3D printing aided design system for dental restoration material

Abstract

The invention relates to the technical field of three-dimensional model design, and discloses a 3D printing aided design system for a dental restoration material, which comprises a data acquisition module, an initial model design module, a design optimization module and a result output module, an initial model design module and a design optimization module are arranged, so that a prosthesis model with the highest matching degree with an individual is selected from a tooth form database to serve as an initial model on the basis of construction of an oral cavity model of the individual, and optimal adjustment parameters are obtained on the basis of an improved optimization algorithm; adjusting and optimizing the prosthesis design parameters of the initial model according to the optimal adjustment parameters to obtain an optimized prosthesis model; according to the method, intelligent optimization design is carried out on the repair material through accurate data, optimization of personalized repair material design is carried out in combination with a tooth form database of a human body and individual data of a patient, and then the attractiveness and practicability of the tooth appearance are guaranteed at the same time.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional model design, and more particularly to a 3D printing assisted design system for oral restoration materials. Background Art

[0002] With the continuous development of oral restoration technology, more and more patients need high-quality restorations to restore oral function and aesthetics. Traditional oral restoration design relies on manual model making and casting technology, which has problems such as long design cycles, insufficient accuracy, and limited humanization. The rapid development of 3D printing technology has made its application in the field of oral restoration increasingly widespread.

[0003] A disclosed document with the publication number CN108992193B discloses a tooth restoration assisted design method. A three-dimensional human face model and a high-precision digital three-dimensional model of the patient's teeth are quickly obtained through a three-dimensional scanning device. Through the matching relationship between the two models, the collected texture map is pasted onto the digital three-dimensional model of the teeth and registered into the oral cavity of the human face model. Then, a single or multiple standard dental models are used to replace the single or multiple teeth to be restored. The doctor adjusts the parameters of the standard dental model to design the restoration of the teeth to meet aesthetic design and patient requirements. At the same time, the patient can view the three-dimensional model of their teeth after restoration, which is more realistic and reliable. Finally, the designed teeth can be directly printed into solid teeth, and then an impression mold is pressed. Medical restoration material is poured into the impression mold, and then the impression mold is put on the teeth to be restored or the teeth after tooth preparation of the patient. After the restoration material solidifies on the teeth, the impression mold is removed, and finally, the tooth restoration assisted design is realized.

[0004] However, during the design process of oral restoration materials, it is still manually designed, and the restoration materials cannot be intelligently optimized designed through accurate data. The tooth shape database of the human body and the individual data of the patient cannot be combined to optimize the personalized design of the restoration materials, and thus the aesthetics and practicality of the tooth shape cannot be ensured at the same time.

[0005] In view of this, the present invention proposes a 3D printing assisted design system for oral restoration materials to assist in optimizing the design of oral restoration materials with scientific data as the support. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, the present invention provides a 3D printing assisted design system for oral restoration materials to solve the problems existing in the above background art.

[0007] The present invention provides the following technical solutions: A 3D printing assisted design system for oral restoration materials includes a data acquisition module, an initial model design module, a design optimization module, and a result output module; The data acquisition module is used to collect individual oral data; The initial model design module constructs an oral model of an individual based on the data collected by the data acquisition module, and selects the prosthetic model with the highest matching degree with the individual from the dental shape database as the initial model; The design optimization module is used to obtain the optimal adjustment parameters based on the improved optimization algorithm, adjust and optimize the prosthetic design parameters of the initial model according to the optimal adjustment parameters, obtain the optimized prosthetic model, and at the same time obtain the predicted functional performance of the optimized prosthetic; The result output model is used to output the optimized prosthetic model. After obtaining the final prosthetic model, a 3D printing operation is performed.

[0008] Preferably, the individual oral data is the three-dimensional oral data of the patient, the individual is the patient who needs oral repair, and the three-dimensional oral data includes three-dimensional oral image data, tooth shape data, alveolar bone shape data, and oral space data.

[0009] Preferably, the method for obtaining the initial model is as follows: Based on the constructed three-dimensional oral model of the individual, extract the key feature points of the three-dimensional oral model to form a key feature point set; establish the key feature point set of the dental shape database; Place each prosthetic model in the dental shape database into the three-dimensional oral model of the individual, obtain the corresponding coordinates of each key feature point in the key feature point set of each prosthetic model, and use the key feature points in the three-dimensional oral model of the individual and the same key feature points corresponding to the prosthetic model in the dental shape database as a group of key feature point pairs, that is, each prosthetic model has m groups of key feature point pairs and corresponds one by one to the three-dimensional oral model of the individual; obtain the similarity of each group of key feature point pairs, and perform weighted average to obtain the average similarity of the key feature point pairs of each prosthetic model. Take the value of the average similarity as the value of the matching degree, and select the prosthetic model corresponding to the highest value of the matching degree as the initial model, that is, the prosthetic model corresponding to the largest value of the average similarity of the key feature point pairs as the initial model.

[0010] Preferably, the key feature point set formed by the key feature points of the three-dimensional oral model is expressed as: , where DJ_i is the key feature point set of the i-th tooth body, and TZ ij is the coordinate of the j-th key feature point in the key feature point set corresponding to the i-th tooth body, j = 1, 2, 3,..., m; i = 1, 2, 3,..., n; n is the total number of tooth bodies of the individual; m is the number of key feature points; The key feature point set of the dental shape database is expressed as: , where SJ_a is the set of key feature points of the a-th prosthesis model in the tooth profile database, and sj aj is the coordinate of the j-th key feature point in the set of key feature points of the a-th prosthesis model; j = 1, 2, 3, …, m.

[0011] Preferably, the coordinate point similarity of each group of key feature point pairs is calculated using the Euclidean distance: , where Dis_aj is the similarity of the j-th group of key feature point pairs in the a-th prosthesis model, and TZ ij _r is the coordinate of TZ ij in the r-th dimension, sj aj _r is the coordinate of sj aj in the r-th dimension, and R is the total dimension; the similarities of the m groups of key feature points are weighted and averaged to obtain the average similarity; , and PJ_a is the matching degree of the a-th prosthesis model.

[0012] Preferably, the specific method for obtaining the optimal adjustment parameter is as follows: Obtain an adjustment parameter set, where the adjustment parameter set is the adjustment data of each key feature point of the prosthesis model; the adjustment parameter set can be expressed as: , where U f is the f-th adjustment parameter set, uj is the adjustment data of the j-th key feature point, and j = 1, 2, 3, …, m; based on the improved simulated annealing algorithm, obtain the optimal adjustment parameter set from the adjustment parameter set, and the adjustment parameters in the optimal adjustment parameter set are the optimal adjustment parameters.

[0013] Preferably, the improved simulated annealing algorithm includes the following steps: Step S11: Represent the adjustment parameter set as U1, U2, U3, …, U f 、…、U F ; Step S12: Preset the initial temperature T max 、the lowest temperature T min 、the temperature reduction coefficient δ, and the maximum number of iterations φ, and let the current temperature T = T max ; Step S13: Randomly set a feasible solution ρ, and the feasible solution ρ is the adjustment parameter set; Step S14: Determine the fitness function; Step S15: Calculate the fitness corresponding to the feasible solution; take the feasible solution as the current point, perform random perturbation in the neighborhood of the current point to obtain a new feasible solution ρ´, and calculate the fitness corresponding to the new feasible solution ρ´; Step S16: Calculate the fitness difference, and the expression of the fitness difference is: , where λ* is the fitness difference, λ ρ is the fitness of the feasible solution corresponding to the current point, λ ρ´ is the fitness of the new feasible solution; If the fitness difference λ * > 0, then let ρ = ρ´, that is, assign the value of the new feasible solution ρ´ to the feasible solution ρ; if the fitness difference λ * ≤ 0, then calculate the probability P, and let ρ = ρ´ according to the probability P; the calculation formula of the probability P is expressed as: , where e is the natural constant; Step S17: Loop steps S15 to S16 until the number of loops reaches the maximum number of iterations, then the loop ends and enter step S18; Step S18: Cool down the current temperature to obtain a new current temperature, cool down the current temperature T in step S12, and use the cooled value as the new current temperature, and represent it as T in the next iteration; Reset the number of iterations, that is, use the reduced value of the maximum number of iterations as the new maximum number of iterations, and represent it as φ in the next iteration; if the reduced maximum number of iterations is not an integer, then round up the reduced maximum number of iterations to make the reduced maximum number of iterations an integer; Step S19: Loop steps S15 to S18; until the current temperature T < T min at this time, the loop ends; obtain the adjustment parameter set corresponding to the feasible solution ρ, which is the optimal adjustment parameter set, and the adjustment parameters in the optimal adjustment parameter set are the optimal adjustment parameters.

[0014] Preferably, the formula for cooling down the current temperature to obtain a new current temperature in step S18 is expressed as: T new = T × δ, where T new is the new current temperature after cooling, and T is the current temperature before cooling; The formula for resetting the number of iterations is expressed as: φ new = φ × δ, where φ new is the reset maximum number of iterations.

[0015] Preferably, the fitness function is expressed as: , where SP ρ is the functional performance gain of the adjustment parameter set corresponding to the feasible solution ρ; The way to obtain the functional performance gain of the adjustment parameter set is: Input the adjusted key feature point data corresponding to the adjustment parameter set and the individual oral data into the functional performance prediction model to obtain the predicted functional performance. Input the key feature point data of the initial model and the human oral data into the functional performance prediction model to obtain the initial functional performance. Subtract the initial functional performance from the predicted functional performance to obtain the functional performance gain corresponding to the adjustment parameter set.

[0016] Preferably, the training method of the functional performance prediction model is as follows: Take the adjusted key feature point data corresponding to each adjustment parameter set and the individual oral data as a set of analysis data. Pre-collect d sets of analysis data, and obtain the corresponding functional performance for each of the d sets of analysis data; d is an integer greater than 1. Convert the analysis data and the corresponding functional performance into a corresponding set of feature vectors; Take each set of feature vectors as the input of the functional performance prediction model. The functional performance prediction model takes a functional performance prediction value corresponding to each set of analysis data as the output, takes the actual value of the functional performance corresponding to each set of analysis data as the prediction target, and takes minimizing the sum of the prediction errors of all analysis data as the training target; The formula for the prediction error is expressed as: , where ε p is the prediction error, p is the group number of the feature vector corresponding to the analysis data, θ p is the functional performance prediction value corresponding to the p-th set of analysis data, μ p is the actual value of the functional performance corresponding to the p-th set of analysis data. Train the functional performance prediction model until the sum of the prediction errors reaches convergence and then stop training.

[0017] Technical effects and advantages of the present invention: The present invention is provided with an initial model design module and a design optimization module, which is beneficial to constructing an individual oral model, selecting the prosthetic model with the highest matching degree with the individual from the tooth type database as the initial model, obtaining the optimal adjustment parameters based on the improved optimization algorithm, adjusting and optimizing the prosthetic design parameters of the initial model according to the optimal adjustment parameters, and obtaining the optimized prosthetic model; By introducing a neural network into the optimization algorithm and integrating the trained neural network model into the fitness calculation of the simulated annealing algorithm, the non-linear and multi-modal data modeling capabilities of the neural network can be fully utilized to effectively capture complex data relationships; According to the parallel computing processing mechanism, the computing efficiency is improved, and complex data relationships are captured through deep learning technology to obtain the optimal adjustment parameters, thereby ensuring the practicability of the prosthetic model; Through precise data, intelligent optimization design of prosthetic materials is carried out, combined with the human tooth type database and the individual data of the patient, to optimize the personalized design of prosthetic materials, and then simultaneously ensure the aesthetics and practicability of the tooth shape. Description of the Drawings

[0018] Figure 1 This is the structural diagram of the 3D printing assisted design system for oral repair materials of the present invention. Detailed implementation manners

[0019] Next, the technical solutions in the present invention will be clearly and completely described in conjunction with the accompanying drawings in the present invention. In addition, the forms of each structure described in the following embodiments are merely examples. A 3D printing assisted design system for oral repair materials involved in the present invention is not limited to the structures described in the following embodiments. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0020] As Figure 1 shown, the present invention provides a 3D printing assisted design system for oral repair materials, including a data acquisition module, an initial model design module, a design optimization module, and a result output module; The data acquisition module is used to collect individual oral data. The individual oral data is the three-dimensional oral data of a patient. The individual is a patient who needs oral repair. The three-dimensional oral data includes, but is not limited to, three-dimensional oral imaging data, tooth morphology data, alveolar bone morphology data, and oral space data, etc. The three-dimensional oral imaging data is the three-dimensional internal oral image obtained by devices such as CBCT (cone beam computed tomography) and oral CT. The tooth morphology data includes, but is not limited to, the size, shape, position, crown, and root morphology of the teeth. The alveolar bone morphology data includes, but is not limited to, the density, height, and width of the alveolar bone, etc. The oral space data includes, but is not limited to, the size of the oral cavity space and the gaps between teeth, etc.; The initial model design module constructs an oral model of an individual based on the data collected by the data acquisition module, and selects the prosthetic model with the highest matching degree with the individual from the tooth type database as the initial model; the prosthetic is an artificially made structure for repairing or replacing missing or damaged teeth; The design optimization module is used to obtain the optimal adjustment parameters based on the improved optimization algorithm, adjust and optimize the prosthetic design parameters of the initial model according to the optimal adjustment parameters, obtain the optimized prosthetic model, and simultaneously obtain the predicted functional performance of the optimized prosthetic; its purpose is to introduce a neural network into the optimization algorithm, integrate the trained neural network model into the fitness calculation of the simulated annealing algorithm, and be able to give full play to the non-linear and multi-modal data modeling capabilities of the neural network, effectively capture complex data relationships; according to the parallel computing processing mechanism, improve the computing efficiency, and capture complex data relationships through deep learning technology to obtain the optimal adjustment parameters, so as to ensure the practicability of the prosthetic model; The result output model is used to output the optimized prosthesis model, which is used as the auxiliary design result for 3D printing. The operator can directly perform 3D printing on this design result, or use it as an auxiliary basis to take improvement measures that meet the 3D printing conditions, and then perform 3D printing.

[0021] In this embodiment, it should be specifically noted that the specific method for constructing the oral model of the individual is as follows: using computer-aided design software, an oral three-dimensional model of the individual is constructed according to the oral three-dimensional data. The computer-aided design software can be any one of three-dimensional modeling software such as 3Shape Dental System, Geomagic, and Materialise Magics; the oral three-dimensional model is a standard three-dimensional model with complete dental tissues in the individual's oral cavity in an ideal state; The acquisition method of the initial model is as follows: Based on the constructed oral three-dimensional model of the individual, key feature points of the oral three-dimensional model are extracted to form a key feature point set, , where DJ_i is the key feature point set of the i-th dental tissue, and TZ ij is the coordinate of the j-th key feature point in the key feature point set corresponding to the i-th dental tissue, j = 1, 2, 3,..., m; i = 1, 2, 3,..., n; n is the total number of dental tissues of the individual; m is the number of key feature points; the key feature points are the key landmark points for oral restoration, including but not limited to the dental margin, gingival margin, and interdental line, etc.; A key feature point set of the dental type database is established, , where SJ_a is the key feature point set of the a-th prosthesis model in the dental type database, and sj aj is the coordinate of the j-th key feature point in the key feature point set of the a-th prosthesis model; the acquisition method of this coordinate is: placing the prosthesis model in the position of the dental tissue to be repaired in the constructed oral three-dimensional model of the individual, and obtaining the corresponding coordinate at this dental tissue position, j = 1, 2, 3,..., m; Place each prosthetic model in the dental database into the individual's three-dimensional oral model, obtain the corresponding coordinates of each key feature point in the key feature point set of each prosthetic model, and use the key feature points in the individual's three-dimensional oral model and the same key feature points corresponding to the prosthetic model in the dental database as a group of key feature point pairs. Therefore, each prosthetic model has m groups of key feature point pairs and corresponds one-to-one with the individual's three-dimensional oral model; obtain the similarity of each group of key feature point pairs, and perform weighted average to obtain the average similarity of the key feature point pairs of each prosthetic model. Take the value of the average similarity as the value of the matching degree, and select the prosthetic model corresponding to the highest value of the matching degree as the initial model, that is, the prosthetic model corresponding to the maximum value of the average similarity of the key feature point pairs as the initial model; The coordinate point similarity of each group of key feature point pairs is calculated using the Euclidean distance: , where Dis_aj is the similarity of the jth group of key feature point pairs in the a-th prosthetic model, TZ ij _r is the coordinate of TZ ij on the r-th dimension, sj aj _r is the coordinate of sj aj on the r-th dimension, and R is the total dimension; perform weighted average on the similarities of m groups of key feature points to obtain the average similarity; , where PJ_a is the matching degree of the a-th prosthetic model; Therefore, by obtaining the key feature point set corresponding to the tooth to be repaired, the prosthetic model with the highest matching degree with the tooth to be repaired can be obtained.

[0022] In this embodiment, it should be specifically noted that the specific method for obtaining the optimal adjustment parameter is as follows: Obtain an adjustment parameter set, where the adjustment parameter set is the adjustment data of each key feature point of the prosthetic model; the adjustment parameter set can be expressed as: , where U f is the f-th adjustment parameter set, uj is the adjustment data of the j-th key feature point, and j = 1, 2, 3,..., m; the adjustment data can be obtained by combining historical adjustment data with the adjustment opinions of domain professionals. The historical adjustment data is the historical adjustment data of the same prosthetic model, and the adjustment data is the adjustment value of the key feature point; the adjustment parameter is the key feature point that needs to be adjusted; Based on the improved optimization algorithm, obtain the optimal adjustment parameter set from the adjustment parameter set. The adjustment parameters in the optimal adjustment parameter set are the optimal adjustment parameters. The optimization algorithm uses the simulated annealing algorithm and is improved by integrating a neural network model, including the following steps: Step S11: Represent the adjustment parameter set as U1, U2, U3,..., U f ,..., UF ; Step S12: Preset the initial temperature T max , the minimum temperature T min , the cooling coefficient δ, and the maximum number of iterations φ, and let the current temperature T = T max ; Step S13: Randomly set a feasible solution ρ, and the feasible solution ρ is the set of adjustment parameters; Step S14: Determine the fitness function; Step S15: Calculate the fitness corresponding to the feasible solution; using the feasible solution as the current point, perform a random perturbation within the neighborhood of the current point to obtain a new feasible solution ρ´, and calculate the fitness corresponding to the new feasible solution ρ´; Step S16: Calculate the fitness difference, and the expression of the fitness difference is: , where λ * is the fitness difference, λ ρ is the fitness of the feasible solution corresponding to the current point, and λ ρ´ is the fitness of the new feasible solution; If the fitness difference λ * > 0, then let ρ = ρ´, that is, assign the value of the new feasible solution ρ´ to the feasible solution ρ; if the fitness difference λ * ≤ 0, then calculate the probability P, and let ρ = ρ´ according to the probability P; the calculation formula of the probability P is expressed as: , where e is the natural constant; Step S17: Loop steps S15 to S16 until the number of loops reaches the maximum number of iterations, then the loop ends and enters step S18; Step S18: Cool down the current temperature to obtain a new current temperature, and the formula is expressed as: T new = T × δ, where T new is the new current temperature after cooling, and T is the current temperature before cooling; cool down the current temperature T in step S12, and use the cooled value as the new current temperature and represent it as T in the next iteration; Reset the number of iterations, and the formula is expressed as: φ new = φ × δ, where φ new is the reset maximum number of iterations; that is, use the reduced value of the maximum number of iterations as the new maximum number of iterations and represent it as φ in the next iteration; if the reduced maximum number of iterations is not an integer, then round up the reduced maximum number of iterations to make it an integer; Step S19: Loop steps S15 to S18; until the current temperature T < T minWhen it is satisfied, the loop ends; obtaining the adjustment parameter set corresponding to the feasible solution ρ is the optimal adjustment parameter set, and the adjustment parameters in the optimal adjustment parameter set are the optimal adjustment parameters.

[0023] In this embodiment, it should be specifically noted that the fitness function is expressed as: , where SP ρ is the functional performance gain of the adjustment parameter set corresponding to the feasible solution ρ; the acquisition method of the functional performance gain of the adjustment parameter set is: Input the adjusted key feature point data corresponding to the adjustment parameter set and the individual oral data into the functional performance prediction model to obtain the predicted functional performance. Input the key feature point data of the initial model and the human oral data into the functional performance prediction model to obtain the initial functional performance. Subtract the initial functional performance from the predicted functional performance to obtain the functional performance gain corresponding to the adjustment parameter set; The training method of the functional performance prediction model is: Take the adjusted key feature point data corresponding to each adjustment parameter set and the individual oral data as a set of analysis data. Pre-collect d sets of analysis data, and obtain the corresponding functional performance for each of the d sets of analysis data; d is an integer greater than 1. Convert the analysis data and the corresponding functional performance into a corresponding set of feature vectors; The functional performance can select any one or more of the marginal seal of the prosthesis, the correctness of the occlusal contact point position, the occlusal force distribution, the occlusal balance index, and the service life of the prosthesis as the evaluation index of the functional performance. The marginal seal is the degree of seal at the interface between the prosthesis and the tooth tissue, also known as the marginal adaptation. Good marginal seal can prevent oral fluids and bacteria from entering the interface between the prosthesis and the tooth, avoiding subsequent oral problems; The higher the correctness of the occlusal contact point position, the better the chewing function, and at the same time, it can also avoid damage to the prosthesis caused by uneven occlusal pressure; The occlusal force distribution can be measured by an instrument, such as the T-Scan Novus occlusal force measurement system, which can evaluate whether the occlusal force is evenly distributed on the teeth, which is related to chewing efficiency and the durability of the prosthesis; The occlusal balance index represents the balance state of the forces between teeth during occlusion. If the occlusal balance index is too low, it will cause the prosthesis to loosen or be damaged; Take each set of feature vectors as the input of the functional performance prediction model. The functional performance prediction model takes a functional performance prediction value corresponding to each set of analysis data as the output, takes the actual value of the functional performance corresponding to each set of analysis data as the prediction target, and takes minimizing the sum of the prediction errors of all analysis data as the training target; The formula for the prediction error is expressed as: , where ε p is the prediction error, p is the group number of the feature vector corresponding to the analysis data, θ p is the functional performance prediction value corresponding to the p-th set of analysis data, μ pUse the actual functional performance value corresponding to the p-th group of analysis data to train the functional performance prediction model until the sum of prediction errors converges and then stop training; The functional performance prediction model is specifically a neural network model, mainly including an input layer, a convolutional layer, a pooling layer, a fully connected layer, an activation function, a loss function, an optimizer, a regularizer, and an output layer.

[0024] In this embodiment, it should be specifically noted that the result output module optimizes and adjusts the initial model based on the optimal adjustment parameters, and outputs the optimized prosthesis model after obtaining it. The optimized prosthesis model can be directly used as the final prosthesis model for 3D printing, or the optimized prosthesis model can be used as the basic model, and appropriate adjustments can be made based on customer requirements and 3D printing requirements to obtain the final prosthesis model for 3D printing and then perform 3D printing operations.

[0025] Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

[0026] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A 3D printing assisted design system for oral restoration materials, characterized in that: It includes a data acquisition module, an initial model design module, a design optimization module, and a result output module; The data acquisition module is used to acquire individual oral data; The initial model design module constructs an oral model of an individual based on the data acquired by the data acquisition module, and selects the prosthetic model with the highest matching degree with the individual from the dental type database as the initial model; The design optimization module is used to obtain the optimal adjustment parameters based on the improved optimization algorithm, adjust and optimize the prosthetic design parameters of the initial model according to the optimal adjustment parameters, obtain the optimized prosthetic model, and at the same time obtain the predicted functional performance of the optimized prosthetic; The result output model is used to output the optimized prosthetic model. After obtaining the final prosthetic model, a 3D printing operation is performed.

2. The 3D printing assisted design system for oral repair materials according to claim 1, characterized in that: The individual oral data is the three-dimensional oral data of a patient. The individual is a patient who needs oral restoration. The three-dimensional oral data includes three-dimensional oral image data, tooth morphology data, alveolar bone morphology data, and oral space data.

3. The 3D printing assisted design system for oral restoration materials according to claim 2, wherein: The acquisition method of the initial model is as follows: Based on the constructed three-dimensional oral model of the individual, extract the key feature points of the three-dimensional oral model to form a key feature point set; establish a key feature point set of the dental type database; Place each prosthetic model in the dental type database in the three-dimensional oral model of the individual, obtain the corresponding coordinates of each key feature point in the key feature point set of each prosthetic model, and use the key feature points in the three-dimensional oral model of the individual and the same key feature points corresponding to the prosthetic model in the dental type database as a group of key feature point pairs. That is, each prosthetic model has m groups of key feature point pairs, and they correspond one by one to the three-dimensional oral model of the individual; obtain the similarity of each group of key feature point pairs, and perform weighted average to obtain the average similarity of the key feature point pairs of each prosthetic model. Take the value of the average similarity as the value of the matching degree, and select the prosthetic model corresponding to the highest value of the matching degree as the initial model, that is, the prosthetic model corresponding to the maximum value of the average similarity of the key feature point pairs as the initial model.

4. The 3D printing assisted design system for oral restoration materials according to claim 3, wherein: The key feature point set composed of the key feature points of the oral three-dimensional model is expressed as: , where DJ_i is the key feature point set of the i-th tooth, and TZ ij is the coordinate of the j-th key feature point in the key feature point set corresponding to the i-th tooth, j = 1, 2, 3, …, m; i = 1, 2, 3, …, n; n is the total number of teeth of the individual; m is the number of key feature points; The key feature point set of the tooth profile database is expressed as: , where SJ_a is the key feature point set of the a-th prosthesis model in the tooth profile database, and sj aj is the coordinate of the j-th key feature point in the key feature point set of the a-th prosthesis model; j = 1, 2, 3,..., m.

5. The 3D printing assisted design system for oral restoration materials according to claim 4, wherein: The coordinate point similarity of each group of key feature point pairs is calculated using the Euclidean distance: , where Dis_aj is the similarity of the j-th group of key feature point pairs in the a-th prosthesis model, and TZ ij _r is the coordinate of TZ ij on the r-th dimension, sj aj _r is the coordinate of sj aj on the r-th dimension, and R is the total dimension; the similarities of m groups of key feature points are weighted and averaged to obtain the average similarity; , and PJ_a is the matching degree of the a-th prosthesis model.

6. The 3D printing assisted design system for oral restoration materials according to claim 5, characterized in that: The specific acquisition method of the optimal adjustment parameters is as follows: Obtain an adjustment parameter set, where the adjustment parameter set is the adjustment data of each key feature point of the prosthesis model; the adjustment parameter set can be expressed as: , where U f is the f-th adjustment parameter set, uj is the adjustment data of the j-th key feature point, j = 1, 2, 3, …, m; obtain the optimal adjustment parameter set from the adjustment parameter set based on the improved simulated annealing algorithm, and the adjustment parameters in the optimal adjustment parameter set are the optimal adjustment parameters.

7. The 3D printing assisted design system for oral restoration materials according to claim 6, characterized in that: The improved simulated annealing algorithm includes the following steps: Step S11: Represent the adjustment parameter sets as U1, U2, U3, …, U f , …, U F ; Step S12: Preset the initial temperature T max , the minimum temperature T min , the cooling coefficient δ, and the maximum number of iterations φ, and set the current temperature T = T max ; Step S13: Randomly set a feasible solution ρ, and the feasible solution ρ is the adjustment parameter set; Step S14: Determine the fitness function; Step S15: Calculate the fitness corresponding to the feasible solution; use the feasible solution as the current point, perform random perturbation in the neighborhood of the current point to obtain a new feasible solution ρ´, and calculate the fitness corresponding to the new feasible solution ρ´; Step S16: Calculate the fitness difference, and the expression of the fitness difference is: , where λ * is the fitness difference, λ ρ is the fitness of the feasible solution corresponding to the current point, and λ ρ´ is the fitness of the new feasible solution; If the fitness difference λ * > 0, then let ρ = ρ´, that is, assign the value of the new feasible solution ρ´ to the feasible solution ρ; if the fitness difference λ * ≤ 0, then calculate the probability P, and let ρ = ρ´ according to the probability P; the calculation formula of the probability P is expressed as: , where e is the natural constant; Step S17: Loop steps S15 to S16 until the number of loops reaches the maximum number of iterations, the loop ends, and enter step S18; Step S18: Cool down the current temperature to obtain a new current temperature. Cool down the current temperature T in step S12, and use the cooled value as the new current temperature, and represent it as T in the next iteration; Reset the number of iterations, that is, use the value of the reduced maximum number of iterations as the new maximum number of iterations and represent it as φ in the next iteration; if the reduced maximum number of iterations is not an integer, round up the reduced maximum number of iterations to make it an integer. Step S19: Loop through steps S15 to S18; until the current temperature T < T min , the loop ends; the set of adjustment parameters corresponding to the feasible solution ρ is the optimal set of adjustment parameters, and the adjustment parameters in the optimal set of adjustment parameters are the optimal adjustment parameters.

8. The 3D printing assisted design system for oral restoration materials according to claim 7, characterized in that: In the step S18, the formula for obtaining the new current temperature by cooling the current temperature is expressed as: T new = T × δ, where T new is the new current temperature after cooling, and T is the current temperature before cooling; The formula for resetting the number of iterations is expressed as: φ new = φ × δ, where φ new is the maximum number of iterations after resetting.

9. The 3D printing assisted design system for oral restoration materials according to claim 8, characterized in that: The fitness function is expressed as: , where SP ρ is the functional performance gain of the adjustment parameter set corresponding to the feasible solution ρ. The method for obtaining the functional performance gain of the adjusted parameter set is as follows: Input the adjusted key feature point data corresponding to the adjusted parameter set and the individual oral data into the functional performance prediction model to obtain the predicted functional performance. Input the key feature point data of the initial model and the human oral data into the functional performance prediction model to obtain the initial functional performance. Subtract the initial functional performance from the predicted functional performance to obtain the functional performance gain corresponding to the adjusted parameter set.

10. A 3D printing assisted design system for oral restoration materials according to claim 9, characterized in that: The training method of the functional performance prediction model is as follows: Take the adjusted key feature point data corresponding to each adjusted parameter set and the individual oral data as a set of analysis data. Pre-collect d sets of analysis data and obtain the corresponding functional performance for each of the d sets of analysis data; d is an integer greater than 1. Convert the analysis data and the corresponding functional performance into a corresponding set of feature vectors. Taking each group of feature vectors as the input of the functional performance prediction model, the functional performance prediction model outputs a functional performance prediction value corresponding to each group of analysis data, takes the actual functional performance value corresponding to each group of analysis data as the prediction target, and takes minimizing the sum of prediction errors of all analysis data as the training target; the formula for the prediction error is expressed as: , where ε p is the prediction error, p is the group number of the feature vector corresponding to the analysis data, θ p is the functional performance prediction value corresponding to the p-th group of analysis data, μ p is the actual functional performance value corresponding to the p-th group of analysis data, and the functional performance prediction model is trained until the sum of prediction errors reaches convergence and then the training stops.

Images