Artificial organ production system utilizing big data infrastructure

Abstract

The present invention relates to an artificial organ production system that utilizes a big data infrastructure. The artificial organ production system of the present invention includes an input unit that accepts user requirements for the artificial organ to be produced as input values, a database unit that stores data necessary for artificial organ production, a control unit that produces the artificial organ by inputting the input values ​​into an artificial organ production algorithm learned based on the database unit, and a bioprinter that produces the artificial organ under the control of the control unit.

Description

[Technical Field]

[0001] This application claims the benefit of priority based on Korean Patent Application No. 10-2022-0169461 dated December 7, 2022, and all contents disclosed in the documents of that Korean patent application are incorporated herein by reference.

[0002] The present invention relates to an artificial organ production system, and more particularly to a system for producing artificial organs using a big data platform. [Background technology]

[0003] Bioprinting is a method for creating artificial organs, enabling the spatial structuring of living cells and other biological products, biomaterials, biochemicals, or biocompatible materials by sequential placement via layer-by-layer deposition under computer control to develop living tissues and organs for tissue engineering, regenerative medicine, pharmacokinetics, and more general biological research.

[0004] When using bioprinting to create artificial organs, in order to create artificial organs that meet the desired conditions, the creation is carried out after experiments to meet those conditions.

[0005] Artificial organs are used in a variety of situations, most notably for medical training. Artificial organs for training are required for various cases. However, there are limitations to the amount of research and experimentation required to create artificial organs that fit various cases. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Korean Patent No. 10-954953 Summary of the Invention [Problem to be solved by the invention]

[0007] Therefore, an object of the present invention is to provide an artificial organ production system that utilizes a big data infrastructure related to artificial organs, which can produce artificial organs that meet the user's requirements. [Means for solving the problem]

[0008] In order to achieve the above-mentioned object, the present invention provides an artificial organ production system that utilizes a big data infrastructure, including an input unit that accepts user requirements for the artificial organ to be produced as input values; a database unit that stores big data necessary for producing the artificial organ; a control unit that inputs the input values ​​into an artificial organ production algorithm learned based on the database unit and calculates process variables that can produce the artificial organ; and a bioprinter that produces the artificial organ using the process variables in accordance with the control of the control unit.

[0009] The input values ​​may include the hardness, tensile strength, toughness and ductility of the artificial organ. The input value may be a number.

[0010] The database section can store data on a plurality of raw materials required for producing an artificial organ, the weight of each raw material, pretreatment conditions, temperature, catalyst conditions, and reaction time.

[0011] The database unit may store mechanical property data including Young's Modulus, Ultimate Tensile Strength, Toughness, and Fracture Point according to the ratio of raw materials when multiple raw materials are synthesized.

[0012] The database section can store data on agarose, acrylamide, N,N'-methylenebis(acrylamide) (MBA), tetramethylethylenediamine (TEMED), deionized water, ammonium sulfate (APS), and temperature.

[0013] The bioprinter may be a 3D bioprinter. The control unit can generate an artificial organ modeling file (STL) including the process variables using the artificial organ production algorithm learned with the input values, and provide the generated artificial organ modeling file (STL) to the bioprinter to bioprint the artificial organ.

[0014] The artificial organ production algorithm may include a data inverse design model that generates virtual data required for machine learning using inverse design technology, and an artificial organ production model that performs machine learning using the virtual data generated by the data inverse design model, and then controls the bioprinter to produce a user-customized artificial organ using user requirements as input values.

[0015] The data reverse design model may include a proxy model that extracts features from raw materials through molecular fingerprints and predicts target physical properties (mechanical properties, electrocautery properties, and 3D printability), an inverse design deep learning model that selects process variables and raw materials that can synthesize the target physical properties, and a virtual data model that estimates the data distribution in the synthesized variable space and generates virtual data.

[0016] The artificial organ production model may include a machine learning model that derives the optimal design for each organ tissue based on the user's requirements (the patient's physiological (BMI, 3D tissue model), genetic, and demographic (age, gender) information), an artificial intelligence model that designs an automated synthesis process including process variables using machine learning utilizing the big data infrastructure, and a printer control model that applies the process variables to control the bioprinter to produce a user-customized artificial organ. [Effects of the Invention]

[0017] According to the present invention, it is possible to produce artificial organs that meet the requirements of users by utilizing a big data infrastructure related to artificial organs. In other words, by utilizing a big data infrastructure secured in the process of producing various artificial organs, it is possible to produce user-customized artificial organs by bioprinting, using the user's requirements as input values.

[0018] The input values ​​are numerical values ​​for the hardness, tensile strength, toughness, and ductility of the artificial organ, so by inputting the values, it is possible to customize and produce an artificial organ that matches the input values. This has the advantage that artificial organs with the same specifications can be mass-produced with a single input.

[0019] The artificial organ production system of the present invention utilizes a big data infrastructure, thereby reducing the time required for experimentation, manufacturing, and verification. [Brief explanation of the drawings]

[0020] [Figure 1] FIG. 1 is a block diagram showing an artificial organ production system utilizing a big data infrastructure according to an embodiment of the present invention. [Figure 2] FIG. 2 is a block diagram showing a database unit in FIG. 1. [Figure 3] 1 is a graph showing the stress-strain characteristics of monomers (agarose, acrylamide) and an additive (MBA). [Figure 4] 1 is a graph showing the stress-strain characteristics of monomers (agarose, acrylamide) and an additive (MBA). [Figure 5] 1 is a graph showing the stress-strain characteristics of monomers (agarose, acrylamide) and an additive (MBA). [Figure 6] 1 is a graph showing Young's Modulus, Ultimate Tensile Strength, and Toughness among the mechanical properties of additives. [Figure 7] 1 is a graph showing fracture points among the mechanical properties of additives. [Figure 8] 1 is a table showing the results of data processing of raw materials used in the production of artificial organs. [Figure 9] 1 is a graph showing mechanical properties according to the ratio of monomers (agarose, acrylamide) and additives (MBA). [Figure 10] FIG. 1 is a block diagram illustrating an algorithm for creating an artificial organ according to an embodiment of the present invention. [Figure 11] FIG. 11 is a detailed block diagram showing a model constituting the artificial organ production algorithm of FIG. 10. DETAILED DESCRIPTION OF THE INVENTION

[0021] It should be noted that in the following description, only the parts necessary for understanding the embodiments of the present invention will be described, and descriptions of other parts will be omitted to the extent that they do not deviate from the gist of the present invention.

[0022] The terms or phrases used in the following specification and claims should not be interpreted as being limited to their ordinary or dictionary meanings, but should be interpreted in a meaning and concept consistent with the technical idea of ​​the present invention, based on the principle that the inventor can appropriately define the concept of the term in order to best describe his invention. Therefore, it should be understood that the embodiments described in this specification and the configurations shown in the drawings are merely preferred embodiments of the present invention and do not represent all of the technical ideas of the present invention, and therefore, at the time of filing this application, there may be various equivalents and modifications that can be substituted therefor.

[0023] Hereinafter, embodiments of the present invention will be described in more detail with reference to the accompanying drawings. Fig. 1 is a block diagram showing an artificial organ production system utilizing a big data infrastructure according to an embodiment of the present invention, and Fig. 2 is a block diagram showing the database unit of Fig. 1.

[0024] Referring to Figures 1 and 2, the artificial organ production system (100) according to this embodiment is a system for producing user-customized artificial organs by bioprinting using a trained artificial organ production algorithm (50) that utilizes a big data platform.

[0025] The artificial organ construction system (100) according to this embodiment includes an input unit (10), a database unit (20), a control unit (30), and a bioprinter (40). The input unit (10) accepts user requirements for the artificial organ to be constructed as input values. The database unit (20) stores big data necessary for constructing the artificial organ. The control unit (30) inputs the input values ​​into an artificial organ construction algorithm (50) learned based on the database unit (20) and calculates process variables that will enable the construction of the artificial organ. The bioprinter (40) then constructs the artificial organ using the process variables under the control of the control unit (30).

[0026] Here, the input unit (10) accepts the requirements for the artificial organ to be created from the user as input values. The user's requirements can include physiological (BMI, 3D tissue model), genetic, and demographic (age, gender) information of the patient. The input values ​​can include information on the type and physical properties of the artificial organ. The information on physical properties includes the hardness, tensile strength, toughness, and ductility of the artificial organ, and the corresponding input values ​​are entered as numerical values.

[0027] Alternatively, information regarding physical properties may be input verbally. The control unit 30 can convert the verbally input information regarding physical properties into numerical values ​​using a big data infrastructure.

[0028] The input unit (10) may be a physical input device such as a keyboard or a mouse that directly accepts requirements from a user, or may be a communication unit that accepts requirements via a user terminal.

[0029] The database unit (20) stores big data used in the production of various artificial organs. Such database unit (20) stores data on the raw materials required for the production of artificial organs, the weight of each raw material, pretreatment conditions, temperature, catalyst conditions, and reaction time. For example, the database unit (20) Data required for producing artificial organs, such as agarose, acrylamide, N,N'-methylenebis(acrylamide) (MBA), tetramethylethylenediamine (TEMED), deionized water, ammonium sulfate (APS), and temperature, can be stored. These are examples of materials according to one embodiment, but are not limited to these.

[0030] As shown in Figures 3 to 9, when synthesizing raw materials used in creating artificial organs, mechanical property data (Young's Modulus, Ultimate Tensile Strength, Toughness, Fracture point) is collected according to the ratio of monomers (agarose, acrylamide) or additives (MBA). In Figure 8, "Agar" represents agarose, and "AM" represents acrylamide.

[0031] When synthesizing the raw materials, it can be seen that as the amount of monomer and additives increases, toughness and strength increase.

[0032] Therefore, when synthesizing raw materials to create artificial organs, the synthesis conditions (process variables) of the raw materials that meet the user's requirements can be derived through machine learning based on the mechanical property data corresponding to the amounts of the monomers and additives.

[0033] Here, Figures 3 to 5 are graphs showing the stress-strain characteristics of monomers (agarose, acrylamide) and additive (MBA). Figure 3 shows the stress-strain characteristics of agarose. Figure 4 shows the stress-strain characteristics of acrylamide. And Figure 5 shows the stress-strain characteristics of additive (MBA).

[0034] The database unit (20) can store an artificial organ modeling file (STL) generated using an artificial organ production algorithm (50) trained using user requirements as input values.

[0035] The bioprinter (40) produces an artificial organ using the artificial organ production algorithm (50) learned under the control of the control unit (30). The bioprinter (40) is a 3D bioprinter. That is, the bioprinter (40) accepts an artificial organ modeling file (STL) created using the learned artificial organ production algorithm (50) and produces an artificial organ using 3D bioprinting.

[0036] Here, artificial organs are those produced by 3D bioprinting, and may include, but are not limited to, artificial skin, artificial blood vessels (for bleeding and hemostasis), artificial peritoneum, and the like.

[0037] 3D bioprinting involves scanning a target structure in three dimensions to obtain an image, and then using cells and bioink to create a three-dimensional structure from the scanned image. Common 3D printing technologies include extrusion-based, inkjet-based, and laser-based processes, of which extrusion-based processes are the most widely used for bioprinting, which creates three-dimensional structures together with cells. Since the core of bioprinting is the inclusion of living cells, bioink, a material that mixes with cells to form three-dimensional structures, is the most important part of 3D bioprinting.

[0038] On the other hand, when the artificial organ according to this embodiment is used for medical education, it is not necessary that all of the raw materials constituting the bioink are biocompatible substances.

[0039] The control unit 30 is a microprocessor that performs overall control of the artificial organ production system 100. The control unit 30 controls the production of a user-customized artificial organ through bioprinting using an artificial organ production algorithm 50 trained using a big data platform. That is, the control unit 30 generates an artificial organ modeling file (STL) using the trained artificial organ production algorithm 50, using the user's requirements for the artificial organ to be produced as input values. The control unit 30 then provides the artificial organ modeling file (STL) to the bioprinter 40, thereby producing the user-customized artificial organ through bioprinting.

[0040] The artificial organ production algorithm (50) of this embodiment utilizes a big data platform to analyze trends, selects an algorithm with a high prediction rate, and reverse-designs raw materials. Furthermore, it can provide customized artificial organs to users using machine learning to minimize trial and error based on their diverse requirements, such as gender and age.

[0041] There are various algorithms for machine learning, and an appropriate algorithm may be selected depending on the data. Machine learning algorithms include supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning.

[0042] In supervised learning, a computer learns from data that contains labels. Here, a label means the correct value for the data, and the computer determines which correct value a new input resembles. In other words, supervised learning is a method of learning data using data with correct answers.

[0043] Such supervised learning algorithms include, but are not limited to, k-Nearest Neighbors, Linear Regression, Logistic Regression, Gaussian Process Regression, Support Vector Machines (SVMs), Decision Trees and Random Forests, and Neural Networks.

[0044] Unsupervised learning is a method in which a computer learns from data that does not contain labels. In other words, since the input data does not contain the correct answer, unsupervised learning is a method in which the computer itself groups similar data together and predicts the results for new data.

[0045] Such unsupervised learning algorithms include, but are not limited to, clustering, visualization and dimensionality reduction, and association rule learning. Clustering algorithms include k-means, hierarchical cluster analysis (HCA), and expectation maximization. Visualization and dimensionality reduction algorithms include principal component analysis (PCA), kernel PCA, locally linear embedding (LLE), and t-distributed stochastic neighbor embedding (t-SNE). Association rule learning algorithms may include Apriori and Eclat.

[0046] Semi-supervised learning is a hybrid of supervised and unsupervised learning, where learning is performed using data that contains labels. In other words, semi-supervised learning aims to improve the performance of supervised learning, but when there are limited resources to label data, it is a learning method that uses labels only for part of the data.

[0047] In addition, reinforcement learning involves taking a certain action in a given environment and learning through trial and error while receiving feedback. Reinforcement learning involves having the program make a decision without providing explicit data to the machine learning model, and learning occurs by receiving feedback on whether that decision was correct or incorrect. Here, the feedback is called a reward, and the model learns to make decisions that will result in a positive reward for the next decision and to avoid decisions that will result in a negative reward as much as possible.

[0048] Such an artificial organ construction algorithm (50) includes a data inverse design model (60) and an artificial organ construction model (70), as shown in Figures 10 and 11. Here, Figure 10 is a block diagram showing the artificial organ construction algorithm (50) according to an embodiment of the present invention. Figure 11 is a detailed block diagram showing the models constituting the artificial organ construction algorithm (50) of Figure 10.

[0049] The data inverse design model (60) generates data necessary for learning the artificial organ production model (70) using inverse design technology.

[0050] The artificial organ production model (70) then performs machine learning using the data generated by the data inverse design model (60), and then controls the bioprinter (40) to produce a user-customized artificial organ using the user's requirements as input values.

[0051] The data reverse design model (60) will be explained as follows. Generally, the physical properties of human organs vary greatly from patient to patient. Therefore, in order to attract many customers, it is necessary to secure raw materials for producing artificial organs with a variety of physical properties. However, due to the diversity and complexity of raw materials, directly synthesizing all raw materials with the target physical properties is time-consuming and cost-intensive. Therefore, in this embodiment, a reverse design technique for raw materials through machine learning based on a data-driven reverse design model (60) is used.

[0052] The data-based inverse design model (60) generated virtual data for training the artificial organ construction model (70) as follows: In other words, to address the lack of experimental data and improve the performance of the inverse design model, the data-based inverse design model (60) used a technique to generate new data points (virtual data) on demand.

[0053] Such a data-based reverse design model (60) can include a surrogate model (61), a reverse design deep learning model (63), and a virtual data model (65). In this case, the data-based reverse design model (60) was developed with the goal of achieving 90% accuracy in predicting target physical properties and 80% accuracy in estimating reverse design process variables.

[0054] The surrogate model (61) extracts features from raw materials through molecular fingerprinting and predicts target properties (mechanical properties, electrocautery properties, and 3D printability).

[0055] The inverse design deep learning model (63) selects process variables and raw materials that can synthesize target properties.

[0056] The virtual data model (65) then generates virtual data by estimating the data distribution in a wide synthetic variable space.

[0057] The artificial organ production model (70) includes a machine learning model (71), an artificial intelligence model (73), and a printer control model (75).

[0058] The optimal design parameters for artificial organs vary greatly depending on the user (patient)'s physiological, age, gender, and other conditions, but there has been no method to predict these and provide customized artificial organs. Even if the optimal target design parameters are determined, the experimental data used for machine learning in the data-based inverse design model (60) has limitations in estimating the entire variable space.

[0059] Therefore, in this embodiment, the artificial organ construction model (70) estimates the design parameters of the target organ for various patient conditions and constructs an efficient process for systematically modifying the manufacturing process and material composition ratios based on limited data.

[0060] The artificial organ fabrication model (70) was developed with the goal of 80% accuracy in predicting the patient's optimized design variables and an average of five iterations of modifying the fabrication process and composition to obtain the target variables.

[0061] Machine learning models (71) derive optimal designs for specific organ tissues based on user requirements, such as physiological (BMI, 3D tissue models, etc.), genetic, and demographic (age, gender, etc.) information for various patients.

[0062] The artificial intelligence model (73) can design an automated synthesis process including process variables using various machine learning techniques based on big data. The artificial intelligence model (73) can select and use an appropriate machine learning algorithm depending on the data contained in the big data. For example, the artificial intelligence model (73) can design an automated synthesis process including process variables optimized based on, but not limited to, Gaussian processes and Bayesian methods.

[0063] The printer control model (75) then applies the process variables designed by the artificial intelligence model (73) to the control of the 3D bioprinter to create a user-customized artificial organ.

[0064] The process of creating an artificial organ by inputting input values ​​into the artificial organ creation algorithm (50) is explained as follows.

[0065] In addition, artificial organs are produced using 3D bioprinting by inputting artificial organ modeling files (STL) into a 3D bioprinter.

[0066] According to this embodiment, it is possible to fabricate artificial organs that meet the requirements of users by utilizing a big data infrastructure related to artificial organs. In other words, it is possible to fabricate user-customized artificial organs by bioprinting using the big data infrastructure secured in the process of fabricating various artificial organs as input values ​​for the requirements of users.

[0067] The input values ​​are numerical values ​​for the hardness, tensile strength, toughness, and ductility of the artificial organ, so it is possible to input numerical values ​​and create an artificial organ that matches them, which has the advantage of making mass production possible.

[0068] Furthermore, the artificial organ production system (100) according to this embodiment utilizes a big data infrastructure, thereby reducing the time required for experiments, manufacturing, and verification.

[0069] Meanwhile, the embodiments disclosed in this specification and the drawings are merely specific examples to aid in understanding and are not intended to limit the scope of the present invention. It is obvious to those skilled in the art to which the present invention pertains that other modifications based on the technical concept of the present invention can be implemented in addition to the embodiments disclosed herein. [Explanation of symbols]

[0070] 10: Input section 20: Database section 30: Control unit 40: Bioprinter 50: Artificial organ creation algorithm 60: Data inverse design model 61: Surrogate model 63: Reverse design deep learning model 65: Virtual data model 70: Artificial organ production model 71: Machine learning model 73: Artificial intelligence model 75: Printer control model 100: Artificial organ production system

Claims

1. an input unit that receives user requirements for the artificial organ to be produced as input values; A database section that stores the big data necessary for creating artificial organs, and a control unit that inputs the input values ​​into an artificial organ construction algorithm learned based on the database unit, and calculates process variables that can be used to construct an artificial organ; a bioprinter that produces an artificial organ using the process variables under the control of the control unit; An artificial organ production system utilizing a big data platform, comprising:

2. 2. The artificial organ production system utilizing a big data platform according to claim 1, wherein the input values ​​include hardness, tensile strength, toughness, and ductility of the artificial organ.

3. 2. The artificial organ production system utilizing the big data infrastructure according to claim 1, wherein the input values ​​are numerical values.

4. The artificial organ production system utilizing the big data platform described in claim 1, characterized in that the database unit stores data regarding the multiple raw materials required for artificial organ production, the weight of each raw material, pretreatment conditions, temperature, catalyst conditions, and reaction time.

5. The artificial organ production system utilizing a big data infrastructure according to claim 4, wherein the database unit stores mechanical property data including Young's Modulus, Ultimate Tensile Strength, Toughness, and Fracture point according to the ratio of raw materials when synthesizing multiple raw materials.

6. The artificial organ production system utilizing a big data infrastructure according to claim 1, wherein the database unit stores data on agarose, acrylamide, N,N'-methylenebis(acrylamide) (MBA), tetramethylethylenediamine (TEMED), deionized water, ammonium sulfate (APS), and temperature.

7. 2. The artificial organ production system utilizing a big data platform according to claim 1, wherein the bioprinter is a 3D bioprinter.

8. The artificial organ production system utilizing the big data infrastructure described in claim 1, characterized in that the control unit generates an artificial organ modeling file (STL) including the process variables using the artificial organ production algorithm learned with the input values, and provides the generated artificial organ modeling file (STL) to the bioprinter to bioprint the artificial organ.

9. The artificial organ creation algorithm A data reverse design model that uses reverse design technology to generate virtual data necessary for machine learning; an artificial organ production model that performs machine learning using the virtual data generated by the data reverse design model, and then controls the bioprinter to produce a user-customized artificial organ using user requirements as input values; 2. An artificial organ production system utilizing the big data platform of claim 1, comprising:

10. The data reverse design model is A surrogate model that extracts features from raw materials through molecular fingerprinting and predicts target properties (mechanical properties, electrocautery properties, and 3D printability) A reverse design deep learning model that selects process variables and raw materials that can synthesize target properties, a virtual data model that generates virtual data by estimating the data distribution in a synthetic variable space; 10. An artificial organ production system utilizing the big data infrastructure according to claim 9, comprising:

11. The artificial organ production model comprises: A machine learning model that derives the optimal design for each organ tissue based on the user's requirements (patient's physiological (BMI, 3D tissue model), genetic, and demographic (age, gender) information). an artificial intelligence model that designs an automated synthetic process including process variables through machine learning based on the big data; a printer control model that applies the process variables to control the bioprinter to create a user-customized artificial organ; 11. An artificial organ production system utilizing the big data platform according to claim 10, comprising:

Images