Methods and devices for designing and evaluating valves

Abstract

The present invention discloses a method and device for designing and evaluating a valve, the method comprising: obtaining a patient's coronary CTA image sequence, and constructing a physical model of the aortic root, aortic valve, and left ventricular outflow tract based on the CTA image sequence; determining different artificial valve physical models, blood flow related parameters, and the release position of the artificial valve; based on the physical model and blood flow related parameters, using a valve replacement model pre-trained based on the acquired real CTA image sequence to sequentially simulate the process of placing each artificial valve physical model in the release position to obtain corresponding simulation results; determining a valve implantation risk index based on the simulation results; and selecting the design parameters of the artificial valve based on the valve implantation risk index. Using the scheme of the present invention, individualized artificial valve data can be obtained efficiently, quickly, and accurately, providing an effective reference for clinical transcatheter aortic valve replacement surgery, and also providing an effective reference for the design of new valve models.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to a method and device for designing and evaluating a valve. Background Art

[0002] The aortic valve is a valve structure where the beginning of the aorta connects to the left ventricle. It is composed of three semilunar valves that inhibit the blood flow injected into the aorta from flowing back into the left ventricle. Because of its central position, the aortic valve is closely related to various heart chambers and valves. The aortic valve can be narrowed (aortic stenosis), insufficiency (aortic regurgitation) and other causes due to various causes, resulting in a series of pathophysiological changes. At present, for patients with aortic valve disease who cannot undergo traditional surgery and have symptoms such as chest pain, TAVR (Transcatheter Aortic Valve Replacement) is usually used as a treatment method. After the catheter carrying the artificial valve enters the left ventricle from the aorta, the artificial valve is delivered to the position of the native aortic valve for release.

[0003] CT (Computed Tomography) images are the imaging evaluation criteria used when selecting artificial valves before TAVR surgery. They can provide a reference for valve selection and surgical approach. Therefore, relevant computer software has emerged in the industry to help doctors make anatomical structures on CT images, but this type of software has the following disadvantages:

[0004] 1. The measurement content only includes the anatomical parameters of the aortic root, and does not include functional parameters such as blood reflux and the pressure of the artificial valve on the aorta, which will lead to an incomplete evaluation. Only experienced doctors can evaluate problems such as severe blood reflux, coronary artery blockage, and excessive pressure of the artificial valve on the blood vessel wall, resulting in rupture of the blood vessel wall and bleeding during surgery without the help of functional parameters;

[0005] 2. Manual measurement may result in different results between different doctors. Summary of the invention

[0006] The present invention provides a method and device for designing and evaluating a valve, so as to obtain individualized artificial valve data efficiently, quickly and accurately, and provide an effective reference for clinical artificial valve model selection, artificial valve design and clinical transcatheter aortic valve replacement surgery.

[0007] To this end, the present invention provides the following technical solutions:

[0008] In one aspect, the present invention provides a method for designing and evaluating a valve, the method comprising:

[0009] Acquire a coronary CTA image sequence of the patient, and construct a physical model of the aortic root, aortic valve, and left ventricular outflow tract based on the CTA image sequence;

[0010] Identify different physical models of artificial valves;

[0011] Acquiring blood flow related parameters of the patient;

[0012] Determine the release position of the prosthetic valve;

[0013] Based on the physical models of the aortic root, aortic valve and left ventricular outflow tract, and the blood flow related parameters, using a valve replacement model for simulating the release process of an artificial valve at the aortic root that is pre-trained based on the acquired real CTA image sequence, the process of placing each artificial valve physical model at the release position is sequentially simulated to obtain simulation results corresponding to each artificial valve physical model;

[0014] Determining the valve implantation risk index of each artificial valve physical model according to the simulation results;

[0015] The design parameters of the artificial valve are selected according to the valve implantation risk index.

[0016] Optionally, the valve replacement model for simulating the release process of the artificial valve at the aortic root obtained by training based on the acquired real CTA image sequence includes:

[0017] Acquire real CTA image sequences and corresponding blood flow related parameters;

[0018] Based on the acquired real CTA image sequences, physical models of the aortic root, aortic valve and left ventricular outflow tract were constructed, as well as a physical model of the artificial valve;

[0019] Obtaining a training data set and corresponding partial differential equation parameters according to the physical models of the aortic root, the aortic valve and the left ventricular outflow tract, the physical model of the artificial valve, and the blood flow related parameters;

[0020] The training data set and the corresponding partial differential equation parameters are used to train a valve replacement model for simulating the release process of an artificial valve at the aortic root.

[0021] Optionally, constructing the physical models of the aortic root, aortic valve and left ventricular outflow tract based on the acquired real CTA image sequence data includes:

[0022] Based on the real CTA image sequences collected, the segmentation model of the aortic root, aortic valve and left ventricular outflow tract is trained;

[0023] constructing a three-dimensional image of the aortic root using the segmentation model of the aortic root blood vessels, aortic valve and left ventricular outflow tract;

[0024] The biomechanical parameters of the aortic root blood vessels, aortic valve and left ventricular outflow tract in the three-dimensional image of the aortic root are determined, and a physical model of the aortic root, aortic valve and left ventricular outflow tract is generated.

[0025] Optionally, the training of the segmentation model of the aortic root, aortic valve and left ventricular outflow tract based on the acquired real CTA image sequence includes:

[0026] Segmenting the aortic root, aortic valve and left ventricular outflow tract of the real CTA image sequence to obtain a training data set and a corresponding mask;

[0027] The training data set and the corresponding mask are used to train a segmentation model of the aortic root, aortic valve and left ventricular outflow tract.

[0028] Optionally, the segmentation model of the aortic root, the aortic valve and the left ventricular outflow tract obtained by training with the training data set and the corresponding mask includes:

[0029] Using the training data set and the corresponding mask, a U-net network or a fully convolutional network is used for deep learning training to obtain a segmentation model of the aortic root, aortic valve and left ventricular outflow tract; or

[0030] The training data set and the corresponding mask are used to train the segmentation models of the aortic root, the aortic valve and the left ventricular outflow tract using a threshold segmentation method or a region growing method.

[0031] Optionally, the biomechanical parameters include any one or more of the following: Young's modulus, Poisson's ratio, and blood vessel rupture strength.

[0032] Optionally, determining different artificial valve physical models includes: selecting different physical parameters of the artificial valve to obtain a plurality of different artificial valve physical models.

[0033] Optionally, the physical parameters include: dimensional parameters and mechanical parameters; the dimensional parameters include any one or more of the following: implant shortening rate, valve frame height when the artificial valve is fully expanded, skirt height, maximum diameter of the inflow site, maximum diameter of the outflow site, minimum diameter of the waist; the mechanical parameters include any one or more of the following: radial resistance, chronic outward force.

[0034] Optionally, the blood flow related parameters include any one or more of the following: heart rate, blood output per beat, blood pressure, and blood viscosity.

[0035] Optionally, determining the valve implantation risk index of each artificial valve physical model according to the simulation result includes:

[0036] According to the simulation results, calculating the physiological functional parameters corresponding to each artificial valve physical model;

[0037] The valve implantation risk index of each artificial valve physical model is determined according to the physiological functional parameters.

[0038] Optionally, the physiological functional parameters include any one or more of the following: aortic root tear risk index, aortic root bleeding risk index, coronary artery occlusion risk index, valve detachment risk index, and blood reflux percentage.

[0039] Optionally, selecting the design parameters of the artificial valve according to the risk index includes: selecting parameters of the artificial valve physical model with the smallest risk index as the design parameters.

[0040] Optionally, the method further comprises:

[0041] Generating a corresponding three-dimensional image of the artificial valve according to different physical parameters of the artificial valve;

[0042] Generating a three-dimensional image of the aortic root, aortic valve and left ventricular outflow tract according to the patient's coronary CTA image sequence;

[0043] reconstructing a three-dimensional image of the artificial valve implantation process based on the three-dimensional images of the aortic root, the aortic valve and the left ventricular outflow tract, and the three-dimensional image of the artificial valve;

[0044] According to the simulation results and the patient's coronary CTA image sequence, a DSA image of an artificial valve implant is simulated.

[0045] In another aspect, the present invention further provides a device for designing and evaluating a valve, the device comprising:

[0046] A first model building module is used to obtain a coronary CTA image sequence of a patient and build a physical model of the aortic root, aortic valve and left ventricular outflow tract based on the CTA image sequence;

[0047] A second model building module is used to determine different artificial valve physical models;

[0048] A parameter acquisition module, used to acquire blood flow related parameters of the patient;

[0049] A release position determination module, used to determine the release position of the artificial valve;

[0050] A release process simulation module is used to sequentially simulate the process of placing each artificial valve physical model to the release position based on the physical models of the aortic root, aortic valve and left ventricular outflow tract, and the blood flow related parameters, using a valve replacement model for simulating the release process of the artificial valve at the aortic root obtained by pre-training based on the acquired real CTA image sequence, to obtain simulation results corresponding to each artificial valve physical model;

[0051] A risk index calculation module, used to determine the valve implantation risk index of each artificial valve physical model according to the simulation results;

[0052] The design parameter selection module is used to select the design parameters of the artificial valve according to the valve implantation risk index.

[0053] Optionally, the device further comprises: a valve replacement model training module, wherein the valve replacement model training module comprises:

[0054] A data acquisition unit, used to acquire real CTA image sequences and corresponding blood flow related parameters;

[0055] A physical model building unit, used to build a physical model of the aortic root, aortic valve and left ventricular outflow tract, and a physical model of an artificial valve based on the real CTA image sequence acquired by the data acquisition unit;

[0056] A training data generating unit, configured to obtain a training data set and corresponding partial differential equation parameters according to the physical models of the aortic root, the aortic valve and the left ventricular outflow tract, the physical model of the artificial valve, and the blood flow related parameters;

[0057] A training unit is used to use the training data set and the corresponding partial differential equation parameters to train a valve replacement model for simulating the release process of an artificial valve at the aortic root.

[0058] Optionally, the physical model building unit includes:

[0059] A segmentation subunit, configured to segment the aortic root, aortic valve and left ventricular outflow tract of the real CTA image sequence to obtain a training data set and a corresponding mask;

[0060] The segmentation model training subunit is used to obtain the segmentation model of the aortic root, the aortic valve and the left ventricular outflow tract by using the training data set and the corresponding mask training.

[0061] Optionally, the device further comprises: a DSA image generation module;

[0062] The second model building module is further used to generate a corresponding three-dimensional image of the artificial valve according to different physical parameters of the artificial valve;

[0063] The first model building module is further used to generate a three-dimensional image of the aortic root, aortic valve and left ventricular outflow tract according to the patient's coronary CTA image sequence;

[0064] The release process simulation module is further used to reconstruct a three-dimensional image of the artificial valve implantation process based on the three-dimensional images of the aortic root, the aortic valve and the left ventricular outflow tract, and the three-dimensional image of the artificial valve;

[0065] The DSA image generation module is used to simulate a DSA image of an artificial valve implantation according to the simulation result and the patient's coronary CTA image sequence.

[0066] The method and device for designing and evaluating valves provided by the present invention are pre-trained based on the collected real CTA image sequence to obtain a valve replacement model for simulating the release process of an artificial valve at the aortic root. When selecting an artificial valve for a patient who needs to undergo transcatheter aortic valve replacement surgery, a physical model of the aortic root, aortic valve and left ventricular outflow tract is constructed based on the patient's coronary CTA image sequence, and the valve replacement model is used to simulate the process of placing the artificial valve physical model at a determined release position, and the simulation results corresponding to various artificial valve physical models are obtained, and then the valve implantation risk index of each artificial valve physical model is determined, and the design parameters of the artificial valve are selected according to the valve implantation risk index, guiding the development of artificial valves and the selection of artificial valve models in TAVR surgery, so that individualized artificial valve data can be obtained efficiently, quickly and accurately, and provide an effective reference for clinical artificial valve model selection, artificial valve design and clinical transcatheter aortic valve replacement surgery. By utilizing the scheme of the present invention, it is possible to simulate artificial valve sizes that are not available on the market, and to adjust the sizes of different parts of the artificial valve in a personalized manner according to the patient's own cardiovascular condition, thereby making the artificial valve better adapted to the patient's own condition.

[0067] Furthermore, the solution of the present invention can simulate DSA (Digital subtraction angiography) images of the aortic root based on CTA images, and can simulate the implantation of artificial valves on DSA images, which is closer to the real image seen by doctors during TAVR surgery, thereby helping doctors to evaluate the surgery more comprehensively. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 is a flow chart of a valve replacement model obtained by training in the method of the present invention;

[0069] Figure 2 is a flow chart of constructing a physical model of the aortic root, aortic valve and left ventricular outflow tract in the method of the present invention;

[0070] Figure 3 is a flow chart of an embodiment of a method for designing and evaluating a valve of the present invention;

[0071] Figure 4 It is a schematic diagram of the structure of the device for designing and evaluating a valve of the present invention;

[0072] Figure 5 It is a structural schematic diagram of a valve replacement model training module in the device for designing and evaluating a valve of the present invention;

[0073] Figure 6 It is another structural schematic diagram of the device for designing and evaluating a valve of the present invention. DETAILED DESCRIPTION

[0074] In order to enable persons skilled in the art to better understand the solutions of the embodiments of the present invention, the embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and implementation modes.

[0075] In view of the problems that when existing patients with aortic valve disease are treated with TAVR, an artificial valve needs to be selected before the operation, and different individual artificial valve data have different requirements, the present invention provides a method and device for designing and evaluating a valve. Based on CTA (CT angiography) images, a three-dimensional model and a physical model of the aortic root are reconstructed. According to the models of the aortic root and the artificial valve, the release position of the artificial valve, etc., a pre-established valve replacement model is used to release the artificial valve at the aortic root. According to the simulation results, relevant functional parameters are calculated to help select artificial valve parameters, guide the development of artificial valves and the selection of artificial valve models in TAVR surgery.

[0076] In the embodiment of the present invention, the simulation of the release process of the artificial valve at the aortic root is performed using a pre-established valve replacement model, wherein the valve replacement model is trained based on a real CTA image sequence acquired.

[0077] The valve replacement model may adopt a deep learning model. The training process of the valve replacement model is first described in detail below.

[0078] like Figure 1 FIG. 1 is a flowchart of training a valve replacement model in the method of the present invention, comprising the following steps:

[0079] In step 101, a real CTA image sequence and corresponding blood flow related parameters are acquired.

[0080] CTA combines CT enhancement technology with thin-layer, large-range, and fast scanning technology to clearly display the details of blood vessels in all parts of the body through reasonable post-processing. The real CTA image sequence refers to a large number of CTA image sequences of patients who have undergone aortic valve replacement.

[0081] It should be noted that the real CTA image sequence includes the CTA image sequence of the patient before and after surgery.

[0082] Among them, the blood flow related parameters are the blood flow data detected and recorded before the corresponding patient undergoes transcatheter aortic valve replacement surgery, which may specifically include but are not limited to any one or more of the following: heart rate, blood output per stroke, blood pressure, blood viscosity, etc.

[0083] In step 102, based on the acquired real CTA image sequence, physical models of the aortic root, aortic valve and left ventricular outflow tract are constructed, and a corresponding physical model of the artificial valve is constructed.

[0084] The process of constructing the physical model of the aortic root, aortic valve and left ventricular outflow tract is as follows: Figure 2 As shown, the following steps are included:

[0085] Step 201 : Based on the acquired real CTA image sequence, a segmentation model of the aortic root, aortic valve and left ventricular outflow tract is trained.

[0086] First, the aortic root, aortic valve and left ventricular outflow tract are segmented for the real CTA image sequence to obtain a training data set and a corresponding mask for training the segmentation model; the training data set for training the segmentation model includes: the original CTA images of existing cases collected, and the images segmented on these images.

[0087] Then, the segmentation model of the aortic root, aortic valve and left ventricular outflow tract is obtained by training using the training data set and the corresponding mask. For example, a U-net network or a fully convolutional network can be used to perform deep learning training to obtain the segmentation model of the aortic root, aortic valve and left ventricular outflow tract. Of course, a traditional algorithm, such as a threshold segmentation method or a regional growing method, can also be used to train to obtain the segmentation model of the aortic root, aortic valve and left ventricular outflow tract, which is not limited in this embodiment of the present invention.

[0088] Step 202: construct a three-dimensional image of the aortic root using the segmentation model of the aortic root blood vessels, aortic valve, and left ventricular outflow tract.

[0089] Step 203, determining the biomechanical parameters of the aortic root blood vessels, aortic valve and left ventricular outflow tract in the three-dimensional image of the aortic root, and generating a physical model of the aortic root, aortic valve and left ventricular outflow tract.

[0090] The biomechanical parameters may include, but are not limited to, any one or more of the following: Young's modulus, Poisson's ratio, vascular rupture strength, etc.

[0091] The physical model of the artificial valve can be generated according to the physical parameters of the artificial valve actually used by the corresponding patient, and the physical parameters may include: size parameters and mechanical parameters. The size parameters include but are not limited to any one or more of the following: implant shortening rate, valve frame height when the artificial valve is fully expanded, skirt height, maximum diameter of the inflow part, maximum diameter of the outflow part, minimum waist diameter, etc.; the mechanical parameters may include but are not limited to any one or more of the following: radial resistance, chronic outward force, etc.

[0092] Continue to refer to Figure 1 In step 103, a training data set and corresponding partial differential equation parameters for training a valve replacement model are obtained based on the physical models of the aortic root, aortic valve and left ventricular outflow tract, the physical model of the artificial valve and the blood flow related parameters. The training data set for training the valve replacement model includes: physical models of the aortic root, aortic valve and left ventricular outflow tract before and after surgery based on the collected existing cases, a physical model of blood flow, a physical model of the artificial valve actually used in the patient's surgery (the physical model includes the shape of the object and various physical property parameter values), and the actual morphology of the artificial valve after surgery.

[0093] The partial differential equations used include, but are not limited to, partial differential equations of elastic deformation and partial differential equations of fluid mechanics.

[0094] The partial differential equation of elastic deformation that needs to be calculated is:

[0095]

[0096] Among them, i represents the direction of one of the x, y, and z axes in the three-dimensional space, t represents the time t, and F it is the elastic force on the object in direction i at time t, are the parameters of the partial differential equation to be calculated, is the hemodynamic component in the direction i at time t (i.e., the f it ), anatomical parameters such as the diameter of the aortic root, aortic valve, and left ventricular outflow tract, physical quantities such as Young's modulus and Poisson's ratio, the size and mechanical parameters of the artificial valve, and a vector composed of the reciprocals of the above parameters, the squares of the parameters, and the products of the parameters.

[0097] The fluid mechanics partial differential equation required for calculation is:

[0098]

[0099] Among them, f it is the force generated when blood is pumped out of the left ventricle in direction i at time t. are the parameters of the partial differential equation of fluid mechanics to be calculated, It is a vector composed of the patient's heart rate, stroke volume, blood pressure, blood viscosity, blood density and other blood flow parameters at time point t, anatomical parameters such as the diameter of the aortic root, aortic valve, and left ventricular outflow tract, physical quantities such as Young's modulus and Poisson's ratio, the size and mechanical parameters of the artificial valve, as well as the reciprocals of the above parameters, the squares of the parameters, the products of the parameters, etc.

[0100] In step 104, the training data set and the corresponding partial differential equation parameters are used to train a valve replacement model for simulating the release process of the artificial valve at the aortic root.

[0101] Specifically, the valve replacement model can be trained by using traditional physical simulation algorithms, such as finite element analysis, finite volume analysis algorithms, etc., or by using deep learning algorithms, such as Fourier neural operator method, etc. The specific training process of the corresponding algorithm can refer to the existing technology, which will not be described in detail.

[0102] When selecting an artificial valve for a patient who needs transcatheter aortic valve replacement surgery, the above-mentioned valve replacement model is used to simulate the process of placing the physical model of the artificial valve to a determined release position, and then the design parameters of the artificial valve suitable for the patient are selected based on the simulation results.

[0103] like Figure 3 FIG. 1 is a flow chart of an embodiment of a method for designing and evaluating a valve according to the present invention, comprising the following steps:

[0104] Step 301 : Acquire a coronary CTA image sequence of a patient, and construct a physical model of the aortic root, aortic valve, and left ventricular outflow tract based on the CTA image sequence.

[0105] The construction process of the physical model of the aortic root, aortic valve and left ventricular outflow tract is the same as the construction method of the corresponding model in the previous valve replacement model training process. Figure 2 The description in the process shown will not be repeated here.

[0106] Step 302, determining different artificial valve physical models.

[0107] Since the physiological parameters of different patients are different, the artificial valves corresponding to different patients need to be specifically designed to better meet the patients' corresponding physiological parameters, thereby better ensuring the effect of the artificial valve after implantation and reducing the various risks brought to patients due to inappropriate artificial valve design parameters.

[0108] To this end, in the embodiment of the present invention, artificial valves with various physical parameters can be designed, and the physical parameters can include: size parameters and mechanical parameters. The size parameters and mechanical parameters can refer to the above description and will not be repeated here.

[0109] In specific applications, for example, based on existing artificial valve products, some or all of the parameters can be adjusted to obtain a variety of different physical parameters, and then a corresponding artificial valve physical model can be constructed based on these physical parameters.

[0110] Step 303: Acquire blood flow related parameters of the patient.

[0111] As mentioned above, blood flow related parameters may include but are not limited to any one or more of the following: heart rate, blood output per beat, blood pressure, blood viscosity, etc.

[0112] It should be noted that the above steps 301 to 303 are not performed in any particular order and can be performed in parallel.

[0113] Step 304, determining the release position of the artificial valve.

[0114] The release position of the artificial valve can be determined based on the positions of the aortic root, aortic valve and left ventricular outflow tract in the patient's coronary CTA images.

[0115] Step 305, based on the physical models of the aortic root, aortic valve and left ventricular outflow tract, and the blood flow related parameters, the valve replacement model is used to sequentially simulate the process of placing each artificial valve physical model to the release position to obtain simulation results corresponding to each artificial valve physical model.

[0116] The simulation results are the physical states such as the shape and stress of the artificial valve, aortic root, aortic valve, and left ventricular outflow tract when they are finally stable after the artificial valve is released.

[0117] Step 306: Determine the valve implantation risk index of each artificial valve physical model according to the simulation results.

[0118] Specifically, the physiological functional parameters corresponding to each artificial valve physical model can be calculated according to the simulation results, wherein the physiological functional parameters can include but are not limited to any one or more of the following: aortic root tear risk index, aortic root bleeding risk index, coronary artery blockage risk index, valve loss risk index, blood reflux percentage, etc. Then, the valve implantation risk index of each artificial valve physical model can be determined by comprehensively considering the physiological functional parameters, for example, weighted calculation of each physiological functional parameter is performed to obtain the valve implantation risk index corresponding to each artificial valve physical model.

[0119] Step 307: Select design parameters of an artificial valve according to the valve implantation risk index.

[0120] For multiple different artificial valve physical models, the parameters of the artificial valve physical model with the smallest risk index are selected as design parameters. In this way, a personalized artificial valve can be designed and manufactured for the patient according to the selected design parameters.

[0121] The method for designing and evaluating valves provided by the present invention, when selecting an artificial valve for a patient who needs to undergo transcatheter aortic valve replacement surgery, uses a valve replacement model pre-trained based on the collected real CTA image sequence to simulate the implantation of artificial valves with various design parameters, and determines the valve implantation risk index of each artificial valve physical model according to the simulation results, selects the design parameters of the artificial valve according to the valve implantation risk index, guides the development of artificial valves and the selection of artificial valve models in TAVR surgery, so that individualized artificial valve data can be obtained efficiently, quickly and accurately, and provides an effective reference for clinical artificial valve model selection, artificial valve design and clinical transcatheter aortic valve replacement surgery. Using the scheme of the present invention, it is possible to simulate artificial valve sizes that are not available on the market, and adjust the sizes of different parts of the artificial valve in a personalized manner according to the patient's own cardiovascular condition, so that the artificial valve can be better adapted to the patient's own condition.

[0122] Furthermore, in another embodiment of the method for designing and evaluating a valve of the present invention, some or all of the following steps may also be included:

[0123] (1) generating a corresponding three-dimensional image of the artificial valve according to different physical parameters of the artificial valve;

[0124] (2) generating a three-dimensional image of the aortic root, aortic valve, and left ventricular outflow tract based on the patient's coronary CTA image sequence;

[0125] (3) reconstructing a three-dimensional image of the artificial valve implantation process based on the three-dimensional images of the aortic root, the aortic valve and the left ventricular outflow tract, and the three-dimensional image of the artificial valve;

[0126] (4) Simulating a DSA image of an artificial valve implant based on the simulation results and the patient's coronary CTA image sequence.

[0127] Specifically, the coronary CTA image sequence is first used to simulate the DSA angiography of the coronary arteries; then, based on the final shape and position of the artificial valve in the simulation results, the artificial valve is drawn on the simulated DSA angiography of the coronary arteries to obtain a DSA image simulating the implantation of the artificial valve.

[0128] Utilizing the above-mentioned three-dimensional images, a DSA image of the aortic root can be simulated based on the CTA image, and the implantation of artificial valves on the DSA image can be simulated, which is closer to the real image seen by the doctor during TAVR surgery, thereby helping the doctor to evaluate the surgery more comprehensively.

[0129] Accordingly, the present invention also provides a device for designing and evaluating a valve, such as Figure 4 Shown is a schematic diagram of the structure of the device for designing and evaluating a valve according to the present invention.

[0130] In this embodiment, the device 400 for designing and evaluating a valve includes the following modules:

[0131] A first model building module 401 is used to obtain a coronary CTA image sequence of a patient, and to build a physical model of the aortic root, aortic valve and left ventricular outflow tract based on the CTA image sequence;

[0132] A second model building module 402 is used to determine different artificial valve physical models;

[0133] A parameter acquisition module 403 is used to acquire blood flow related parameters of the patient;

[0134] A release position determination module 404, used to determine the release position of the artificial valve;

[0135] The release process simulation module 405 is used to sequentially simulate the process of placing each artificial valve physical model to the release position based on the physical models of the aortic root, the aortic valve and the left ventricular outflow tract, and the blood flow related parameters, using a valve replacement model for simulating the release process of the artificial valve at the aortic root obtained by pre-training based on the acquired real CTA image sequence, and obtain the simulation results corresponding to each artificial valve physical model;

[0136] A risk index calculation module 406, for determining the valve implantation risk index of each artificial valve physical model according to the simulation results;

[0137] The design parameter selection module 407 is used to select the design parameters of the artificial valve according to the valve implantation risk index.

[0138] It should be noted that the valve replacement model can be obtained by training the corresponding valve replacement model training module based on the acquired real CTA image sequence. The valve replacement model training module can be used as a part of the device for designing and evaluating the valve of the present invention, or it can be independent of the device, and this embodiment of the present invention is not limited to this.

[0139] like Figure 5 The figure is a structural schematic diagram of a valve replacement model training module in the device for designing and evaluating valves of the present invention.

[0140] The valve replacement model training module 500 includes the following units:

[0141] A data acquisition unit 501 is used to acquire real CTA image sequences and corresponding blood flow related parameters;

[0142] A physical model building unit 502, for building a physical model of the aortic root, aortic valve and left ventricular outflow tract, and a physical model of an artificial valve based on the real CTA image sequence acquired by the data acquisition unit;

[0143] A training data generating unit 503 is used to obtain a training data set and corresponding partial differential equation parameters according to the physical models of the aortic root, the aortic valve and the left ventricular outflow tract, the physical model of the artificial valve, and the blood flow related parameters;

[0144] The training unit 504 is used to use the training data set and the corresponding partial differential equation parameters to train a valve replacement model for simulating the release process of the artificial valve at the aortic root.

[0145] The physical model building unit 502 may specifically include the following subunits:

[0146] A segmentation subunit, configured to segment the aortic root, aortic valve and left ventricular outflow tract of the real CTA image sequence to obtain a training data set and a corresponding mask;

[0147] The segmentation model training subunit is used to obtain the segmentation model of the aortic root, the aortic valve and the left ventricular outflow tract by using the training data set and the corresponding mask training.

[0148] The specific implementation of the above modules, units, and sub-units can be

[0149] For more details about the working principle and working method of the above-mentioned device for designing and evaluating a valve, please refer to the relevant description in the previous embodiment of the method for designing and evaluating a valve, which will not be repeated here.

[0150] The device for designing and evaluating valves provided by the present invention, when selecting an artificial valve for a patient who needs to undergo transcatheter aortic valve replacement surgery, uses a valve replacement model pre-trained based on the acquired real CTA image sequence to simulate the implantation of artificial valves with various design parameters, and determines the valve implantation risk index of each artificial valve physical model according to the simulation results, selects the design parameters of the artificial valve according to the valve implantation risk index, guides the development of artificial valves and the selection of artificial valve models in TAVR surgery, thereby efficiently, quickly and accurately obtaining individualized artificial valve data, and providing an effective reference for clinical artificial valve model selection, artificial valve design and clinical transcatheter aortic valve replacement surgery. Using the scheme of the present invention, it is possible to simulate artificial valve sizes that are not available on the market, and adjust the sizes of different parts of the artificial valve in a personalized manner according to the patient's own cardiovascular condition, so that the artificial valve can be better adapted to the patient's own condition.

[0151] Furthermore, if Figure 6 As shown, in another embodiment of the device for designing and evaluating a valve of the present invention, it may also include: a DSA image generating module 601.

[0152] and Figure 4 The difference between the embodiment shown is that in this embodiment, in the device 400 for designing and evaluating a valve,

[0153] The second model building module 402 is further used to generate a corresponding three-dimensional image of the artificial valve according to different physical parameters of the artificial valve;

[0154] The first model building module 401 is also used to generate a three-dimensional image of the aortic root, aortic valve and left ventricular outflow tract according to the patient's coronary CTA image sequence;

[0155] The release process simulation module 405 is also used to reconstruct a three-dimensional image of the artificial valve implantation process based on the three-dimensional images of the aortic root, the aortic valve and the left ventricular outflow tract, and the three-dimensional image of the artificial valve;

[0156] The DSA image generation module 601 is used to simulate a DSA image of an artificial valve implantation according to the simulation result and the patient's coronary CTA image sequence.

[0157] Utilizing the above-mentioned three-dimensional images, a DSA image of the aortic root can be simulated based on the CTA image, and the implantation of artificial valves on the DSA image can be simulated, which is closer to the real image seen by the doctor during TAVR surgery, thereby helping the doctor to evaluate the surgery more comprehensively.

[0158] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or apparatus.

[0159] The various modules / units included in the various devices and products described in the above embodiments may be software modules / units or hardware modules / units, or may be partially software modules / units and partially hardware modules / units.

[0160] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. Moreover, the system embodiments described above are merely schematic, in which the modules and units described as separate components may or may not be physically separated, that is, they may be located on one network unit, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.

[0161] The embodiments of the present invention are described in detail above. The present invention is described in detail using specific implementation methods herein. The description of the above embodiments is only used to help understand the method and device of the present invention. It is only a part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should belong to the scope of protection of the present invention, and the content of this specification should not be understood as limiting the present invention. Therefore, any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for designing and evaluating a valve, characterized in that the method include: Obtain the patient's coronary CTA image sequence, and construct a physical model of the aortic root, aortic valve, and left ventricular outflow tract based on the CTA image sequence; Identify different physical models of artificial valves; Obtaining blood flow related parameters of the patient; Determine the release position of the prosthetic valve; Based on the physical models of the aortic root, aortic valve and left ventricular outflow tract, and blood flow related parameters, a valve replacement model for simulating the release process of an artificial valve at the aortic root, which is pre-trained based on the acquired real CTA image sequence, is used to sequentially simulate the process of placing each artificial valve physical model in the release position to obtain simulation results corresponding to each artificial valve physical model; According to the simulation results, the valve implantation risk index of each artificial valve physical model is determined; Select the design parameters of the artificial valve according to the valve implantation risk index; The valve replacement model for simulating the release process of the artificial valve at the aortic root obtained by training based on the acquired real CTA image sequence includes: acquiring the real CTA image sequence and the corresponding blood flow related parameters; constructing the physical model of the aortic root, the aortic valve and the left ventricular outflow tract, and constructing the physical model of the artificial valve based on the acquired real CTA image sequence; obtaining the training data set and the corresponding partial differential equation parameters according to the physical model of the aortic root, the aortic valve and the left ventricular outflow tract, the physical model of the artificial valve, and the blood flow related parameters; using the training data set and the corresponding partial differential equation parameters, the valve replacement model for simulating the release process of the artificial valve at the aortic root is trained; The method of constructing the physical model of the aortic root, aortic valve and left ventricular outflow tract based on the acquired real CTA image sequence includes: training a segmentation model of the aortic root, aortic valve and left ventricular outflow tract based on the acquired real CTA image sequence; constructing a three-dimensional image of the aortic root using the segmentation model of the aortic root, aortic valve and left ventricular outflow tract; determining the biomechanical parameters of the aortic root blood vessels, aortic valve and left ventricular outflow tract in the three-dimensional image of the aortic root, and generating a physical model of the aortic root, aortic valve and left ventricular outflow tract; The partial differential equations are partial differential equations of elastic deformation and partial differential equations of fluid mechanics; The elastic deformation partial differential equation is: Among them, i represents the direction of one of the x, y, and z axes in three-dimensional space, t represents time t, and Fit is the elastic force on the object in the i direction at time t. are the parameters of the partial differential equation to be calculated, It is a vector composed of the hemodynamic component in the direction i at time point t, anatomical parameters such as the diameter of the aortic root, aortic valve, and left ventricular outflow tract, physical quantities such as Young's modulus and Poisson's ratio, the size and mechanical parameters of the artificial valve, as well as the reciprocal of the above parameters, the square of the parameters, and the product of the parameters. The partial differential equation of fluid mechanics is: Among them, fit is the force generated by blood pumping out of the left ventricle in direction i at time t, are the parameters of the partial differential equation of fluid mechanics to be calculated, It is a vector composed of the patient's heart rate, stroke volume, blood pressure, blood viscosity, blood density and other blood flow parameters at time point t, anatomical parameters such as the diameter of the aortic root, aortic valve, and left ventricular outflow tract, physical quantities such as Young's modulus and Poisson's ratio, the size and mechanical parameters of the artificial valve, as well as the reciprocals of the above parameters, the squares of the parameters, the products of the parameters, etc.

2. The method according to claim 1, characterized in that The training of the segmentation model of the aortic root, aortic valve and left ventricular outflow tract based on the acquired real CTA image sequence includes: performing segmentation of the aortic root, aortic valve and left ventricular outflow tract on the real CTA image sequence to obtain a training data set and a corresponding mask; and using the training data set and the corresponding mask to train the segmentation model of the aortic root, aortic valve and left ventricular outflow tract.

3. The method according to claim 2, characterized in that The segmentation model of the aortic root, aortic valve and left ventricular outflow tract obtained by training with the training data set and the corresponding mask includes: using the training data set and the corresponding mask, using a U-net network or a fully convolutional network for deep learning training to obtain the segmentation model of the aortic root, aortic valve and left ventricular outflow tract; or using the training data set and the corresponding mask, using a threshold segmentation method or a region growing method to train to obtain the segmentation model of the aortic root, aortic valve and left ventricular outflow tract.

4. The method according to claim 1, characterized in that: Determining different artificial valve physical models includes: selecting different physical parameters of the artificial valve to obtain a plurality of different artificial valve physical models.

5. The method according to claim 4, characterized in that The physical parameters include: size parameters and mechanical parameters; the size parameters include any one or more of the following: implant shortening rate, valve frame height when the artificial valve is fully expanded, skirt height, maximum diameter of the inflow part, maximum diameter of the outflow part, minimum diameter of the waist; the mechanical parameters include any one or more of the following: radial resistance, chronic outward force.

6. The method according to claim 4, characterized in that The blood flow related parameters include any one or more of the following: heart rate, blood output per stroke, blood pressure, and blood viscosity.

7. The method according to claim 4, characterized in that Determining the valve implantation risk index of each artificial valve physical model according to the simulation results includes: calculating the physiological functional parameters corresponding to each artificial valve physical model according to the simulation results; and determining the valve implantation risk index of each artificial valve physical model according to the physiological functional parameters.

8. The method according to claim 7, characterized in that The physiological functional parameters include any one or more of the following: aortic root tear risk index, aortic root bleeding risk index, coronary artery occlusion risk index, valve detachment risk index, and blood reflux percentage.

9. The method according to claim 1, characterized in that: Selecting the design parameters of the artificial valve according to the valve implantation risk index includes: selecting the parameters of the artificial valve physical model with the smallest risk index as the design parameters.

10. The method according to any one of claims 4 to 9, characterized in that: The method also includes: generating a corresponding three-dimensional image of the artificial valve according to different physical parameters of the artificial valve; generating a three-dimensional image of the aortic root, aortic valve and left ventricular outflow tract according to the patient's coronary CTA image sequence; reconstructing a three-dimensional image of the artificial valve implantation process based on the three-dimensional images of the aortic root, aortic valve and left ventricular outflow tract, and the three-dimensional image of the artificial valve; and simulating a DSA image of the artificial valve implantation according to the simulation results and the patient's coronary CTA image sequence.

11. A device for designing and evaluating a valve using the method according to any one of claims 1 to 10, characterized in that: The device includes: The first model building module is used to obtain the patient's coronary CTA image sequence and build a physical model of the aortic root, aortic valve and left ventricular outflow tract based on the CTA image sequence; A second model building module is used to determine different artificial valve physical models; A parameter acquisition module, used to obtain blood flow related parameters of the patient; A release position determination module, used to determine the release position of the artificial valve; A release process simulation module is used to sequentially simulate the process of placing each artificial valve physical model into the release position based on the physical models of the aortic root, aortic valve and left ventricular outflow tract, and blood flow related parameters, using a valve replacement model for simulating the release process of the artificial valve at the aortic root obtained by pre-training based on the acquired real CTA image sequence, and obtain simulation results corresponding to each artificial valve physical model; A risk index calculation module, used to determine the valve implantation risk index of each artificial valve physical model according to the simulation results; The design parameter selection module is used to select the design parameters of the artificial valve according to the valve implantation risk index.

12. The device according to claim 11, characterized in that The device also includes: a valve replacement model training module, specifically including: A data acquisition unit, used to acquire real CTA image sequences and corresponding blood flow related parameters; A physical model building unit, used for building a physical model of the aortic root, aortic valve and left ventricular outflow tract based on a real CTA image sequence acquired by the data acquisition unit, and building a physical model of an artificial valve; A training data generating unit, for obtaining a training data set and corresponding partial differential equation parameters according to the physical models of the aortic root, the aortic valve and the left ventricular outflow tract, the physical model of the artificial valve, and blood flow related parameters; The training unit is used to use the training data set and the corresponding partial differential equation parameters to train a valve replacement model for simulating the release process of the artificial valve at the aortic root.

13. The device according to claim 12, characterized in that The physical model building unit includes: The segmentation subunit is used to segment the aortic root, aortic valve and left ventricular outflow tract of the real CTA image sequence to obtain a training data set and corresponding masks; The segmentation model training subunit is used to obtain the segmentation model of the aortic root, the aortic valve and the left ventricular outflow tract by using the training data set and the corresponding mask training.

14. The device according to claim 13, characterized in that The device also includes: a DSA image generation module; The second model building module is also used to generate a corresponding three-dimensional image of the artificial valve according to different physical parameters of the artificial valve; The first model building module is also used to generate a three-dimensional image of the aortic root, aortic valve and left ventricular outflow tract based on the patient's coronary CTA image sequence; The release process simulation module is further used to reconstruct a three-dimensional image of the artificial valve implantation process based on the three-dimensional images of the aortic root, the aortic valve and the left ventricular outflow tract, and the three-dimensional image of the artificial valve; The DSA image generation module is used to simulate a DSA image of an artificial valve implantation according to the simulation result and the patient's coronary CTA image sequence.

Images