Apparatus and method for determining implant prosthetic option using machine learning model

Abstract

A method for determining a prosthetic option for an implant is characterized by comprising the steps of: acquiring an implant image including an implant fixture region; identifying the manufacturer of the implant based on the implant image through a first machine learning model; estimating the detailed specifications of the implant fixture based on the implant fixture region through a second machine learning model and identifying an implant model corresponding to the detailed specifications; and determining a prosthetic option corresponding to the upper diameter of the implant fixture by referring to a database.

Description

Technology Field

[0001] The present disclosure relates to an apparatus and method for determining implant prosthetic options. More specifically, the present disclosure relates to an apparatus and method for determining implant prosthetic options using a machine learning model. Background Technology

[0002] As decades have passed since the introduction of implant treatment in dentistry, the need for retreatment of aging implant prostheses is increasing. To restore the superstructure of an implant, the manufacturer, model name, and detailed specifications of each implant may be required. However, difficulties may arise in identifying the manufacturer, model name, and detailed specifications of the implant when the retreatment process is performed by a different attending physician or when it is difficult or impossible to access existing medical records.

[0003] Traditionally, the process has relied on conjectures based on doctors' experience. For example, the conventional method involved taking periapical or panoramic radiographs of the implant fixture, ordering prosthetic options based on doctors' estimates, and fitting them into the patient's oral cavity. If the ordered prosthesis did not fit the implant fixture, this involved a trial-and-error process of purchasing and fitting different parts, which could consume a significant amount of time and money. Prior art literature

[0004] (Patent Document 001) US2001-0021498 (Patent Document 002) US2021-0030520 (Patent Document 003) US2012-0308963 The problem to be solved

[0005] The purpose of the embodiments disclosed in this disclosure is to provide a method for identifying the manufacturer of an implant through an implant image, estimating detailed specifications, and determining prosthetic options.

[0006] The problems that this disclosure aims to solve are not limited to those mentioned above, and other unmentioned problems will be clearly understood by a person skilled in the art from the description below. means of solving the problem

[0007] An apparatus for determining a prosthetic option for an implant according to one aspect of the present disclosure for achieving the above-described technical problem can acquire an implant image including an implant fixture region, identify the manufacturer of the implant based on the implant image through a first machine learning model, estimate the detailed specifications of the implant fixture based on the implant fixture region through a second machine learning model, identify an implant model corresponding to the detailed specifications, and determine a prosthetic option corresponding to the upper diameter of the implant fixture by referring to a database.

[0008] In addition to this, a computer program stored on a computer-readable recording medium for executing the present disclosure may be further provided.

[0009] In addition, a computer-readable recording medium for recording a computer program for executing a method for implementing the present disclosure may be further provided. Effects of the invention

[0010] According to the aforementioned means for solving the problem of the present disclosure, by determining the prosthetic option of an unknown implant fixture using a machine learning model, the effect of shortening the time for implant retreatment and improving accuracy is provided.

[0011] The effects of the present disclosure are not limited to those mentioned above, and other unmentioned effects will be clearly understood by a person skilled in the art from the description below. Brief explanation of the drawing

[0012] FIG. 1 illustrates an implant prosthetic option determination system of the present disclosure. FIG. 2 is a block diagram of an implant prosthetic option determination device of the present disclosure. FIG. 3 is a flowchart illustrating the operation of the device of the present disclosure. Figure 4 illustrates the operation of the first machine learning model. Figures 5a and 5b illustrate the operation of the second machine learning model. Figure 6a illustrates the operation of determining prosthetic options using a database. Figure 6b illustrates the relationships between data stored in a database. Specific details for implementing the invention

[0013] Throughout this disclosure, the same reference numerals denote the same components. This disclosure does not describe all elements of the embodiments, and general content in the art to which this disclosure pertains or content that overlaps between embodiments is omitted. The terms 'part, module, component, block' as used in the specification may be implemented in software or hardware, and depending on the embodiments, a plurality of 'parts, modules, components, blocks' may be implemented as a single component, or a single 'part, module, component, block' may include a plurality of components.

[0014] Throughout the specification, when a part is described as being "connected" to another part, this includes not only cases where they are directly connected but also cases where they are indirectly connected, and indirect connections include connections made via a wireless communication network.

[0015] Furthermore, when it is stated that a part "includes" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0016] Throughout the specification, when it is stated that a component is located "on" another component, this includes not only cases where a component is in contact with another component, but also cases where another component exists between the two components.

[0017] The terms first, second, etc. are used to distinguish one component from another, and the components are not limited by the aforementioned terms.

[0018] Singular expressions include plural expressions unless there is an obvious exception in the context.

[0019] In each step, identification codes are used for convenience of explanation and do not describe the order of the steps; the steps may be performed differently from the specified order unless a specific order is clearly indicated in the context.

[0020] The operating principles and embodiments of the present disclosure will be described below with reference to the attached drawings.

[0021] In this specification, the term "device according to the present disclosure" includes all various devices capable of performing computational processing and providing results to a user. For example, the device according to the present disclosure may include a computer, a server device, and a portable terminal, or may be in the form of any one of these.

[0022] Here, the computer may include, for example, a notebook, desktop, laptop, tablet PC, slate PC, etc. equipped with a web browser.

[0023] The above server device is a server that processes information by communicating with an external device, and may include an application server, a computing server, a database server, a file server, a game server, a mail server, a proxy server, and a web server.

[0024] The above portable terminal may include, for example, all types of handheld-based wireless communication devices such as PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), WiBro (Wireless Broadband Internet) terminals, smartphones, etc., as well as wearable devices such as watches, rings, bracelets, anklets, necklaces, glasses, contact lenses, or head-mounted devices (HMDs).

[0025] FIG. 1 illustrates an implant prosthetic option determination system of the present disclosure.

[0026] In one embodiment, the implant (110) may be composed of a fixture (114) and a prosthesis (112). The fixture (114) and the prosthesis (112) may have various detailed specifications and shapes depending on the manufacturer.

[0027] In one embodiment, the implant prosthetic option determination device (100) (hereinafter, the device) can identify the manufacturer of the implant fixture (114) using a machine learning model. In one embodiment, the device (100) can estimate the detailed specifications of the implant fixture (114) and determine options for the prosthesis (112) (e.g., prosthetic-related parts) based on the detailed specifications of the implant fixture (114) (e.g., length and / or upper diameter).

[0028] FIG. 2 is a block diagram of an implant prosthetic option determination device of the present disclosure.

[0029] Referring to FIG. 2, the device (100) according to the present disclosure may include a control unit (200), an image acquisition unit (210), and / or a memory (220). In one embodiment, the memory (220) may store a first machine learning model (230), a second machine learning model (240), and / or a database.

[0030] In one embodiment, the control unit (200) may be implemented with a memory (220) that stores data for an algorithm or a program that reproduces the algorithm for controlling the operation of components within the device, and at least one processor (not shown) that performs the aforementioned operation using the data stored in the memory (220). In this case, the memory (220) and the processor may each be implemented as separate chips. Alternatively, the memory (220) and the processor may be implemented as a single chip.

[0031] In addition, the control unit (200) can control one or a combination of the components described above in order to implement various embodiments according to the present disclosure described in FIGS. 2 to 8 below on the device.

[0032] In one embodiment, the image acquisition unit (210) can acquire an implant image. For example, the implant image may be a periapical radiograph and / or a panoramic radiograph. The implant image may include an implant fixture region.

[0033] In one embodiment, the memory (220) can store data supporting various functions of the device (100) and a program for the operation of the control unit (200), and can store input / output data (e.g., music files, still images, video, etc.), and can store a plurality of application programs (or applications) running on the device (100), data for the operation of the device (100), and commands. At least some of these application programs may be downloaded from an external server via wireless communication.

[0034] Such memory (220) may include at least one type of storage medium among flash memory type, hard disk type, SSD type (Solid State Disk type), SSD type (Silicon Disk Drive type), multimedia card micro type, card type memory (e.g., SD or XD memory, etc.), RAM (random access memory; RAM), SRAM (static random access memory), ROM (read-only memory; ROM), EEPROM (electrically erasable programmable read-only memory), PROM (programmable read-only memory), magnetic memory, magnetic disk, and optical disk. Additionally, the memory (220) may be a database that is separated from the device (100) but connected via wired or wireless connection.

[0035] In one embodiment, the first machine learning model (230) can identify the implant manufacturer based on the implant image. In one embodiment, the first machine learning model (230) may use a deep learning method based on a deep neural network. For example, the first machine learning model (230) may be based on a convolutional neural network (CNN) method. The specific operation of the first machine learning model (230) will be described later.

[0036] In one embodiment, the second machine learning model (240) can estimate the detailed specifications of the implant fixture based on the manufacturer and the implant image. In one embodiment, the second machine learning model (240) may use a deep learning method or a rule-based learning method. For example, the detailed specifications may be the length and / or upper diameter of the implant fixture.

[0037] In one embodiment, the database (250) may store information on the implant manufacturer (or model), information on the upper diameter of the fixture, information on prosthetic options, information on related parts according to the prosthetic options, and the relationships between them.

[0038] In one embodiment, the device for determining the prosthetic option of an implant (100) can acquire an implant image including an implant fixture region, identify the manufacturer of the implant based on the implant image through a first machine learning model (230), estimate the detailed specifications of the implant fixture based on the implant fixture region through a second machine learning model (240), identify an implant model corresponding to the detailed specifications, and determine a prosthetic option corresponding to the upper diameter of the implant fixture by referring to a database (250). For example, the implant image may include a periapical radiograph and a panoramic radiograph.

[0039] In one embodiment, the first machine learning model (230) can be connected to the second machine learning model (240) in a cascade structure.

[0040] In one embodiment, the first machine learning model (230) can identify the manufacturer based on the balance between the diameter and length of the implant fixture.

[0041] In one embodiment, the second machine learning model (240) is based on a deep learning method, and the second machine learning model (240) is trained using a plurality of implant images, and the plurality of implant images may include images in which the implant fixture region is specified by a bounding box or images in which feature points of the implant fixture are marked. For example, the feature points of the implant fixture may include the upper two end points and the lowest point of the implant fixture.

[0042] In one embodiment, the second machine learning model (240) may be based on a predefined rule-based learning method using the number of screw threads of the implant fixture and the length according to the number of screw threads.

[0043] In one embodiment, the detailed specifications may include the length and upper diameter of the implant fixture.

[0044] In one embodiment, the database (250) may store information on the implant model, information on the upper diameter of the implant fixture, information on prosthetic options, information on related parts according to the prosthetic options, and the relationships between them.

[0045] The components illustrated in FIG. 2 are not essential for implementing the device (100) according to the present disclosure, so the device (100) described in this specification may have more or fewer components than the components listed above.

[0046] Meanwhile, each component illustrated in Figure 2 refers to a software and / or hardware component such as a Field Programmable Gate Array (FPGA) and an Application Specific Integrated Circuit (ASIC).

[0047] FIG. 3 is a flowchart illustrating the operation of the device of the present disclosure.

[0048] In operation 300, the control unit (200) of the device (100) can acquire an implant image through the image acquisition unit (210). For example, the implant image may be a periapical radiograph and / or a panoramic radiograph. The implant image may include an implant fixture area.

[0049] In operation 310, the control unit (200) can identify the manufacturer of the implant using the first machine learning model (230). Implant models produced by a specific implant manufacturer may have similar characteristics. The first machine learning model (230) may be based on a CNN method. The first machine learning model (230) may be trained using multiple implant images. The first machine learning model (230), once trained, can classify the manufacturer of the implant images as a specific manufacturer. The specific operation of the first machine learning model (230) will be described later.

[0050] In operation 320, the control unit (200) can estimate the detailed specifications of the implant fixture using the second machine learning model (240). For example, the detailed specifications may include the upper diameter and / or length of the implant fixture.

[0051] In one embodiment, for the estimation of detailed specifications, the control unit (200) may input an image of the implant fixture region into the second machine learning model (240). The second machine learning model (240) may estimate the detailed specifications of the implant fixture using a deep learning or rule-based learning method. The specific operation of the second machine learning model (240) will be described later.

[0052] In operation 330, the control unit (200) can identify an implant model based on detailed specifications (e.g., ratio of upper diameter to length) and determine a prosthetic option corresponding to the upper diameter of the implant fixture. In one embodiment, the control unit (200) can refer to a database (250) to determine the implant model and the prosthetic option corresponding to the upper diameter of the fixture.

[0053] Figure 4 illustrates the operation of the first machine learning model.

[0054] In one embodiment, the control unit (200) may input an implant image (400) into a first machine learning model (230). The implant image (400) may include an implant fixture region (405).

[0055] In one embodiment, the first machine learning model (230) can identify the implant manufacturer based on the implant image (400). The first machine learning model (230) may be an artificial intelligence model trained on multiple implant images based on a CNN method.

[0056] In one embodiment, the implant manufacturer according to the implant image (400) may be Company A. Company A may be a manufacturer that produces a plurality of implant models (412, 414, 416, 418). The plurality of implant models (412, 414, 416, 418) manufactured by Company A may have common or similar features. For example, a common or similar feature may be a balance based on the length and diameter of the implant fixture. As another example, a common or similar feature may further include the number of threads on the implant fixture. In one embodiment, the first machine learning model (230) may learn the features through a plurality of implant images. Based on the learning results, the first machine learning model (230) may classify and identify the manufacturer of the implant image (400) as Company A.

[0057] FIGS. 5a and 5b illustrate the operation of a second machine learning model. In FIG. 5a, the second machine learning model (240) may use a deep learning method. In FIG. 5b, the second machine learning model (240) may use a rule-based learning method.

[0058] Referring to FIG. 5a, the first machine learning model (230) may be connected to the second machine learning model (240) in a cascade manner. The second machine learning model (240) may be an artificial intelligence model that has completed training through a plurality of implant images. In one embodiment, the plurality of implant images may be preprocessed images. For example, the plurality of implant images may be images in which the implant fixture region is specified by a bounding box through object detection or manual notation. As another example, the plurality of implant images may be images in which feature points of the implant fixture (e.g., the upper two end points and the lowest point of the fixture) are marked. In yet another embodiment, the plurality of implant images may be original images that have not undergone separate preprocessing. In one embodiment, the second machine learning model (240) may be configured to place weight on the upper diameter rather than the length of the implant fixture during training.

[0059] In one embodiment, the control unit (200) may input manufacturer information identified in the first machine learning model (230) and images (500, 505) of the implant fixture region into the second machine learning model (240). The images of the implant fixture region (405) may be images that do not include separate markings, such as image (500), or images that include markings, such as image (505). For example, the marking points in image (505) may include at least one feature point of the fixture (e.g., upper two end points (506, 507) and lowest point (508)) of the fixture. The at least one feature point is exemplary and the embodiments of the present disclosure are not limited thereto.

[0060] In one embodiment, the second machine learning model (240) can identify detailed specifications of the implant fixture based on a deep learning method. For example, the second machine learning model (240) can estimate the upper diameter and length (or the ratio of the upper diameter to the length) of the implant fixture and identify the implant model (510) corresponding to the implant image (400) among the implant models (510, 512, 514) manufactured by the identified manufacturer. The plurality of implant models (510, 512, 514) of FIG. 5a may have the same number of screw threads but different upper diameters.

[0061] Referring to FIG. 5b, the second machine learning model (240) can be connected to the first machine learning model (230) in a cascade manner. The control unit (200) can input manufacturer information identified in the second machine learning model (240) and an image (550) of the implant fixture area into the second machine learning model (240).

[0062] In one embodiment, the second machine learning model (240) can extract the number of screw threads (555) from an image (550) of the fixture area using a rule-based learning method. The rules of the second machine learning model (240) can be generated in advance by a developer. The second machine learning model (240) can estimate the length (l) of the implant fixture based on the number of screw threads. For example, the second machine learning model (240) can estimate the length (L) of the implant fixture to be 12 mm when the number of screw threads is 12. The second machine learning model (240) can calculate the upper diameter of the implant fixture using the ratio (Φ / l) between the upper diameter (Φ) and the length (l) on the image (550) of the implant fixture area. The upper diameter of the implant fixture can be calculated using Equation 1.

[0063]

[0064] In one embodiment, the second machine learning model (240) can identify the implant model (560) corresponding to the implant image (400) among the implant models (560, 562, 564) manufactured by the manufacturer based on the calculated upper diameter of the implant fixture. The plurality of implant models (560, 562, 564) of FIG. 5b may have the same number of screw threads but different upper diameters.

[0065] Fig. 6a illustrates the operation of determining prosthetic options using a database. Fig. 6b illustrates the relationships between data stored in the database.

[0066] Referring to FIG. 6a, the control unit (200) can determine an implant model (600) based on the detailed specifications of the implant fixture identified through operations 300 to 320. The control unit (200) can determine a prosthetic option according to the upper diameter of the implant model (600) by referring to the database (250).

[0067] Referring to FIG. 6b, the database (250) may store implant model information (650), upper diameter information of the fixture (660), prosthetic option information (670), related part information according to the prosthetic option (670, 672, 674), and the relationships between them.

[0068] Referring again to FIG. 6a, the control unit (200) can identify the upper diameter of the implant model (600) as 4.8 WN by referring to the implant model information (650) of the database (250) and the corresponding upper diameter information (660) of the implant fixture. The control unit (200) can determine the prosthetic option of the implant model (600) as Tissue level Wide Neck corresponding to 4.8 WN by using the prosthetic option information (670). The control unit (200) can provide related part information (674) according to the prosthetic option (e.g., Tissue level Wide Neck) determined by referring to the database (250).

[0069] Meanwhile, the disclosed embodiments may be implemented in the form of a recording medium that stores instructions executable by a computer. The instructions may be stored in the form of program code and, when executed by a processor, may generate a program module to perform the operation of the disclosed embodiments. The recording medium may be implemented as a computer-readable recording medium.

[0070] Computer-readable recording media include all types of recording media that store instructions that can be decoded by a computer. Examples include ROM (Read Only Memory), RAM (Random Access Memory), magnetic tape, magnetic disk, flash memory, optical data storage devices, etc.

[0071] As described above, the disclosed embodiments have been explained with reference to the attached drawings. Those skilled in the art will understand that the present disclosure may be practiced in forms different from the disclosed embodiments without changing the technical spirit or essential features of the present disclosure. The disclosed embodiments are illustrative and should not be interpreted restrictively.

Claims

Claim 1 A method performed by a device for determining prosthetic options for an implant, comprising: a step of acquiring an implant image including an implant fixture region; a step of inputting the acquired implant image into a first machine learning model to identify the manufacturer of the implant based on the implant image; a step of estimating detailed specifications including the upper diameter and length of the implant fixture based on the implant fixture region through a second machine learning model and identifying an implant model corresponding to the detailed specifications; and a step of determining a prosthetic option corresponding to the upper diameter of the implant fixture by referring to a database.The device for determining the prosthetic option of the implant includes the first machine learning model and the second machine learning model being connected in a cascade manner, and the device inputs the manufacturer of the implant identified through the first machine learning model and the image of the implant fixture region into the second machine learning model, wherein the second machine learning model learns by giving weight to the upper diameter rather than the length of the implant fixture, extracts the number of screw threads from the image of the implant fixture region using a rule-based learning method, estimates the length (L) of the implant fixture based on the number of screw threads, and, based on the learned result, identifies the implant model corresponding to the implant image among the implant models manufactured by the manufacturer of the implant identified through the first machine learning model, wherein the implant models manufactured by the manufacturer of the implant have the same number of screw threads but different upper diameters, and the first machine learning model learns the number of screw threads of the implant fixture as a characteristic through a plurality of implant images so as to identify the manufacturer from the implant image, and the second machine learning model [is] of the estimated implant fixture Calculate the upper diameter (D) of the implant fixture by applying the ratio (Φ / l) between the length (L) and the upper diameter (Φ) and length (l) on the image of the implant fixture region to the following Equation 1, [Equation 1]; A method for determining a prosthetic option, which identifies an implant model corresponding to the calculated upper diameter (D) among the above implant models. Claim 2 A method for determining prosthetic options according to claim 1, wherein the implant images include a periapical radiograph and a panoramic radiograph. Claim 3 delete Claim 4 A method for determining a prosthetic option according to claim 1, wherein the first machine learning model identifies the manufacturer based on the balance between the diameter and length of the implant fixture. Claim 5 A method for determining a prosthetic option according to claim 4, wherein the second machine learning model is trained using a plurality of implant images based on a deep learning method, and the plurality of implant images include images in which the implant fixture region is specified by a bounding box or images in which feature points of the implant fixture are marked. Claim 6 A method for determining prosthetic options according to claim 5, wherein the characteristic points of the implant fixture include the upper two end points and the lowest point of the implant fixture. Claim 7 In claim 6, the second machine learning model is based on a predefined rule-based learning method using the number of screw threads of the implant fixture and the length according to the number of screw threads, a method for determining a prosthetic option. Claim 8 delete Claim 9 A method for determining a prosthetic option according to claim 1, wherein the database includes information on the implant model, information on the upper diameter of the implant fixture, information on prosthetic options, information on related parts according to the prosthetic options, and the relationships between them. Claim 10 A program stored on a computer-readable recording medium to execute a method for determining a prosthetic option according to any one of claims 1, 2, 4 through 7 and 9, combined with a computer. Claim 11 A computer-readable recording medium combined with a computer, storing a program for executing a method for determining a prosthetic option according to any one of claims 1, 2, 4 through 7 and 9. Claim 12 A control unit that controls the operation of an image acquisition unit; and a memory in which a first machine learning model, a second machine learning model, and a database are stored;The control unit acquires an implant image including an implant fixture region using the image acquisition unit, inputs the acquired implant image into a first machine learning model to identify the manufacturer of the implant based on the implant image, estimates the detailed specifications of the implant fixture based on the implant fixture region through the second machine learning model, identifies an implant model corresponding to the detailed specifications, and determines a prosthetic option corresponding to the upper diameter of the implant fixture by referring to a database, wherein the first machine learning model and the second machine learning model are connected in a cascade manner, and the device for determining the prosthetic option of the implant inputs the manufacturer of the implant identified through the first machine learning model and the image of the implant fixture region into the second machine learning model, wherein the second machine learning model learns by giving more weight to the upper diameter than to the length of the implant fixture, extracts the number of screw threads from the image of the implant fixture region using a rule-based learning method, estimates the length (L) of the implant fixture based on the number of screw threads, and based on the learned result, the The method is characterized by identifying the implant model corresponding to the implant image among the implant models manufactured by the manufacturer of the implant identified through the first machine learning model, wherein the implant models manufactured by the manufacturer of the implant have the same number of screw threads but different upper diameters, and the first machine learning model learns the number of screw threads of the implant fixture as a characteristic through a plurality of implant images so as to identify the manufacturer from the implant image, and the second machine learning model calculates the upper diameter (D) of the implant fixture by applying the estimated length (L) of the implant fixture and the ratio (Φ / l) between the upper diameter (Φ) and the length (l) on the image of the implant fixture region to the following Equation 1, [Equation 1]; A prosthetic option determination device that identifies an implant model corresponding to the calculated upper diameter (D) among the above implant models. Claim 13 A prosthetic option determining device according to claim 12, wherein the implant images include a periapical radiograph and a panoramic radiograph. Claim 14 delete Claim 15 A prosthetic option determination device according to claim 12, wherein the first machine learning model identifies the manufacturer based on the balance between the diameter and length of the implant fixture. Claim 16 A prosthetic option determining device according to claim 15, wherein the second machine learning model is trained using a plurality of implant images based on a deep learning method, and the plurality of implant images include an image in which the implant fixture region is specified by a bounding box or an image in which feature points of the implant fixture are marked. Claim 17 A prosthetic option determining device according to claim 16, wherein the characteristic points of the implant fixture include the upper two end points and the lowest point of the implant fixture. Claim 18 In claim 17, the prosthetic option determination device, wherein the second machine learning model is based on a predefined rule-based learning method using the number of screw threads of the implant fixture and the length according to the number of screw threads. Claim 19 delete Claim 20 In claim 12, the prosthetic option determination device, wherein the database includes information on the implant model, information on the upper diameter of the implant fixture, information on prosthetic options, information on related parts according to the prosthetic options, and the relationships between them.

Images