 is located in a central region of a dental arch of the 3D dentition 210. A projection target (i.e., projection surface 220) is placed around the dental arch of the 3D dentition 210. As shown, the projection surface 220 is a partial cylinder surrounding the dental arch. In at least one embodiment, the radius of the projection surface 220 can coincide with the projection center 230. In at least one embodiment, the radius can be selected to surround the dental arch while maintaining a minimum spacing away from the nearest teeth of the 3D dentition 210. In at least one embodiment, the projection center 230 corresponds to the center of the dental arch. In at least one embodiment, the projection center 230 is selected such that for a given radius of the projection surface 220, a radial projection line 232 is tangential or approximately tangential to the third molar of the 3D dentition 210.

[0061] Figure 3 2 . The projection of the 3D dentition 210 onto the cylindrical projection surface 220 is shown in accordance with at least one embodiment. In at least one embodiment, the 3D dentition 210 is projected onto the cylindrical projection surface 220 and then the 3D model / mesh is flattened to produce a flattened dental arch mesh 330. The flattened dental arch mesh 330 can then be rendered using orthographic rendering to generate a panoramic projection. In at least one embodiment, the coordinate system (xyz) is based on the original coordinate system associated with the 3D dentition 210, and the coordinate system of the projection surface 220 is defined as x'-y'-z'. In at least one embodiment, any coordinate of the 3D dentition 210 is transformed to x'-y'-z' using a transform 320 according to the following relationship:

[0062]

[0063] Where r is the distance from the origin of the 3D dentition 210 coordinate system to the projection surface 220. Through the above transformation, a "flattened" dental arch mesh 330 is obtained.

[0064] Figure 4is a workflow 400 illustrating the generation of an X-ray panoramic simulation image 450 (based on a circular projection) according to at least one embodiment. In at least one embodiment, orthographic rendering is applied via a transformation 420, thereby producing panoramic 2D images 430A and / or 430B. In at least one embodiment, since the buccal side of the flattened dental arch mesh 330 faces the projection surface 220, applying the transformation produces a buccal image (panoramic 2D image 430A). In at least one embodiment, a lingual rendering (panoramic 2D image 430B) can be obtained, for example, by rotating the flattened dental arch mesh 330 180° around a vertical axis or by flipping the sign of the depth coordinate (z'). In at least one embodiment, the panoramic 2D images 430A and 430B can retain the original color of the 3D dentition 210. In at least one embodiment, the panoramic 2D images 430A and 430B can be recolored. For example, as Figure 4 As shown, each tooth is recolored in grayscale using, for example, a grayscale pixel value that is the tooth index number multiplied by 5. In at least one embodiment, transformation 440 is applied to panoramic 2D images 430A and 430B to generate an X-ray panoramic simulation image 450, which can be generated by comparing a buccal rendering and a lingual rendering of the same jaw and marking areas with different color values ​​from each other with different colors (e.g., white) to illustrate tooth overlap representing high-density areas of the X-ray image.

[0065] Figure 5 is a comparison of an actual X-ray image 500 and an X-ray panoramic simulation image 450 for the same patient according to at least one embodiment. As shown, the simulated rendering of the X-ray panoramic simulation image 450, including the marked / highlighted areas, is very similar to the original X-ray image 500, including the identification of the high-density areas. In at least one embodiment, the simulation process can be calibrated to more closely resemble the X-ray image, for example by adjusting the position of the projection center and the position and orientation of the projection surface 220. For example, such calibration is beneficial if the patient's jaw was not facing / orthogonal to the X-ray film when the X-ray was captured. In at least one embodiment, these parameters can be iterated and multiple X-ray panoramic simulation images can be generated in order to identify the best fitting simulation image.

[0066] Figure 6AA dental arch curve following modeling method for generating a 2D projection of a 3D dentition 210 according to at least one embodiment is shown. In at least one embodiment, a plurality of projection surfaces 620 (e.g., a plurality of connected or continuous projection surfaces 620) are used as projection targets. In at least one embodiment, the dental arch of the 3D dentition is segmented into a plurality of dental arch segments 642 (e.g., 7 segments as shown) based on an angular span around the center of the dental arch. For each of the dental arch segments 642, a center vertex 644 is calculated. The center vertex 644 is used to connect the segments 642 in a segment-by-segment manner to produce a dental arch mesh 640. Once the dental arch mesh 640 is generated, the dental arch mesh 640 can be scaled radially outward from the projection center 630 to form a projection surface 620 that contains the dental arch. In one embodiment, a smoothing algorithm is applied to the projection surface 620 to produce a smoother transition between the segments 622 by eliminating / reducing discontinuities caused by the presence of joints.

[0067] Figure 6B FIG6 is a workflow 650 illustrating the generation of a panoramic projection 695 from a 3D dentition 660 based on a dental arch curve following method in accordance with at least one embodiment. In at least one embodiment, the projection surface 670 can be generated based on (1) a polynomial fitting process or (2) a process similar to the projection surface 620 using a smoothing algorithm. Based on the projection surface 670, a transformation 680 is applied to the 3D dentition to generate a flattened dental arch mesh 690. Then, orthographic rendering is applied to the flattened dental arch mesh 690 to generate a final panoramic rendering (panoramic projection 695) of the 3D dentition 660. As described above with respect to Figure 4 As discussed, to switch from a buccal view to a lingual view, the sign of the depth coordinate in the flattened mesh can be switched, the vertex order of all triangular mesh faces can be reversed, or the flattened dental arch mesh 690 can be rotated around its vertical axis before applying orthographic rendering.

[0068] Figure 7 A graphical user interface (GUI) 700 is shown that displays various renderings of a 3D dentition according to at least one embodiment. The dental professional can utilize the GUI 700 to display a panoramic 2D image of a patient's dentition. As shown, the GUI 700 includes a buccal rendering 710, a lingual rendering 720, and an occlusal rendering 730. For example, the occlusal rendering 730 can be generated by projecting from the top of the 3D dentition downward onto a plane below the 3D dentition. In other embodiments, the maxillary dentition can be shown alone, or the GUI 700 can include renderings of the upper and lower dentitions (e.g., up to 6 views total). In at least one embodiment, the GUI 700 can allow the dental professional to mark dental features that are observable in the various renderings. In at least one embodiment, marking a dental feature in one rendering can cause a similar mark to appear in a corresponding location in another rendering.

[0069] Figure 8A A dental arch curve following modeling method for generating a 2D projection of a 3D dentition 810 using a hybrid projection surface according to at least one embodiment is shown. In at least one embodiment, the projection surface 820 is generated based on three segments joined at its edges, each portion corresponding to a projection sub-surface. In other embodiments, more than three segments are utilized. As shown, the projection surface 820 is formed by a planar portion 822 and a cylindrical portion 826 connected to the planar portion 822 at its edge, thereby producing a symmetrical shape that substantially surrounds the 3D dentition 810. A smooth transition between the planar portion 822 and the cylindrical portion 826 can be utilized to reduce or eliminate discontinuities. In at least one embodiment, the cylindrical portion 826 includes a first portion of the 3D dentition 810, and the planar portion 822 extends beyond the first portion and toward the posterior portion of the 3D dentition (left and right posterior molars). The angle θ corresponds to the angle between two projection lines 824 extending from the projection center 830 to the edge where the planar portion 822 and the cylindrical portion 826 meet. In at least one embodiment, angle θ is between about 110° and about 130° (e.g., 120°). In at least one embodiment, angle θ, the location of projection center 830, and the orientation and location of projection surface 820 are used as adjustable parameters to optimize / minimize distortion in the resulting panoramic image. Figure 8B and Figure 8C 2D buccal and 2D lingual renderings of a 3D dentition 810 and a maxillary 3D dentition using a hybrid surface modeling approach, respectively, are shown. In at least one embodiment, the renderings can be presented in a GUI for inspection, evaluation, and marking (as described above with respect to Figure 7 described).

[0070] Various other embodiments involve different geometries of the projection target.A more general dental arch curve following method is now described to further minimize potential distortions in the rendering process. Figure 9A A polynomial dental arch curve modeling method for generating a 2D projection of the 3D dentition 210 is shown in accordance with at least one embodiment. As shown, a parabolic projection surface 920 surrounds the dental arch of the 3D dentition 210. The parabolic projection surface 920 is an illustrative example of a polynomial curve modeling method, and it is contemplated that higher-order polynomials may be used, as will be understood by one of ordinary skill in the art. Such an embodiment may be preferable to, for example, a cylindrical projection target (e.g., projection surface 220) because the parabolic surface can more closely follow the patient's dental arch. Figure 9B A 2D panoramic rendering is shown superimposed onto X-ray images of the patient's upper and lower dental arches, demonstrating the accuracy of the polynomial arch curve modeling method.

[0071] In practice, while all panoramic X-ray imaging systems closely follow the curves of the patient's dental arches during the imaging process, the actual projection trajectory can vary between panoramic X-ray imaging systems, depending on the underlying design and manufacturing of the imaging system. Furthermore, the patient's positioning relative to the panoramic X-ray imaging system can also affect the resulting X-ray image. To improve the simulation of X-ray images, certain embodiments parameterize both the projection trajectory and relative jaw positioning to more accurately simulate panoramic images.

[0072] Figure 10 to Figure 1 2 shows methods related to generating a panoramic 2D image based on a 3D model of a dental site, the 3D model of the dental site being generated based on one or more intraoral scans. These methods may be performed by processing logic that may include hardware (e.g., circuitry, dedicated logic, programmable logic, microcode, etc.), software (e.g., instructions running on a processing device to perform hardware simulation), or a combination thereof. In at least one embodiment, at least some of the operations of these methods are performed by executing a dental modeling application (e.g., Figure 1 The dental modeling application 116 may be executed on a computing device (e.g., a dental modeling logic 116) configured to perform the dental modeling. The dental modeling application 116 may be, for example, a component of an intraoral scanning device comprising a handheld intraoral scanner and a computing device operatively coupled (e.g., via a wired or wireless connection) to the handheld intraoral scanner. Alternatively or additionally, the dental modeling application may be executed on a computing device in a dentist's office or dental laboratory.

[0073] For simplicity of explanation, these methods are depicted and described as a series of actions. However, the actions according to the present disclosure can occur in various orders and / or simultaneously, and other actions are not presented and described herein. In addition, not all of the actions shown are necessary for implementing the methods according to the disclosed subject matter. In addition, those skilled in the art will understand and appreciate that, alternatively, these methods can also be represented as a series of interrelated states via state diagrams or events.

[0074] Now refer to Figure 10, which shows a flow chart of a method 1000 for generating a panoramic 2D image according to at least one embodiment. At box 1005, a computing device (e.g., computing device 105 of dental clinic 108 or dental laboratory 110) receives a 3D model of a dental site. In at least one embodiment, the 3D model is generated based on one or more intraoral scans. For example, the intraoral scans can be performed by a scanner (e.g., scanner 150) that generates one or more intraoral scan datasets. In at least one embodiment, the intraoral scan datasets can include 3D point clouds, 2D images, and / or 3D images of specific teeth and / or regions of the dental site. The intraoral scan datasets can be processed (e.g., via an intraoral scanning application 115 implementing dental modeling logic 116) to generate a 3D model of a dental site, such as a 3D dentition of a patient's mandible, maxilla, or both (e.g., any of 3D dentition 210, 660, or 810).

[0075] At block 1010, a computing device (e.g., implementing modeling logic 118) generates a projection target shaped to substantially surround a dental arch represented by a dental site. In at least one embodiment, the projection target is a cylindrical surface (e.g., projection surface 220) that substantially surrounds the dental arch. In at least one embodiment, the projection target comprises a curve in the shape of a polynomial curve, such as a parabolic surface (e.g., projection surface 620) that substantially surrounds the dental arch.

[0076] In at least one embodiment, the projection target is a hybrid surface (e.g., projection surface 920) formed by a cylindrical surface (e.g., cylindrical portion 926), a first planar surface extending from an edge of the cylindrical surface, and a second planar surface (e.g., planar portion 922). In at least one embodiment, the cylindrical surface, the first planar surface, and the second planar surface collectively define a continuous surface that substantially surrounds the dental arch. In at least one embodiment, the angle between the first planar surface and the second planar surface is about 110° to about 130° (e.g., 120°).

[0077] In module 1015, a computing device (e.g., implementing panoramic 2D image processing logic 119) calculates surface projections by projecting the 3D model of the dental site onto one or more surfaces of a projection target. In at least one embodiment, the surface projections are calculated based on a projection path around the dental arch.

[0078] At block 1020, a computing device (e.g., implementing the panoramic 2D image processing logic 119) generates at least one panoramic two-dimensional (2D) image based on the surface projection. In at least one embodiment, the at least one panoramic 2D image is generated by orthographic rendering of a flattened mesh generated by projecting the 3D model along a projection path around the dental arch (e.g., applying either of the transformations 420 or 780).

[0079] Figure 11 FIG. 1 is a flow chart showing a method 1100 for generating a panoramic 2D image based on a multi-surface projection target according to at least one embodiment. For example, the method 1100 may follow the method of FIG. Figure 7 A and Figure 7 B. At block 1105, a computing device (e.g., computing device 105 of dental clinic 108 or dental laboratory 110) receives a 3D model of a dental site. In at least one embodiment, the 3D model of the dental site may include 3D dentition (e.g., any of 3D dentition 210, 660, or 810) of the patient's mandible, maxilla, or both.

[0080] At block 1110, a plurality of vertices (e.g., center vertex 644) are calculated along a dental arch represented by a dental part (e.g., 3D dentition 210). In at least one embodiment, one or more of the plurality of vertices are located at the center of a tooth. In at least one embodiment, the number of vertices is greater than 5 (e.g., 10, 50, or more).

[0081] At block 1115, an initial projected target (eg, dental arch mesh 640) is calculated. In at least one embodiment, the initial projected target is formed by a plurality of surface segments (eg, segment 742) that are connected in series to one another at the locations of vertices.

[0082] At block 1120 , a projection target (eg, projection surface 620 ) is generated by scaling the initial projection target relative to the center of the dental arch within the center region of the dental arch so that the projection target substantially surrounds the dental arch. The resulting projection target includes a plurality of segments (eg, segment 622 ).

[0083] At block 1125, a surface projection is calculated by projecting the 3D model of the dental site onto each of the surface segments of the projection target. In at least one embodiment, a smoothing algorithm is applied to the projection target to reduce potential discontinuities in the final rendering, thereby improving rendering quality. At block 1130, at least one panoramic two-dimensional (2D) image is generated based on the surface projections.

[0084] Figure 12AA flow chart of a method 1200 for generating an X-ray panoramic simulation image according to at least one embodiment is shown. For example, the method 1200 may follow the steps of Figure 4 The workflow 400 is described. At block 1205, a computing device (e.g., computing device 105 of dental clinic 108 or dental laboratory 110) receives a 3D model of a dental site. In at least one embodiment, the 3D model of the dental site may include 3D dentition (e.g., any of 3D dentition 210, 660, or 810) of the patient's mandible, maxilla, or both.

[0085] At block 1210 , a projection target is generated. The projection target can be shaped to substantially surround the dental arch represented by the dental site. The projection target can correspond to any of the projection targets described above with respect to methods 1000 and 1100 .

[0086] At block 1215, a first surface projection is calculated by projecting the 3D model of the dental site onto one or more surfaces of the projection target along the buccal direction (e.g., based on the transformation 420). Figure 4 The mathematical operations described transform the coordinates of the 3D dentition to calculate the projection.

[0087] At block 1220, a second surface projection is calculated by projecting the 3D model of the dental site along a lingual direction onto one or more surfaces of the projection target (e.g., based on transformation 420). In at least one embodiment, the projection along the lingual direction can be performed by flipping the sign of the depth coordinate before or after applying the second surface projection, reversing the order of vertices of all mesh faces of the 3D model, or rotating the second surface projection about a vertical axis of the second surface projection.

[0088] At block 1225, at least one panoramic 2D image is generated by combining the first surface projection and the second surface projection (e.g., applying transformation 440). In at least one embodiment, the resulting panoramic 2D image corresponds to an X-ray panoramic simulation image (e.g., X-ray panoramic simulation image 450). In at least one embodiment, generating the panoramic 2D image includes marking regions of the panoramic 2D image that correspond to overlapping regions of the 3D model identified from the first surface projection and the second surface projection.

[0089] The following embodiments relate to any of methods 1000, 1100, or 1200. In at least one embodiment, the dental site corresponds to a single jaw. In such an embodiment, the first panoramic 2D image may correspond to a buccal rendering, and the second panoramic 2D image may correspond to a lingual rendering. For example, the buccal and lingual renderings of the jaws may be displayed separately in the GUI, or together with an occlusal rendering of the dental site, or together with a similar rendering of the contralateral jaw. In at least one embodiment, the occlusal rendering is generated by projecting a 3D model of the dental site onto a flat surface from the occlusal side of the dental site.

[0090] In at least one embodiment, a computing device may generate a panoramic 2D image for display and label one or more dental features in the image. Each labeled dental feature has an associated location within the panoramic 2D image. In at least one embodiment, the computing device determines a corresponding location in a 3D model from which the panoramic 2D image was generated and assigns a label for the dental feature to the corresponding location. In at least one embodiment, the 3D model includes the one or more labels when displayed. In at least one embodiment, labeling may be performed, for example, in response to user input to directly label the dental feature.

[0091] In at least one embodiment, a trained machine learning model may be used for labeling. For example, a trained machine learning model may be trained to identify and label dental features in a panoramic 2D image, a 3D dentition, or both. In at least one embodiment, model training according to embodiments of the present disclosure may be implemented using one or more workflows. In various embodiments, the model training workflow may be executed at a server, which may or may not include an intraoral scanning application. The model training workflow and the model application workflow may be executed by processing logic executed by a processor of a computing device. For example, one or more of these workflows may be implemented by one or more machine learning modules implemented in the intraoral scanning application 115, by the dental modeling logic 116, or in a Figure 13 1300 as shown and described in detail.

[0092] The model training workflow is used to train one or more machine learning models (e.g., deep learning models) to perform one or more classification, segmentation, detection, recognition, prediction, and other tasks on intraoral scan data (e.g., 3D intraoral scans, height maps, 2D color images, 2D NIRI images, 2D fluorescence images, etc.) and / or 3D surfaces generated based on the intraoral scan data. The model application workflow is used to apply one or more trained machine learning models to perform classification, segmentation, detection, recognition, prediction, and other tasks on intraoral scan data (e.g., 3D scans, height maps, 2D color images, NIRI images, etc.) and / or 3D surfaces generated based on the intraoral scan data. One or more of the machine learning models can receive and process 3D data (e.g., 3D point clouds, 3D surfaces, parts of 3D models, etc.). One or more of the machine learning models can receive and process 2D data (e.g., 2D panoramic images, height maps, projections of 3D surfaces onto planes, etc.).

[0093] Many different machine learning outputs are described herein. A specific number and arrangement of machine learning models are described and illustrated. However, it should be understood that the number and type of machine learning models used, as well as the arrangement of such machine learning models, can be modified to achieve the same or similar end results. Therefore, the described and illustrated arrangements of machine learning models are merely examples and should not be construed as limiting.

[0094] In various embodiments, one or more machine learning models are trained to perform one or more of the following tasks. Each task can be performed by a separate machine learning model. Alternatively, a single machine learning model can perform each task or a subset of the tasks. Additionally or alternatively, different machine learning models can be trained to perform different combinations of tasks. In one example, one or several machine learning models can be trained, where the trained ML model is a single shared neural network with multiple shared layers and multiple higher-level different output layers, where each output layer outputs a different prediction, classification, identification, etc. The tasks that the one or more trained machine learning models can be trained to perform are as follows:

[0095] I) Canonical Position Determination—This may include determining the canonical position and / or orientation of a 3D surface or object in an intraoral scan, or determining the canonical position of an object in a 2D image.

[0096] II) Scan / 2D Image Evaluation - This may include determining quality metric values ​​associated with regions of intraoral scans, 2D images, and / or 3D surfaces. This may include assigning quality values ​​to individual scans, 3D surfaces, portions of 3D surfaces, 3D models, portions of 3D models, 2D images, portions of 2D images, etc.

[0097] III) Moving tissue (excess tissue) identification / removal - this may include performing pixel-level identification of moving tissue (e.g., tongue, fingers, lips, etc.) from intraoral scans and / or 2D images

[0098] / classify, and optionally remove such mobile tissue from intraoral scans, 2D images, and / or 3D surfaces. Mobile tissue identification and removal is described in U.S. Patent Application Publication No. 2020 / 0349698A1, entitled “Excess Material Removal Using Machine Learning,” the entire contents of which are incorporated herein by reference.

[0099] IV) Dental Features in 2D or 3D Images - This can include performing point-level or pixel-level classification on the 3D model and / or 2D image to classify points / pixels as being part of a dental feature. This can include segmenting the 3D surface and / or 2D image. Points / pixels can be classified into two or more categories. A minimal classification scheme can include a dental feature category and a non-dental feature category. In other examples, additional dental categories can be identified, such as a hard tissue or tooth category, a soft tissue or gum category, and a margin category.

[0100] One type of machine learning model that can be used to perform some or all of the above tasks is an artificial neural network, such as a deep neural network. An artificial neural network typically includes a feature representation component with a classifier or regression layer that maps features to a desired output space. For example, a convolutional neural network (CNN) hosts multiple layers of convolution filters. Pooling is performed at the lower layers, and nonlinear problems can be solved. A multi-layer perceptron is typically attached above the lower layers to map the top features extracted by the convolution layer to a decision (e.g., classification output). Deep learning is a class of machine learning algorithms that uses a cascade of multiple layers of nonlinear processing units for feature extraction and transformation. Each successive layer uses the output from the previous layer as input. Deep neural networks can learn in a supervised (e.g., classification) and / or unsupervised (e.g., pattern analysis) manner. A deep neural network includes a hierarchical structure in which different layers learn different levels of representation corresponding to different levels of abstraction. In deep learning, each level learns to transform its input data into a slightly more abstract and complex representation. For example, in an image recognition application, the raw input might be a matrix of pixels; the first representation layer might abstract the pixels and encode edges; the second layer might construct and encode the arrangement of edges; the third layer might encode higher-level shapes (e.g., teeth, lips, gums, etc.); and the fourth layer might recognize the scanned character. Notably, the deep learning process can autonomously learn which features are best placed at which level. The "depth" in "deep learning" refers to the number of layers through which the data is transformed. More precisely, a deep learning system has a tangible credit assignment path (CAP) depth. The CAP is the chain of transformations from input to output. The CAP describes the underlying causal relationship between input and output. For feedforward neural networks, the CAP depth can be the depth of the network, which can be the number of hidden layers plus one. For recurrent neural networks, where a signal can propagate through a layer more than once, the CAP depth can be infinite.

[0101] In at least one embodiment, a graph neural network (GNN) architecture that operates on three-dimensional data is used. Unlike traditional neural networks that operate on two-dimensional data, GNNs can receive three-dimensional data (e.g., 3D surfaces) as input and can output predictions, estimates, classifications, etc. based on the three-dimensional data.

[0102] In at least one embodiment, a U-net architecture is used for one or more machine learning models. U-net is a type of deep neural network that combines an encoder and a decoder with appropriate concatenation between them to capture both local and global features. The encoder is a series of convolutional layers that increase the number of channels while reducing height and width when processing from input to output, while the decoder increases height and width and reduces the number of channels. Layers with the same image height and width from the encoder can be concatenated with the output from the decoder. Any or all convolutional layers from the encoder and decoder can use traditional or depth-wise separable convolutions.

[0103] In at least one embodiment, one or more machine learning models are recurrent neural networks (RNNs). An RNN is a type of neural network that includes a memory that enables the neural network to capture time dependencies. An RNN is able to learn an input-output mapping that depends on both current input and past input. The RNN will process past and future scans and make predictions based on this continuous scan information. The RNN can be trained using a training data set to generate a fixed number of outputs (for example, to classify time-varying data such as video data into a fixed number of categories). One type of RNN that can be used is a long short-term memory (LSTM) neural network.

[0104] A common architecture for this task is the LSTM (Long Short-Term Memory). Unfortunately, the LSTM is not well-suited for images because it does not capture spatial information as well as convolutional networks. To address this, a variant of the LSTM, the ConvLSTM, can be utilized, which incorporates convolution operations within the LSTM cell. The ConvLSTM is a variant of the LSTM (Long Short-Term Memory) that incorporates convolution operations within the LSTM cell. The ConvLSTM replaces matrix multiplication with a convolution operation at each gate in the LSTM cell. This allows it to capture underlying spatial features through convolution operations in multidimensional data. The main difference between the ConvLSTM and the LSTM is the number of input dimensions. Because the LSTM input data is one-dimensional, it is not suitable for spatial sequence data such as video, satellite, and radar imagery datasets. The ConvLSTM is designed to accept 3D data as its input. In at least one embodiment, a CNN-LSTM machine learning model is used. The CNN-LSTM is an integration of a CNN (convolutional layer) and an LSTM. First, the CNN portion of the model processes the data, and the one-dimensional results are fed into the LSTM model.

[0105] Training of neural networks can be achieved in a supervised learning manner, which involves feeding a training dataset consisting of labeled inputs through the network, observing its output, defining an error (by measuring the difference between the output and the labeled value), and using techniques such as deep gradient descent and backpropagation to adjust the network's weights across all its layers and nodes in such a way that the error is minimized. In many applications, repeating this process for many labeled inputs in the training dataset results in a network that can generate correct outputs when presented with inputs that differ from those present in the training dataset. In high-dimensional settings (such as large images), this generalization can be achieved when a sufficiently large and diverse training dataset is available.

[0106] For the model training workflow, a training dataset containing hundreds, thousands, tens of thousands, hundreds of thousands or more intraoral scans, 2D panoramic images and / or 3D models should be used. In an embodiment, up to millions of patient dentition cases that have undergone restorative procedures and / or orthodontic procedures can be used to form a training dataset, where each case can include various labels of one or more types of useful information. Each case can include, for example, data showing a 3D model, intraoral scan, height map, color image, NIRI image, etc. of one or more dental sites, data showing pixel-level segmentation of the data (e.g., 3D model, intraoral scan, height map, color image, NIRI image, etc.) into various dental classifications (e.g., teeth, gingival margin, mobile tissue, saliva, blood, etc.), data showing one or more assigned scan quality metrics for the data, mobile data associated with the 3D scan, etc. The data can be processed to generate one or more training datasets for training one or more machine learning models. For example, a training dataset can include a first training dataset of 2D panoramic images with labeled dental features (e.g., cracks, chips, gum lines, worn tooth areas, cavities, eruption contours, implant gum lines, implant margins, scan volume margins / curves, etc.) and a second training dataset of 3D dentition with labeled dental features. For example, a machine learning model can be trained to detect blood / saliva, detect moving tissue, segment 2D images and / or 3D models of dental sites (e.g., segment such images / 3D surfaces into one or more dental classes), and the like.

[0107] To implement training, processing logic inputs the training dataset into one or more untrained machine learning models. Before inputting the first input into the machine learning model, the machine learning model may be initialized. Processing logic trains the untrained machine learning model based on the training dataset to generate one or more trained machine learning models that perform various operations as described above.

[0108] Training can be performed by inputting one or more of the panoramic 2D images, scans, or 3D surfaces (or data from the images, scans, or 3D surfaces) into the machine learning model at a time. Each input can include data from a panoramic 2D image, an intraoral scan, or a 3D surface in a training data item from a training dataset. For example, a training data item can include a height map, a 3D point cloud, or a 2D image and an associated probability map, which can be input into the machine learning model.

[0109] The machine learning model processes input to generate output. The artificial neural network includes an input layer consisting of values ​​in the data points (e.g., intensity values ​​and / or height values ​​of pixels in a height map). The next layer is called a hidden layer, and each node at the hidden layer receives one or more of the input values. Each node includes parameters (e.g., weights) to apply to the input values. Therefore, each node essentially inputs the input value into a multivariate function (e.g., a nonlinear mathematical transformation) to generate an output value. The next layer can be another hidden layer or an output layer. In either case, the nodes at the next layer receive the output values ​​from the nodes at the previous layer, each node applies weights to these values, and then generates its own output value. This can be performed at each layer. The last layer is the output layer, where there is a node for each category, prediction, and / or output that the machine learning model can generate. For example, for an artificial neural network, it is trained to determine dental features in a 2D panoramic image or a 3D dentition (e.g., represented by a mesh or point cloud).

[0110] The processing logic can then compare the determined dental features with the marked dental features of the panoramic 2D image or 3D point cloud. The processing logic determines an error (i.e., a positioning error) based on the difference between the output dental features and the known correct dental features. The processing logic adjusts the weights of one or more nodes in the machine learning model based on the error. An error term or delta can be determined for each node in the artificial neural network. Based on the error, the artificial neural network adjusts one or more of its parameters (the weights of one or more inputs of the node) for one or more of its nodes. The parameters can be updated in a backpropagation manner so that the nodes in the highest layer are updated first, followed by the nodes in the next layer, and so on. The artificial neural network includes multiple layers of "neurons", each layer receiving values ​​from the neurons in the previous layer as input. The parameters of each neuron include weights associated with the values ​​received from each neuron in the previous layer. Therefore, adjusting the parameters can include adjusting the weights of each input of one or more neurons assigned to one or more layers in the artificial neural network.

[0111] Once the model parameters have been optimized, model validation can be performed to determine whether the model has improved and to determine the current accuracy of the deep learning model. After one or more rounds of training, processing logic can determine whether a stopping criterion has been met. The stopping criterion can be a target level of accuracy, a target number of processed images from the training dataset, a target change in a parameter over one or more previous data points, a combination thereof, and / or other criteria. In at least one embodiment, the stopping criterion is met when at least a minimum number of data points have been processed and at least a threshold accuracy has been reached. The threshold accuracy can be, for example, 70%, 80%, or 90% accuracy. In at least one embodiment, the stopping criterion is met if the accuracy of the machine learning model has stopped improving. If the stopping criterion has not been met, further training is performed. If the stopping criterion has been met, training can be completed. Once the machine learning model has been trained, a retained portion of the training dataset can be used to test the model.

[0112] Once one or more trained ML models are generated, they can be stored in the data store 125 and can be added to the intraoral scanning application 115 and / or used by the dental modeling logic 116. The intraoral scanning application 115 and / or the dental modeling logic 116 can then use the one or more trained ML models and additional processing logic to identify dental features in the panoramic 2D image. In an embodiment, the trained machine learning model can be trained to perform one or more tasks. In at least one embodiment, the trained machine learning model is trained to perform one or more tasks set forth in U.S. patent application publication number 2021 / 0059796A1, entitled “Automated Detection, Generation, And / or Correction of Dental Features in Digital Models,” the entire contents of which are incorporated herein by reference. In at least one embodiment, the trained machine learning model is trained to perform one or more tasks set forth in U.S. patent application publication No. 2021 / 0321872 A1, entitled “Smart Scanning for Intraoral Scans,” which is incorporated herein by reference in its entirety. In at least one embodiment, the trained machine learning model is trained to perform one or more tasks set forth in U.S. patent application publication No. 2022 / 0202295 A1, entitled “Dental Diagnostics Hub,” which is incorporated herein by reference in its entirety.

[0113] In at least one embodiment, a model application workflow includes a first trained model and a second trained model. The first trained model and the second trained model can each be trained to segment input and identify dental features therefrom, but can be trained to operate on different types of data. For example, the first trained model can be trained to operate on 3D data, and the second trained model can be trained to operate on panoramic 2D images. In at least one embodiment, a single trained machine learning model is used to analyze multiple types of data.

[0114] According to one embodiment, an intraoral scanner generates a series of intraoral scans and 2D images. A 3D surface generator can perform registration between the intraoral scans to stitch the intraoral scans together and generate a 3D surface / model based on the intraoral scans. In addition, a 2D intraoral image (e.g., a color 2D image and / or a NIRI 2D image) can be generated. Furthermore, when generating the intraoral scans and 2D images, motion data can be generated by an IMU of the intraoral scanner and / or based on analysis of the intraoral scans and / or 2D intraoral images.

[0115] Data from the 3D model / surface can be input into a first trained model, which outputs a first dental feature. In at least one embodiment, the first dental feature can be output as a probability map or mask, where each point has an assigned probability of being part of the dental feature and / or an assigned probability of not being part of the dental feature. Similarly, for each panoramic 2D image, data from the panoramic 2D image is input into a second trained model, which outputs a dental feature. In at least one embodiment, the dental features can each be output as a probability map or mask, where each pixel of the input 2D image has an assigned probability of being a dental feature and / or an assigned probability of not being a dental feature.

[0116] In at least one embodiment, the machine learning model is additionally trained to identify teeth, gums, and / or excess material. In at least one embodiment, the machine learning model is also trained to determine one or more specific tooth numbers and / or identify a specific indication (or multiple indications) of the input image. Thus, a single machine learning model can be trained to identify dental features and also trained to generally identify teeth, identify different specific tooth numbers, identify gums, and / or identify other features (e.g., margin lines, etc.). In alternative embodiments, a separate machine learning model is trained for each specific tooth number and each specific indication. Thus, a tooth number and / or indication (e.g., a specific dental prosthesis to be used) can be indicated (e.g., can be input by a user), and an appropriate machine learning model can be selected based on the specific tooth number and / or specific indication.

[0117] In embodiments, a machine learning model may be trained to output identification of dental features and separate information indicative of one or more of the foregoing (e.g., insertion path, model orientation, tooth identification, gingival identification, excess material identification, etc.). In at least one embodiment, the machine learning model (or a different machine learning model) is trained to perform one or more of the following: identify teeth represented in the height map, identify gingival represented in the height map, identify excess material (e.g., material that is not gingival or tooth) in the height map, and / or identify dental features in the height map.

[0118] Various embodiments described herein may utilize other methods of generating panoramic 2D images, including methods described in U.S. patent application publication No. 2021 / 0068773A1, entitled “Dental Panoramic Views,” the entire contents of which are incorporated herein by reference.

[0119] Figure 12B A flow chart is shown of a method 1250 for generating projected segmentation and / or classification information from a panoramic 2D image onto a 3D model of a dental site, according to at least one embodiment. At block 1255, a 3D model of the dental site is generated from an intraoral scan (e.g., by a computing device 105 of a dental clinic 108 or a dental laboratory 110). In at least one embodiment, the 3D model of the dental site can include 3D dentition of the patient's mandible, maxilla, or both (e.g., any of the 3D dentitions 210, 660, 730, 760, or 910).

[0120] At block 1260 , a panoramic 2D image is generated from the 3D model of the dental site, for example using any of the methods 1000 , 1100 , or 1200 described in more detail above.

[0121] At block 1265, the dental features identified in the panoramic 2D image may be segmented / classified using one or more trained ML models. The one or more trained ML models may be trained and used according to the methods discussed in more detail above.

[0122] At block 1270 , information describing the segmentation / classification is projected onto the 3D model of the dental part, for example, by identifying and / or labeling dental features at locations in the 3D model that correspond to dental features of the panoramic 2D image.

[0123] An exemplary process may involve the following operations: (1) generating a 2D panoramic image; (2) labeling features of interest in 2D and mapping the labeled features from 2D to 3D; and (3) training a machine learning model on the 3D model using the labeled features.

[0124] Another exemplary process may involve the following operations: (1) generating a 2D panoramic image; (2) labeling features of interest in the 2D panoramic image; (3) training a machine learning model on the 2D panoramic image using the labeled features; and (4) mapping the results of the machine learning model back to a 3D model.

[0125] Figure 13 A diagrammatic representation of a machine in the example form of a computing device 1300 is shown within which a set of instructions may be executed for causing the machine to perform any one or more of the methodologies discussed herein. In alternative embodiments, the machine may be connected (e.g., using a network) to other machines in a local area network (LAN), an intranet, an extranet, or the Internet. The computing device 1300 may correspond to, for example, Figure 1 105 and / or computing device 106. The machine may operate as a server or a client machine in a client-server network environment, or as a peer machine in a peer-to-peer (or distributed) network environment. The machine may be a personal computer (PC), a tablet computer, a set-top box (STB), a personal digital assistant (PDA), a cellular phone, a network appliance, a server, a network router, a switch or a bridge, or any machine capable of executing a set of instructions (sequential or otherwise) that specify actions for the machine to take. Further, while a single machine is illustrated, 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 methodologies discussed herein.

[0126] The example computing device 1300 includes a processing device 1302, a main memory 1304 (e.g., read-only memory (ROM), flash memory, dynamic random access memory (DRAM) such as synchronous DRAM (SDRAM)), etc.), a static memory 1306 (e.g., flash memory, static random access memory (SRAM), etc.), and a secondary memory (e.g., a data storage device 1328), which communicate with each other via a bus 1308.

[0127] Processing device 1302 represents one or more general-purpose processors, such as microprocessors, central processing units, and the like. More specifically, processing device 1302 may 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 another instruction set, or a processor implementing a combination of instruction sets. Processing device 1302 may also be one or more special-purpose processing devices, such as an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), a network processor, and the like. Processing device 1302 is configured to execute processing logic (instructions 1326) for performing the operations and steps discussed herein.

[0128] The computing device 1300 may also include a network interface device 1322 for communicating with a network 1364. The computing device 1300 may also include a video display unit 1310 (e.g., a liquid crystal display (LCD) or a cathode ray tube (CRT)), an alphanumeric input device 1312 (e.g., a keyboard), a cursor control device 1314 (e.g., a mouse), and a signal generating device 1320 (e.g., a speaker).

[0129] The data storage device 1328 may include a machine-readable storage medium (or, more specifically, a non-transitory computer-readable storage medium) 1324 having stored thereon one or more sets of instructions 1326 embodying any one or more of the methods or functionality described herein, such as instructions for the dental modeling logic 116. A non-transitory storage medium refers to a storage medium other than a carrier wave. The instructions 1326 may also reside, completely or at least partially, within the main memory 1304 and / or the processing device 1302 during execution by the computer 1300, with the main memory 1304 and the processing device 1302 also constituting computer-readable storage media.

[0130] The computer-readable storage medium 1324 may also be used to store the dental modeling logic 116, which may include one or more machine learning modules and may perform the operations described above. The computer-readable storage medium 1324 may also store a software library including methods for the dental modeling logic 116. Although the computer-readable storage medium 1124 is shown as a single medium in the example embodiment, the term "computer-readable storage medium" should be considered to include a single medium or multiple media (e.g., a centralized or distributed database, and / or associated caches and servers) that store one or more instruction sets. The term "computer-readable storage medium" should also be considered to include any medium, other than a carrier wave, that is capable of storing or encoding an instruction set for execution by a machine and causing the machine to perform any one or more methods of the present disclosure. Therefore, the term "computer-readable storage medium" should be considered to include, but is not limited to, solid-state memory and optical and magnetic media.

[0131] The following exemplary embodiments are now described:

[0132] Embodiment 1: A method comprising: receiving a three-dimensional (3D) model of a dental part generated based on one or more intraoral scans; generating a projection target, the projection target being shaped to substantially surround a dental arch represented by the dental part; calculating a surface projection by projecting the 3D model of the dental part onto one or more surfaces of the projection target; and generating at least one panoramic two-dimensional (2D) image of the dental part based on the surface projections.

[0133] Embodiment 2: The method of embodiment 1, wherein the surface projection is calculated based on a projection path around the dental arch.

[0134] Embodiment 3: The method according to embodiment 2, wherein the at least one panoramic 2D image is generated by performing orthographic rendering on a flattened mesh, wherein the flattened mesh is generated by projecting the 3D model along the projection path.

[0135] Embodiment 4: The method according to any of the preceding embodiments, wherein the projection target comprises a cylindrical surface substantially surrounding the dental arch.

[0136] Embodiment 5: The method according to any of the preceding embodiments, wherein the projection target comprises a polynomial curved surface substantially surrounding the dental arch.

[0137] Embodiment 6: A method according to any one of the preceding embodiments, wherein the projection target includes: a cylindrical surface; a first planar surface extending from a first edge of the cylindrical surface; and a second planar surface extending from a second edge of the cylindrical surface opposite to the first edge, wherein the cylindrical surface, the first planar surface and the second planar surface together define a continuous surface substantially surrounding the dental arch.

[0138] Embodiment 7: The method of Embodiment 6, wherein the angle between the first planar surface and the second planar surface is about 110° to about 130°.

[0139] Embodiment 8: The method according to any of the preceding embodiments, wherein the dental site corresponds to a single jaw, wherein the first panoramic 2D image corresponds to a buccal rendering, and wherein the second panoramic 2D image corresponds to a lingual rendering.

[0140] Example 9: The method according to Example 8 further includes: generating buccal rendering, lingual rendering, and optional occlusal rendering of the dental part for display, wherein the occlusal rendering is generated by projecting the 3D model of the dental part from the occlusal side of the dental part onto a flat surface.

[0141] Example 10: The method according to any one of the preceding embodiments further includes: generating a panoramic 2D image for display; marking a dental feature at a first position in the panoramic 2D image; determining a second position of the 3D model corresponding to the first position of the panoramic 2D image; and assigning the label of the dental feature to the second position of the 3D model, wherein the 3D model can be displayed together with the label.

[0142] Example 11: A method according to Example 10, wherein labeling the dental features includes one or more of the following: receiving user input to directly label the dental features, or using a trained machine learning model that has been trained to identify and label dental features in panoramic 2D images.

[0143] Example 12: A method comprising: receiving a three-dimensional (3D) model of a dental part generated based on one or more intraoral scans; generating a plurality of vertices along a dental arch represented by the dental part; calculating a projection target, the projection target comprising a plurality of surface segments connected in series to each other at the locations of the vertices; scaling the projection target relative to a dental arch center located within a central region of the dental arch so that the projection target substantially surrounds the dental arch; calculating a surface projection by projecting the 3D model of the dental part onto each of the surface segments of the projection target; and generating at least one panoramic two-dimensional (2D) image of the dental part based on the surface projection.

[0144] Embodiment 13: The method of embodiment 12, wherein one or more of the plurality of vertices are positioned at a center of a tooth, and wherein the number of vertices is greater than five.

[0145] Example 14: A method comprising: receiving a three-dimensional (3D) model of a dental part generated based on one or more intraoral scans; generating a projection target, the projection target being shaped to substantially surround a dental arch represented by the dental part; calculating a first surface projection by projecting the 3D model of the dental part along a buccal direction onto one or more surfaces of the projection target; calculating a second surface projection by projecting the 3D model of the dental part along a lingual direction onto one or more surfaces of the projection target; and generating at least one panoramic two-dimensional (2D) image by combining the first surface projection and the second surface projection.

[0146] Embodiment 15: The method of embodiment 14, wherein generating the at least one panoramic 2D image comprises marking regions of the panoramic 2D image corresponding to overlapping regions of the 3D model identified from the first surface projection and the second surface projection.

[0147] Example 16: An intraoral scanning system comprising: an intraoral scanner; and a computing device operably connected to the intraoral scanner, wherein the computing device is used to perform the method according to any one of Examples 1 to 15 in response to generating one or more intraoral scans using the intraoral scanner.

[0148] Embodiment 17: A non-transitory computer-readable medium comprising instructions, which, when executed by a processing device, cause the processing device to perform the method according to any one of embodiments 1 to 15.

[0149] Example 18: A system comprising: a memory; and a processing device for executing instructions from the memory to perform a method, the method comprising: receiving a three-dimensional (3D) model of a dental part generated based on one or more intraoral scans; generating a projection target, the projection target being shaped to substantially surround a dental arch represented by the dental part; calculating a surface projection by projecting the 3D model of the dental part onto one or more surfaces of the projection target; and generating at least one panoramic two-dimensional (2D) image of the dental part based on the surface projection.

[0150] Embodiment 19: A system according to embodiment 18, wherein the surface projection is calculated based on a projection path around the dental arch, and wherein the at least one panoramic 2D image is generated by orthographic rendering of a flattened mesh, wherein the flattened mesh is generated by projecting the 3D model along the projection path.

[0151] Example 20: The system of Example 18 or Example 19, wherein the projection target comprises a cylindrical surface substantially surrounding the dental arch.

[0152] Embodiment 21: The system of any one of Embodiments 18 to 20, wherein the projection target comprises a polynomial curved surface substantially surrounding the dental arch.

[0153] Example 22: A system according to any one of Examples 18 to 21, wherein the projection target includes: a cylindrical surface; a first planar surface extending from a first edge of the cylindrical surface; and a second planar surface extending from a second edge of the cylindrical surface opposite to the first edge, wherein the cylindrical surface, the first planar surface and the second planar surface together define a continuous surface substantially surrounding the dental arch, and wherein the angle between the first planar surface and the second planar surface is approximately 110° to approximately 130°.

[0154] Example 23: A system according to any one of Examples 18 to 22, wherein the dental part corresponds to a single jaw, wherein the first panoramic 2D image corresponds to a buccal rendering, and wherein the second panoramic 2D image corresponds to a lingual rendering, and wherein the method further comprises: generating a buccal rendering, a lingual rendering, and optionally an occlusal rendering of the dental part for display, wherein the occlusal rendering is generated by projecting a 3D model of the dental part from the occlusal side of the dental part onto a flat surface.

[0155] Example 24: The system according to any one of Examples 18 to 22 further includes: generating a panoramic 2D image for display; marking a dental feature at a first position in the panoramic 2D image; determining a second position of the 3D model corresponding to the first position of the panoramic 2D image; and assigning the label of the dental feature to the second position of the 3D model, wherein the 3D model can be displayed together with the label.

[0156] Example 25: A system according to Example 24, wherein labeling the dental features includes one or more of the following: receiving user input to directly label the dental features, or using a trained machine learning model that has been trained to identify and label dental features in panoramic 2D images.

[0157] While the present disclosure is described with respect to the specific application of human dental assessment, the present disclosure is not limited thereto. The techniques described herein can also be applied to any other medical application. For example, the techniques described can be used for general imaging, and in particular for imaging and characterizing elements of human or animal anatomy (e.g., eyes, noses, other facial features, skeletal structures, etc.).

[0158] Claim language or other language herein reciting "at least one of" a set and / or "one or more" a set means that one member of the set or multiple members of the set (in any combination) satisfies the claim. For example, claim language reciting "at least one of A and B" or "at least one of A or B" means A, B, or A and B. In another example, claim language reciting "at least one of A, B, and C" or "at least one of A, B, or C" means A, B, C, or A and B, or A and C, or B and C, or A, B, and C. Language reciting "at least one of" a set and / or "one or more" a set does not limit the set to the items listed in the set. For example, claim language reciting "at least one of A and B" or "at least one of A or B" can mean A, B, or A and B, and can additionally include items not listed in the set of A and B.

[0159] It should be understood that the above description is intended to be illustrative and not limiting. Many other embodiments will become apparent upon reading and understanding the above description. Although embodiments of the present disclosure have been described with reference to specific example embodiments, it will be appreciated that the present disclosure is not limited to the described embodiments, but may be implemented with modifications and alterations within the spirit and scope of the appended claims. Accordingly, the description and drawings should be regarded as illustrative and not restrictive. Therefore, the scope of the present disclosure should be determined with reference to the appended claims and the full scope of equivalents to which such claims are entitled.

Claims

1. A method comprising: receiving a three-dimensional 3D model of a dental part generated from one or more intraoral scans; generating a projected target shaped to substantially surround a dental arch represented by the dental site; calculating a surface projection by projecting the 3D model of the dental part onto one or more surfaces of the projection target; as well as At least one panoramic two-dimensional (2D) image of the dental part is generated based on the surface projection.

2. The method according to claim 1, wherein The surface projection is calculated based on a projection path around the dental arch.

3. The method according to claim 2, wherein: The at least one panoramic 2D image is generated by orthographic rendering of a flattened mesh generated by projecting the 3D model along the projection path.

4. The method according to any one of claims 1 to 3, wherein: The projection target includes a cylindrical surface substantially surrounding the dental arch.

5. The method according to any one of claims 1 to 4, wherein: The projection target includes a polynomial curved surface substantially surrounding the dental arch.

6. The method according to any one of claims 1 to 5, wherein: The projection targets include: cylindrical surface; a first planar surface extending from a first edge of the cylindrical surface; and a second planar surface extending from a second edge of the cylindrical surface opposite the first edge, wherein the cylindrical surface, the first planar surface, and the second planar surface collectively define a continuous surface substantially surrounding the dental arch.

7. The method according to claim 6, wherein: The angle between the first planar surface and the second planar surface is about 110° to about 130°.

8. The method according to any one of claims 1 to 7, wherein: The dental site corresponds to a single jaw, wherein the first panoramic 2D image corresponds to a buccal rendering, and wherein the second panoramic 2D image corresponds to a lingual rendering.

9. The method according to claim 8, further comprising: The buccal rendering, the lingual rendering, and optionally an occlusal rendering of the dental site are generated for display, the occlusal rendering being generated by projecting the 3D model of the dental site from the occlusal side of the dental site onto a flat surface.

10. The method according to any one of claims 1 to 7, further comprising: generating a panoramic 2D image for display; marking a dental feature at a first position in the panoramic 2D image; determining a second position of the 3D model corresponding to the first position of the panoramic 2D image; as well as A label of the dental feature is assigned to a second location of the 3D model, wherein the 3D model is displayable with the label.

11. The method according to claim 10, wherein: Labeling the dental features includes one or more of: receiving user input to directly label the dental features, or using a trained machine learning model that has been trained to identify and label dental features in the panoramic 2D image.

12. A method comprising: receiving a three-dimensional 3D model of a dental part generated from one or more intraoral scans; generating a plurality of vertices along a dental arch represented by the dental part; calculating a projected target comprising a plurality of surface segments connected in series to one another at the locations of the vertices; scaling the projected object relative to a center of the dental arch within a central region of the dental arch so that the projected object substantially surrounds the dental arch; calculating a surface projection by projecting the 3D model of the dental part onto each of the surface segments of the projection target; as well as At least one panoramic two-dimensional (2D) image of the dental part is generated based on the surface projection.

13. The method according to claim 12, wherein: One or more of the plurality of vertices is located at a center of a tooth, and wherein the number of vertices is greater than five.

14. A method comprising: receiving a three-dimensional 3D model of a dental part generated from one or more intraoral scans; generating a projected target shaped to substantially surround a dental arch represented by the dental site; calculating a first surface projection by projecting the 3D model of the dental site onto one or more surfaces of the projection target along a buccal direction; calculating a second surface projection by projecting the 3D model of the dental part onto one or more surfaces of the projection target along a lingual direction; as well as At least one panoramic two-dimensional (2D) image is generated by combining the first surface projection and the second surface projection.

15. The method according to claim 14, wherein Generating the at least one panoramic 2D image includes marking regions of the panoramic 2D image corresponding to overlapping regions of the 3D model identified from the first surface projection and the second surface projection.

16. An intraoral scanning system comprising: intraoral scanner; as well as A computing device operatively connected to the intraoral scanner, wherein the computing device is configured to perform the method of any one of claims 1 to 15 in response to generating one or more intraoral scans using the intraoral scanner.

17. A non-transitory computer-readable medium comprising instructions which, when executed by a processing device, cause the processing device to perform the method according to any one of claims 1 to 16.

18. A system comprising: Memory; as well as a processing device for executing instructions from the memory to perform a method, the method comprising: receiving a three-dimensional 3D model of a dental part generated from one or more intraoral scans; generating a projected target shaped to substantially surround a dental arch represented by the dental site; calculating a surface projection by projecting the 3D model of the dental part onto one or more surfaces of the projection target; and At least one panoramic two-dimensional (2D) image of the dental part is generated based on the surface projection.

19. The system according to claim 18, wherein: The surface projection is calculated based on a projection path around the dental arch, and wherein the at least one panoramic 2D image is generated by orthographic rendering of a flattened mesh generated by projecting the 3D model along the projection path.

20. A system according to claim 18 or claim 19, wherein The projection target includes a cylindrical surface substantially surrounding the dental arch.

21. A system according to any one of claims 18 to 20, wherein: The projection target includes a polynomial curved surface substantially surrounding the dental arch.

22. A system according to any one of claims 18 to 21, wherein The projection targets include: cylindrical surface; a first planar surface extending from a first edge of the cylindrical surface; and a second planar surface extending from a second edge of the cylindrical surface opposite the first edge, wherein the cylindrical surface, the first planar surface, and the second planar surface collectively define a continuous surface substantially surrounding the dental arch, and wherein an angle between the first planar surface and the second planar surface is from about 110° to about 130°.

23. A system according to any one of claims 18 to 22, wherein: The dental site corresponds to a single jaw, wherein the first panoramic 2D image corresponds to a buccal rendering, and wherein the second panoramic 2D image corresponds to a lingual rendering, and wherein the method further comprises: The buccal rendering, the lingual rendering, and optionally an occlusal rendering of the dental site are generated for display, the occlusal rendering being generated by projecting the 3D model of the dental site from the occlusal side of the dental site onto a flat surface.

24. The system according to any one of claims 18 to 23, further comprising: generating a panoramic 2D image for display; marking a dental feature at a first position in the panoramic 2D image; determining a second position of the 3D model corresponding to the first position of the panoramic 2D image; as well as A label of the dental feature is assigned to a second location of the 3D model, wherein the 3D model is displayable with the label.

25. The system of claim 24, wherein: Labeling the dental features includes one or more of: receiving user input to directly label the dental features, or using a trained machine learning model that has been trained to identify and label dental features in the panoramic 2D image.

Images