Prediction method and system for mechanical properties of blood vessel and calcified plaque material

Abstract

The invention discloses a method and a system for predicting the mechanical property of a blood vessel and calcified plaque material. The method comprises the following steps: constructing a three-dimensional geometric model containing an aorta root blood vessel wall, an autologous valve and a calcified plaque; constructing a bracket geometric model actually implanted in the operation; using finite element analysis to simulate the implantation process of the stent, and obtaining the geometric morphology index of the simulated stent valve ring; based on the obtained geometrical morphology indexes of the stent valve ring, the material mechanical property with clinical value of the target patient is determined; obtaining material mechanical properties with clinical values of a plurality of target patients and corresponding clinical feature data thereof, and constructing a data set; based on the constructed data set, a nonlinear regression model is adopted to construct a prediction model of clinical features and material mechanical properties; and performing prediction by using the prediction model. By adopting the method and the system, the mechanical property of the individual material can be obtained noninvasively, and the mechanical property of the new patient material can be predicted by associating the mechanical property with the clinical characteristics.

Description

Technical Field

[0001] This invention relates to the fields of smart healthcare and biomechanics, and in particular to a method and system for predicting the mechanical properties of vascular and calcified plaque materials. Background Technology

[0002] Transcatheter aortic valve replacement (TAVR) has become an important method for treating aortic valve disease. The success and safety of the procedure depend heavily on the accuracy of preoperative planning, which requires understanding the complex biomechanical environment of the aortic root wall, especially the individualized material mechanical properties of the vessel wall, the autologous valve, and calcified plaques. These mechanical properties directly affect the anchoring stability of the stent, the risk of paravalvular leakage, and complications such as conduction block.

[0003] Currently, in preoperative simulations, the properties of tissue materials are typically assigned based on literature values ​​or empirical assumptions from in vitro experiments. This fails to reflect the significant differences between individuals, leading to discrepancies between simulation results and reality, thus limiting its clinical guidance value. While some methods exist for inversely solving material parameters, they are mostly limited to calibration for single cases, lacking a systematic approach. They fail to correlate the real mechanical properties obtained after calibration of a large number of individuals with the broad clinical characteristics of patients, thus failing to form a universal predictive capability for preoperative planning of new patients.

[0004] Therefore, there is an urgent need for a technical solution that can systematically and non-invasively acquire individualized material mechanical properties and correlate them with readily available clinical characteristics, ultimately achieving accurate prediction of material mechanical properties for new patients. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for predicting the mechanical properties of vascular and calcified plaque materials, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a method for predicting the mechanical properties of vascular and calcified plaque materials, comprising the following steps: S1. Obtain the preoperative medical imaging data of the target patient and construct a three-dimensional geometric model based on the preoperative medical imaging data, including the aortic root vessel wall, autologous valve and calcified plaque. S2. Obtain postoperative medical imaging data of the target patient and the stent type and implantation strategy used in the treatment plan, and construct a stent geometric model; S3. Perform finite element analysis. Based on the three-dimensional geometric model, assign initial material property parameters to the aortic root vessel wall, autologous valve, and calcified plaque, and simulate the stent implantation process to obtain the simulated geometric morphology index at the stent annulus. S4. Compare the geometric morphology parameters of the stent annulus obtained by simulation with postoperative medical imaging data to determine the material mechanical properties with clinical value for the target patient. S5. Obtain clinically valuable material mechanical properties and their corresponding clinical characteristics data from multiple target patients, and construct a dataset; S6. Based on the constructed dataset, a nonlinear regression model is used to construct a predictive model of clinical characteristics and material mechanical properties; S7. Based on the clinical characteristic data of new patients, use a predictive model to predict the mechanical properties of materials for new patients.

[0007] Preferably, step S1 specifically includes: acquiring preoperative CT images and preoperative aortic valve ultrasound images of the target patient; segmenting and reconstructing a geometric model of the aortic root vessel wall and calcified plaques based on the preoperative CT images of the target patient; constructing a geometric model of the autologous valve based on the preoperative aortic valve ultrasound images; and fusing the three through 3D registration and Boolean operations to construct a three-dimensional geometric model containing the aortic root vessel wall, the autologous valve, and calcified plaques.

[0008] Preferably, step S3 specifically includes: S31. Mesh the three-dimensional geometric model, set the material mechanical properties of the autologous valve to fixed values, simplify the material mechanical properties of the aortic root vessel wall and calcified plaque to a linear elastic model, and characterize them using Young's modulus; at the same time, set the material property parameters of the stent and its auxiliary devices to fixed values. S32. Using the optimal Latin hypercube sampling method, multiple sets of material parameter combinations are generated in the parameter space of Young's modulus of the aortic root vessel wall and Young's modulus of calcified plaques. S33. For each combination of material parameters, perform finite element simulation of the stent implantation process and obtain the geometric morphological indicators at the stent annulus in each simulation.

[0009] Preferably, the geometric parameters at the stent annulus include major axis length, minor axis length, expansion rate, and ellipticity.

[0010] Preferably, step S4 specifically includes: S41. Based on the obtained geometric morphology indices at the stent annulus in each simulation, a Kriging surrogate model is established to describe the functional relationship between the material mechanical properties and the geometric morphology indices at the stent annulus. S42. Compare the geometric morphology indices at the stent annulus obtained from the simulation with postoperative medical imaging data, and use the NSGA-II multi-objective optimization algorithm to optimize and solve the established Kriging surrogate model to determine the clinically valuable material mechanical properties of the aortic root vessel wall and calcified plaques in the target patient.

[0011] Preferably, step S42 specifically includes: The simulated geometric morphology parameters of the stent annulus were compared with postoperative medical imaging data. The goal was to minimize the difference between the actual and simulated geometric morphology parameters of the stent annulus measured by CT images in the postoperative medical imaging data of the target patient. The NSGA-II multi-objective optimization algorithm was used to optimize and solve the Kriging surrogate model, obtaining the optimal solution set of material mechanical properties that best matches the geometric morphology parameters of the actual and simulated stent annulus. Based on the optimal solution set, the clinically valuable material mechanical properties of the aortic root vessel wall and calcified plaques of the target patient were obtained.

[0012] Preferably, the nonlinear regression model in step S6 includes: ; in, For the mechanical properties of materials, , … For significantly relevant clinical features, , … For regression coefficients, This is the error term.

[0013] Preferably, the clinical features include blood pressure, age, and visceral fat index.

[0014] This invention also provides a system for predicting the mechanical properties of vascular and calcified plaque materials, comprising: Medical imaging equipment is used to acquire preoperative CT image data, postoperative CT image data, preoperative aortic valve ultrasound images, and postoperative aortic valve ultrasound images of target patients. The processor, connected to the medical imaging device, is used to perform three-dimensional geometric model construction, stent geometric model construction, finite element simulation, surrogate model establishment, multi-objective optimization, and prediction model training. The terminal, connected to the processor, is used to visualize the geometric model, simulation process, optimization results, and prediction results.

[0015] Preferably, it also includes a database connected to the processor for storing three-dimensional geometric models, stent geometric models, material parameters, clinical features, and predictive models.

[0016] Therefore, the present invention employs the above-described method and system for predicting the mechanical properties of vascular and calcified plaque materials, which has the following beneficial effects: (1) This invention achieves non-invasive and accurate solution of the individualized mechanical properties of materials for patients through the technical path of "preoperative modeling-simulated implantation-postoperative calibration"; (2) By acquiring a large amount of patient data and establishing a predictive model, mechanical properties that are difficult to measure directly are associated with clinical characteristics that are easy to obtain. This enables the prediction of key tissue mechanical properties in new patients before surgery, providing unprecedented quantitative basis for stent selection, surgical risk assessment and strategy optimization in TAVR preoperative planning, and significantly improving the accuracy and personalization of surgical planning.

[0017] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0018] Figure 1 This is a flowchart of a method according to an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments, and therefore should not be construed as limiting the present invention.

[0020] Example Reference Figure 1 This invention provides a method for predicting the mechanical properties of materials used in blood vessels and calcified plaques, comprising the following steps: S1. Obtain preoperative medical imaging data of the target patient, and construct a three-dimensional geometric model based on the preoperative medical imaging data, including the aortic root vessel wall, autologous valve, and calcified plaques. Specifically, this includes: A CAT scanner was used to scan the aortic root wall of the target patient to obtain preoperative CT images. An ultrasound scanner was also used to scan the target patient to obtain preoperative aortic valve ultrasound images. The preoperative CT images were imported into medical image processing software. Through segmentation and reconstruction, a geometric model of the aortic root wall and calcified plaques was generated. Based on the preoperative aortic valve ultrasound images, a geometric model of the autologous valve was constructed. The three models were then fused using 3D registration and Boolean operations to create a three-dimensional geometric model encompassing the aortic root wall, the autologous valve, and the calcified plaques.

[0021] S2. Obtain postoperative medical imaging data of the target patient, as well as the stent type and implantation strategy used in the treatment plan, and construct a stent geometric model. Postoperative medical imaging data includes postoperative CT images and postoperative aortic valve ultrasound images.

[0022] S3. Perform finite element analysis. Based on a three-dimensional geometric model, assign initial material property parameters to the aortic root vessel wall, autologous valve, and calcified plaque, and simulate the stent implantation process to obtain the simulated geometric morphology indices at the stent annulus. The specific process includes: S31. Mesh the three-dimensional geometric model, set the material mechanical properties of the autologous valve to fixed values, and simplify the material mechanical properties of the aortic root vessel wall and calcified plaques to a linear elastic model, characterized by Young's modulus; simultaneously, set the material property parameters of the stent and its auxiliary devices to fixed values. Specifically, the Young's modulus of the aortic root vessel wall ranges from 23.20 MPa to 104.00 MPa; the Young's modulus of the calcified plaques ranges from 11.80 GPa to 19.58 GPa.

[0023] S32. Using the optimal Latin hypercube sampling method, 20 sets of material parameter combinations are generated in the parameter space of Young's modulus of the aortic root vessel wall and Young's modulus of calcified plaques.

[0024] S33. For each combination of material parameters, perform finite element simulation of the stent implantation process and obtain the geometric morphological indicators at the stent annulus in each simulation. The geometric morphological indicators at the stent annulus include the major axis length, minor axis length, expansion rate, and ellipticity.

[0025] S4. Compare the simulated geometric morphology parameters at the stent annulus with postoperative medical imaging data to determine the clinically valuable material mechanical properties for the target patient. The specific process includes: S41. Based on the obtained geometric morphology indices at the stent annulus in each simulation, a Kriging surrogate model is established to describe the functional relationship between the material mechanical properties and the geometric morphology indices at the stent annulus.

[0026] The Kriging surrogate model is a mathematical approximation model that can quickly predict the scaffold morphology corresponding to arbitrary material properties, replacing computationally expensive finite element simulations. Its core idea is to treat the response values ​​at unknown points as a spatially correlated random field. Given a set of training samples obtained through optimal Latin hypercube sampling... ,in , for input material parameters (Young's modulus of the aortic root and Young's modulus of calcified plaques). To obtain the morphological response (such as the ellipticity of the support) through finite element simulation, the Kriging surrogate model will use the response values... Spatial interpolation and high-precision prediction are performed to construct a continuous response surface based on known samples. The expression is as follows: ; In the formula, It is a global trend function; Given a matrix with zero mean and covariance of , A stationary Gaussian random process, R For parameterized related functions, This is a hyperparameter.

[0027] After training, the surrogate model can handle any new combination of material parameters. Give the predicted value This will reduce the cost of a single function evaluation by several orders of magnitude in subsequent optimizations, making large-scale parameter optimization possible.

[0028] S42. Compare the geometric morphology indices at the stent annulus obtained from the simulation with postoperative medical imaging data, and use the NSGA-II multi-objective optimization algorithm to optimize and solve the established Kriging surrogate model to determine the clinically valuable material mechanical properties of the aortic root vessel wall and calcified plaques in the target patient.

[0029] In this embodiment, step S42 specifically includes the following process: The simulated geometric morphological parameters of the stent annulus were compared with postoperative medical imaging data. The objective was to minimize the difference between the actual and simulated geometric morphological parameters of the stent annulus measured in CT images of the target patient. The NSGA-II multi-objective optimization algorithm was used to optimize the Kriging surrogate model, obtaining the optimal set of material mechanical properties that best matched the actual and simulated geometric morphological parameters of the stent annulus. Each set of material properties in this set represented an optimal trade-off among multiple objectives. The final solution was selected from the optimal set, and this set of material properties was determined as the clinically valuable material mechanical properties of the aortic root vessel wall and calcified plaques in the target patient.

[0030] The optimization process of the NSGA-II multi-objective optimization algorithm is performed on the response surface constructed using the Kriging surrogate model. The NSGA-II algorithm continuously generates, evaluates, and evolves a population of material parameters, and its "fitness" evaluation relies entirely on the rapid invocation of the Kriging surrogate model, rather than the original finite element simulation. Ultimately, the algorithm outputs a Pareto optimal solution set, where each solution represents a set of optimal compromise material parameters that cannot be further improved simultaneously among multiple morphological matching objectives.

[0031] S5. Obtain clinically valuable material mechanical properties and their corresponding clinical characteristics data from at least 200 target patients to construct a dataset. Clinical characteristics include blood pressure, age, and visceral fat index. When constructing the dataset, patient clinical characteristics are the independent variables, and patient material properties are the dependent variables.

[0032] S6. Based on the constructed dataset, a predictive model for clinical characteristics and material mechanical properties is built using a nonlinear regression model. The nonlinear regression model takes the following form: ; in, For the mechanical properties of materials, , … For significantly relevant clinical features, , … For regression coefficients, This is the error term.

[0033] S7. Based on the clinical characteristic data of new patients, use a predictive model to predict the mechanical properties of materials for new patients. The predictive model output includes the Young's modulus of the aortic root vessel wall and the Young's modulus of calcified plaques.

[0034] This invention also provides a system for predicting the mechanical properties of vascular and calcified plaque materials, comprising: Medical imaging equipment, including CAT equipment and ultrasound equipment, is used to acquire preoperative CT image data, postoperative CT image data, preoperative aortic valve ultrasound images, and postoperative aortic valve ultrasound images of target patients. The processor connects to medical imaging equipment and is used to perform 3D geometric model construction, stent geometric model construction, finite element simulation, surrogate model establishment, multi-objective optimization, and predictive model training. The terminal, connected to the processor, is used to visualize geometric models, simulation processes, optimization results, and prediction results.

[0035] The database, connected to the processor, is used to store three-dimensional geometric models, stent geometric models, material parameters, clinical features, and predictive models.

[0036] Therefore, the present invention employs the above-mentioned method and system for predicting the mechanical properties of vascular and calcified plaque materials. Through the reverse engineering approach of "preoperative modeling-simulated implantation-postoperative calibration", it can non-invasively obtain individualized material mechanical properties and, by associating them with readily available clinical characteristics, achieve accurate prediction of the material mechanical properties of new patients.

[0037] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for predicting the mechanical properties of materials used in blood vessels and calcified plaques, characterized in that, Including the following steps: S1. Obtain the preoperative medical imaging data of the target patient and construct a three-dimensional geometric model containing the aortic root vessel wall, autologous valve and calcified plaque based on the preoperative medical imaging data. S2. Obtain postoperative medical imaging data of the target patient and the stent type and implantation strategy used in the treatment plan, and construct a stent geometric model; S3. Perform finite element analysis. Based on the three-dimensional geometric model, assign initial material property parameters to the aortic root vessel wall, autologous valve, and calcified plaque, and simulate the stent implantation process to obtain the simulated geometric morphology index at the stent annulus. S4. Compare the geometric morphology parameters of the stent annulus obtained by simulation with postoperative medical imaging data to determine the material mechanical properties with clinical value for the target patient. S5. Obtain clinically valuable material mechanical properties and their corresponding clinical characteristics data from multiple target patients, and construct a dataset; S6. Based on the constructed dataset, a nonlinear regression model is used to construct a predictive model of clinical characteristics and material mechanical properties; S7. Based on the clinical characteristic data of new patients, use a predictive model to predict the mechanical properties of materials for new patients.

2. The method for predicting the mechanical properties of vascular and calcified plaque materials according to claim 1, characterized in that, Step S1 specifically includes: acquiring preoperative CT images and preoperative aortic valve ultrasound images of the target patient; segmenting and reconstructing the geometric model of the aortic root vessel wall and calcified plaques based on the preoperative CT images of the target patient; constructing the geometric model of the autologous valve based on the preoperative aortic valve ultrasound images; and fusing the three through 3D registration and Boolean operations to construct a three-dimensional geometric model containing the aortic root vessel wall, the autologous valve, and calcified plaques.

3. The method for predicting the mechanical properties of vascular and calcified plaque materials according to claim 1, characterized in that, Step S3 specifically includes: S31. Mesh the three-dimensional geometric model, set the material mechanical properties of the autologous valve to fixed values, simplify the material mechanical properties of the aortic root vessel wall and calcified plaque to a linear elastic model, and characterize them using Young's modulus; at the same time, set the material property parameters of the stent and its auxiliary devices to fixed values. S32. Using the optimal Latin hypercube sampling method, multiple sets of material parameter combinations are generated in the parameter space of Young's modulus of the aortic root vessel wall and Young's modulus of calcified plaques. S33. For each combination of material parameters, perform finite element simulation of the stent implantation process and obtain the geometric morphological indicators at the stent annulus in each simulation.

4. The method for predicting the mechanical properties of vascular and calcified plaque materials according to claim 3, characterized in that: The geometric parameters of the stent annulus include major axis length, minor axis length, expansion rate, and ellipticity.

5. The method for predicting the mechanical properties of vascular and calcified plaque materials according to claim 3, characterized in that, Step S4 specifically includes: S41. Based on the obtained geometric morphology indices at the stent annulus in each simulation, a Kriging surrogate model is established to describe the functional relationship between the material mechanical properties and the geometric morphology indices at the stent annulus. S42. Compare the geometric morphology indices at the stent annulus obtained from the simulation with postoperative medical imaging data, and use the NSGA-II multi-objective optimization algorithm to optimize and solve the established Kriging surrogate model to determine the clinically valuable material mechanical properties of the aortic root vessel wall and calcified plaques in the target patient.

6. The method for predicting the mechanical properties of vascular and calcified plaque materials according to claim 5, characterized in that, Step S42 specifically includes: The simulated geometric morphology parameters of the stent annulus were compared with postoperative medical imaging data. The goal was to minimize the difference between the actual and simulated geometric morphology parameters of the stent annulus measured by CT images in the postoperative medical imaging data of the target patient. The NSGA-II multi-objective optimization algorithm was used to optimize and solve the Kriging surrogate model, obtaining the optimal solution set of material mechanical properties that best matches the geometric morphology parameters of the actual and simulated stent annulus. Based on the optimal solution set, the clinically valuable material mechanical properties of the aortic root vessel wall and calcified plaques of the target patient were obtained.

7. The method for predicting the mechanical properties of vascular and calcified plaque materials according to claim 1, characterized in that, The nonlinear regression model in step S6 includes: ； in, For the mechanical properties of materials, , … For significantly relevant clinical features, , … For regression coefficients, This is the error term.

8. The method for predicting the mechanical properties of vascular and calcified plaque materials according to claim 1, characterized in that: The clinical features include blood pressure, age, and visceral fat index.

9. A system for predicting the mechanical properties of vascular and calcified plaque materials, used to execute the method for predicting the mechanical properties of vascular and calcified plaque materials as described in any one of claims 1-8, characterized in that, include: Medical imaging equipment is used to acquire preoperative CT image data, postoperative CT image data, preoperative aortic valve ultrasound images, and postoperative aortic valve ultrasound images of target patients. The processor, connected to the medical imaging device, is used to perform three-dimensional geometric model construction, stent geometric model construction, finite element simulation, surrogate model establishment, multi-objective optimization, and prediction model training. The terminal, connected to the processor, is used to visualize the geometric model, simulation process, optimization results, and prediction results.

10. The system for predicting the mechanical properties of vascular and calcified plaque materials according to claim 9, characterized in that: It also includes a database connected to the processor for storing three-dimensional geometric models, stent geometric models, material parameters, clinical features, and predictive models.

Images