A method for predicting performance of a prosthetic valve stent based on digital twinning

Abstract

This invention provides a method for predicting the performance of artificial valve stents based on digital twins, comprising the following steps: acquiring the material parameters and geometric characteristics of the valve stent; real-time monitoring of service data of the valve stent under simulated in vitro service; preprocessing the service data; establishing a fatigue mechanical model; establishing model coupling for the fatigue mechanical model, performing reduced-order model analysis, and establishing a dynamically updated digital twin model; using the digital twin model to simulate and calculate the output data of the valve stent, obtaining simulation data of the valve stent digital twin model; comparing the simulation data of the digital twin model with the service data to verify the accuracy of the digital twin model; establishing a remaining life prediction module; establishing a life assessment module, taking the remaining life output from the remaining life prediction model as input, comparing the remaining life with a preset value, and issuing an early warning when the preset value is exceeded. This invention improves the accuracy of predicting the mechanical performance of valve stents.

Description

Technical Field

[0001] This invention belongs to the field of interventional medical device technology, specifically relating to a method for predicting the performance of artificial valve stents based on digital twins. Background Technology

[0002] Common heart valve diseases are caused by functional or structural abnormalities of one or more valve structures (including leaflets, annulus, chordae tendineae, or papillary muscles) due to inflammation, myxoid degeneration, degenerative changes, congenital malformations, ischemic necrosis, trauma, etc., leading to valvular stenosis or regurgitation. Heart valve diseases are particularly common in the elderly.

[0003] Treatment methods for valvular heart disease mainly include drug therapy and surgical valve replacement. Surgical valve replacement can fundamentally resolve the structural and functional abnormalities of the diseased valve, making it the most trusted and preferred treatment method for both patients and doctors with valvular heart disease. Subsequently, the advent of transcatheter aortic valve replacement (TAVR) brought new prospects for artificial valve stents. Compared to surgical procedures, TAVR has advantages such as minimally invasiveness, rapid postoperative recovery, and high efficiency. The valve stents used in TAVR are mainly divided into self-expanding and balloon-expandable types. Both types of stents have their own application prospects and are not inherently superior or inferior. After more than a decade of development, TAVR has become a relatively mature technology for treating valvular heart disease.

[0004] Currently, experimental and simulation methods are commonly used in the design and research of artificial valve stents. However, in practice, experiments require significant time and financial resources, and commonly used experimental studies are still in the in vitro research stage, awaiting further clarification through animal and clinical trials. While the finite element method (FEM) has gradually gained widespread application in the design of artificial valve stents with the advancement of computer technology, it is often limited to static simulations of the overall service status of the stent within the aortic valve. This method is inefficient and cannot verify the accuracy of calculations in real time, failing to capture the real-time service status of the stent. How to monitor, predict, and optimize the service process of artificial valve stents in the aortic valve in real time remains a challenging problem to be solved. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method for predicting the performance of artificial valve stents based on digital twins. By using digital twin technology, the physical simulation and modeling of the valve stent entity are performed in digital space, improving the accuracy of predicting the mechanical performance of the valve stent. This is beneficial for better determining the service status and remaining lifespan of the valve stent and has important guiding significance for the design of valve stents.

[0006] This invention is achieved through the following technical solution:

[0007] A method for predicting the performance of artificial valve stents based on digital twins includes the following steps:

[0008] Step S1. Obtain the material parameters and geometric features of the valve stent;

[0009] Step S2. Monitor the service data of the valve stent in real time during in vitro simulated service;

[0010] Step S3. Perform maximum normalization on the service data detected in step S2;

[0011] Step S4. Based on the geometric features and material parameters of the valve stent in Step S1, establish a fatigue mechanical model;

[0012] Step S5. Couple the service data obtained in step S2 with the fatigue mechanics model established in step S4, perform reduced-order model analysis, and establish a dynamically updated digital twin model.

[0013] Step S6. Import the normalized data from step S3 into the digital twin model established in step S5, and use the digital twin model to simulate and calculate the output data of the valve stent to obtain the simulation data of the valve stent digital twin model.

[0014] Step S7. Compare and digitally fuse the simulation data of the digital twin model in step S6 with the corresponding service data to verify the accuracy of the digital twin model. If the residual sequence is white noise with zero mean, the corresponding valve stent digital twin model is accepted. Otherwise, the parameters of the valve stent digital twin model are adjusted and corrected according to the criterion function to obtain a real-time synchronized valve stent digital twin model.

[0015] Step S8. Establish a remaining lifetime prediction model based on the modified digital twin model in step S7.

[0016] In the above scheme, in step S1, the geometric parameters of the valve stent include the outer diameter, thickness and ellipticity of the valve stent.

[0017] In the above scheme, the material parameters of the valve stent in step S1 include the density, Poisson's ratio and elastic modulus of the valve stent.

[0018] In the above scheme, the service data in step S2 includes blood flow pressure difference, valve load, external temperature, strain state, damage area, and damage depth.

[0019] In the above scheme, step S4, where the fatigue mechanics model is established using Abaqus software, includes the following steps:

[0020] Step S401. Establish a geometric model of the valve stent and aortic sinus based on the geometric features of the valve stent in step S1;

[0021] Step S402. Define the material properties of the valve stent and the aortic sinus according to the material parameters of the valve stent in step S1;

[0022] Step S403. Set the loads and boundary conditions in the Abaqus finite element analysis software;

[0023] Step S404. Establish an expansion mechanical model based on the loads and boundary conditions in step S403;

[0024] Step S405. Based on the expansion mechanical model in step S404, apply fatigue load to establish a fatigue mechanical model;

[0025] Step S406. Calculate the maximum strain of each node in the fatigue mechanical model obtained in step S405, and calculate the working life of the valve stent based on this.

[0026] In the above scheme, in step S5, the reduced-order model analysis adopts the data fitting method of support vector regression.

[0027] In the above scheme, in step S6, the data after normalization in step S3 includes blood flow pressure difference, valve load, external temperature, valve stent strain, damage area, and damage depth.

[0028] In the above scheme, the specific steps of the remaining lifetime prediction module in step S8 are as follows:

[0029] Step S801:: Based on the data from the modified digital twin model, a lifetime prediction model is constructed using the SVR support vector regression mechanism;

[0030] Step S802: Establish training and validation sets based on historical service data of valve stents. Use random sampling to select 90% of the data as the training set and the remaining 10% as the validation set. Set the random seed arbitrarily. Train the life prediction model based on the training set data.

[0031] Step S803: Use grid search to obtain the optimal parameters of the lifetime prediction model, and compare the accuracy of the lifetime prediction model with the preset value to determine whether the accuracy meets the standard. The accuracy calculation formula is as follows:

[0032]

[0033] Where MAPE is the mean absolute percentage error, and n is the number of valve stent samples. This represents the actual value of the remaining lifespan. This is a predicted value for the remaining lifespan;

[0034] Step S804: Use real-time service data as input to the life prediction model and the remaining life as output of the remaining life prediction model.

[0035] The above scheme also includes step S9; step S9 establishes a life assessment module, takes the remaining life output from the remaining life prediction model in step S8 as the input of the life assessment module, compares the remaining life with a preset value, and issues an early warning when the preset value is exceeded.

[0036] The above scheme also includes step S10; step S10 is to establish a human-computer interaction module based on the simulation data in step S6; the human-computer interaction module is used to display the deformation generated during the service of the valve stent.

[0037] Compared with the prior art, the beneficial effects of the present invention are:

[0038] By using digital twin technology to physically simulate and model the physical valve stent in digital space, the accuracy of predicting the mechanical properties of the valve stent can be improved. This is beneficial for better judging the service status and remaining lifespan of the valve stent and has important guiding significance for the design of valve stents. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of a digital twin-based artificial valve stent performance prediction method according to an embodiment of the present invention.

[0040] Figure 2 This is a schematic diagram of a digital twin according to an embodiment of the present invention;

[0041] Figure 3 This is a schematic diagram of data preprocessing according to an embodiment of the present invention;

[0042] Figure 4 This is a schematic diagram of a model reduction method according to an embodiment of the present invention. Detailed Implementation

[0043] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0044] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "front," "rear," "left," "right," "upper," "lower," "axial," "radial," "vertical," "horizontal," "inner," and "outer," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0045] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0046] Figure 1 and 2 The image shows a preferred embodiment of the digital twin-based artificial valve stent performance prediction method, which includes the following steps:

[0047] Step S1. Obtain the material parameters and geometric features of the valve stent;

[0048] According to this embodiment, preferably, the geometric parameters of the valve stent include the outer diameter, thickness, and ellipticity of the valve stent.

[0049] According to this embodiment, preferably, the material parameters of the valve stent include the valve stent's density, Poisson's ratio, and elastic modulus.

[0050] Step S2. Real-time monitoring of the valve stent's service data during simulated in vitro service is achieved using sensors, strain gauges, and image acquisition equipment. Sensors are used to collect the blood flow pressure differential and temperature experienced by the valve stent; strain gauges are used to collect the strain generated by the valve stent under load; and image acquisition equipment is used to collect the damage area and depth of the valve stent.

[0051] According to this embodiment, preferably, in step S2, the service data includes, but is not limited to, strain state, damage area, damage depth, and life cycle.

[0052] Step S3. Perform maximum normalization processing on the service data detected in step S2, including eliminating delay and noise errors, such as... Figure 3 As shown, since the data changes in real time, the method chosen is zero-mean standardization, and the transformation function is (X-Mean) / (Standard deviation), where Mean is the mean of all sample data and Standard deviation is the standard deviation of all sample data.

[0053] Step S4. Based on the geometric features and material parameters of the valve stent in Step S1, establish a fatigue mechanical model;

[0054] Step S5. Couple the service data obtained in step S2 with the fatigue mechanics model established in step S4, and perform a reduced-order model analysis, such as... Figure 4 As shown, a dynamically updated digital twin model is established. Under the premise that the digital twin model has the same functions as the physical entity, one valve stent entity can correspond to multiple digital twins. The aortic valve has multiple parameters, such as valve annulus diameter and valve annulus ellipticity. By implanting digital twins into aortic valves with different parameters, the virtual-real mapping correspondence can be realized, which can be applied to more scenarios.

[0055] Step S6. Import the normalized data from step S3 into the digital twin model established in step S5, and use the digital twin model to simulate and calculate the output data of the valve stent to obtain the simulation data of the valve stent digital twin model.

[0056] Step S7. Compare and digitally fuse the simulation data of the digital twin model in step S6 with the corresponding service data to verify the accuracy of the digital twin model. If the residual sequence is white noise with zero mean, the corresponding valve stent digital twin model is accepted. Otherwise, the parameters of the valve stent digital twin model are adjusted and corrected according to the criterion function to obtain the corrected valve stent digital twin model in real time.

[0057] Step S8. Establish a remaining lifetime prediction model based on the modified digital twin model in step S7.

[0058] According to this embodiment, preferably, in step S4, the fatigue mechanics model uses Abaqus software to write a numerical calculation program, and establishes the model structure and parameters using machine learning algorithms based on service data, including the following steps:

[0059] Step S401. Establish a geometric model of the valve stent and aortic sinus based on the geometric features of the valve stent in step S1;

[0060] Step S402. Define the material properties of the valve stent and the aortic sinus according to the material parameters of the valve stent in step S1;

[0061] Step S403. Set the loads and boundary conditions in the Abaqus finite element analysis software;

[0062] Step S404. Establish an expansion mechanical model based on the loads and boundary conditions in step S403;

[0063] Step S405. Based on the expansion mechanical model in step S404, apply fatigue load to establish a fatigue mechanical model;

[0064] Step S406. Calculate the maximum strain of each node in the fatigue mechanical model obtained in step S405, and calculate the working life of the valve stent based on this.

[0065] According to this embodiment, preferably, in step S5, the reduced-order model analysis adopts the data fitting method of support vector regression.

[0066] According to this embodiment, preferably, in step S6, the data after normalization in step S3 includes blood flow pressure difference, valve load, external temperature, valve stent strain, damage area, and damage depth.

[0067] According to this embodiment, preferably, the remaining lifetime prediction module in step S8 has the following specific steps:

[0068] Step S801: Construct a lifetime prediction model using the SVR support vector regression mechanism;

[0069] Step S802: Establish training and validation sets based on historical service data of valve stents. Use random sampling to select 90% of the data as the training set and the remaining 10% as the validation set. Set the random seed arbitrarily. Train the life prediction model based on the training set data.

[0070] Step S803: Use grid search to obtain the optimal parameters of the lifetime prediction model, and compare the accuracy of the lifetime prediction model with the preset value to determine whether the accuracy meets the standard. The accuracy calculation formula is as follows:

[0071]

[0072] Where MAPE is the mean absolute percentage error, and n is the number of valve stent samples. This represents the actual value of the remaining lifespan. This is a predicted value for the remaining lifespan;

[0073] Step S804: Use real-time service data as input to the life prediction model and the remaining life as output to the remaining life prediction model. The service data includes, but is not limited to, strain state, damage area, and life cycle.

[0074] According to this embodiment, preferably, it also includes step S9; step S9 establishes a life assessment module, sets a preset value, takes the remaining life output from the remaining life prediction model in step S8 as the input of the life assessment module, compares the remaining life with the preset value, and when it exceeds the preset value, it indicates that the valve stent is about to reach its life limit and issues a warning.

[0075] According to this embodiment, preferably, it further includes step S10; step S10 is to establish a human-computer interaction module based on the simulation data in step S6; the human-computer interaction module is used to display the deformation generated during the service of the valve stent. The human-computer interaction module allows patients to see the deformation and warning prompts generated during the service of the valve stent through the display screen, thereby clarifying their physical condition, and improving the real-time performance of the prediction through data interaction feedback.

[0076] This invention utilizes software and multi-platform analysis to establish a fatigue mechanical model of the valve stent and aortic sinus, and performs fusion analysis on simulation data and service data to achieve real-time monitoring of the service status of the valve stent.

[0077] Establishing a real-time remaining life prediction model can accurately predict the remaining life of valve stents, which has important guiding significance for the selection of valve stents in clinical practice and provides reference information for manufacturers to design and manufacture valve stents with better performance.

[0078] In the context of digital twin technology, a remaining life prediction model based on SVR (Support Vector Regression Machine) was established. By combining historical and real-time service data of valve stents, the accuracy of valve stent life prediction was improved.

[0079] This invention enables real-time monitoring and analysis prediction through a digital twin model, improving the real-time and predictive nature of valve stent management during service.

[0080] This invention constructs a digital twin of a valve stent and performs simulation corrections on it under service conditions. Then, the corrected digital twin of the valve stent is used to perform real-time detection of the valve stent, which helps to better determine the service status and remaining lifespan of the valve stent.

[0081] This invention utilizes digital twin technology to physically simulate and model the physical valve stent in digital space, describing the changes of the valve stent under real-world conditions at multiple scales. This enables real-time monitoring, prediction, and optimization of the valve stent, improving the accuracy of predicting its mechanical properties and providing important guidance for valve stent design.

[0082] It should be understood that although this specification is described according to various embodiments, not every embodiment contains only one independent technical solution. This way of describing the specification is only for clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0083] The detailed descriptions listed above are merely specific illustrations of feasible embodiments of the present invention and are not intended to limit the scope of protection of the present invention. All equivalent embodiments or modifications made without departing from the spirit of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting the performance of artificial valve stents based on digital twins, characterized in that, Includes the following steps: Step S1. Obtain the material parameters and geometric features of the valve stent; Step S2. Monitor the service data of the valve stent in real time during in vitro simulated service; Step S3. Perform maximum normalization on the service data detected in step S2; Step S4. Based on the geometric features and material parameters of the valve stent in Step S1, establish a fatigue mechanical model; Step S5. Couple the service data obtained in step S2 with the fatigue mechanics model established in step S4, perform reduced-order model analysis, and establish a dynamically updated digital twin model. Step S6. Import the normalized data from step S3 into the digital twin model established in step S5, and use the digital twin model to simulate and calculate the output data of the valve stent to obtain the simulation data of the valve stent digital twin model. Step S7. Compare and digitally fuse the simulation data of the digital twin model from step S6 with the corresponding service data to verify the accuracy of the digital twin model. If the residual sequence is white noise with zero mean, the corresponding valve stent digital twin model is accepted. Otherwise, adjust and correct the parameters of the valve stent digital twin model according to the criterion function to obtain the corrected valve stent digital twin model. Step S8. Establish a remaining lifetime prediction model based on the modified digital twin model in step S7; In step S4, the fatigue mechanics model is established using Abaqus software, including the following steps: Step S401. Establish a geometric model of the valve stent and aortic sinus based on the geometric features of the valve stent in step S1; Step S402. Define the material properties of the valve stent and the aortic sinus according to the material parameters of the valve stent in step S1; Step S403. Set the loads and boundary conditions in the Abaqus finite element analysis software; Step S404. Establish an expansion mechanical model based on the loads and boundary conditions in step S403; Step S405. Based on the expansion mechanical model in step S404, apply fatigue load to establish a fatigue mechanical model; Step S406. Calculate the maximum strain of each node in the fatigue mechanical model obtained in step S405, and calculate the working life of the valve stent based on this. In step S5, the reduced-order model analysis uses the data fitting method of support vector regression. The specific steps of the remaining lifetime prediction model in step S8 are as follows: Step S801: Based on the data from the modified digital twin model, construct a lifetime prediction model using the SVR support vector regression mechanism; Step S802: Establish training and validation sets based on historical service data of valve stents, using random sampling. 90% of the data is randomly used as the training set, and the remaining 10% is used as the validation set. The random seed can be set arbitrarily, and the lifespan prediction model is trained based on the training set data. Step S803: Use grid search to obtain the optimal parameters of the lifetime prediction model, and compare the accuracy of the lifetime prediction model with the preset value to determine whether the accuracy meets the standard. The accuracy calculation formula is as follows: ； Where MAPE is the mean absolute percentage error, and n is the number of valve stent samples. This represents the actual value of the remaining lifespan. This is a predicted value for the remaining lifespan; Step S804: Use real-time service data as input to the life prediction model and the remaining life as output of the remaining life prediction model.

2. The method for predicting the performance of artificial valve stents based on digital twins according to claim 1, characterized in that, In step S1, the geometric parameters of the valve stent include the outer diameter, thickness, and ellipticity of the valve stent.

3. The method for predicting the performance of artificial valve stents based on digital twins according to claim 1, characterized in that, In step S1, the material parameters of the valve stent include the valve stent's density, Poisson's ratio, and elastic modulus.

4. The method for predicting the performance of artificial valve stents based on digital twins according to claim 1, characterized in that, In step S2, the service data includes blood flow pressure difference, valve load, external temperature, strain state, damage area, and damage depth.

5. The method for predicting the performance of artificial valve stents based on digital twins according to claim 1, characterized in that, In step S6, the data after normalization in step S3 includes blood flow pressure difference, valve load, external temperature, strain of valve stent, damage area, and damage depth.

6. The method for predicting the performance of artificial valve stents based on digital twins according to claim 1, characterized in that, It also includes step S9; In step S9, a life assessment module is established. The remaining life output from the remaining life prediction model in step S8 is used as the input of the life assessment module. The remaining life is compared with a preset value. When the preset value is exceeded, an early warning is issued.

7. The method for predicting the performance of artificial valve stents based on digital twins according to claim 1, characterized in that, It also includes step S10; step S10 is to establish a human-computer interaction module based on the simulation data in step S6; the human-computer interaction module is used to display the deformation generated during the service of the valve stent.

Images