Method for modeling an endoprosthesis to be implanted in a portion of a patient's aorta - Patent Application 20070122999

Abstract

The present invention relates to a computer-implemented method (100) for simulating the interaction of a portion of a patient's aorta with a specialized endoprosthesis to be implanted in the portion of the aorta to determine areas at risk for surgical complications.

Description

[Technical Field]

[0001] The present invention relates to the field of numerical simulation of cardiovascular interventional procedures.

[0002] More particularly, the present invention relates to a computer-implemented method for modeling an endoprosthesis to be implanted in a portion of a patient's aorta to improve surgical outcomes. Summary of the Invention [Problem to be solved by the invention]

[0003] Cardiovascular interventional procedures are often associated with complications such as endoleaks, endoprosthesis migration, endoprosthesis invagination, endoprosthesis occlusion, thrombosis, and endoprosthesis kinking. In this context, there is room for improvement, especially in the preoperative planning phase. The present invention provides a solution that allows for a pre-evaluation of the surgical outcome and also for the personalization of the procedure. [Means for solving the problem]

[0004] Accordingly, the present invention provides a computer-implemented method for simulating the interaction of a portion of a patient's aorta with a specialized endoprosthesis to be implanted therein to determine areas at risk for surgical complications, the method comprising: - receiving a 3D image of a portion of the patient's aorta; - reconstructing a deformable three-dimensional digital model of a portion of the patient's aorta, called an aortic model, based on the 3D images; - placing the three-dimensional digital model of the endoprosthesis in its planned position in the aortic model, and the three-dimensional model of the endoprosthesis and the aortic model together forming an implantation model; - simulating said interaction based on calculations using said implantation model; - determining at least one risk indicator of a surgical complication in a predefined region in the implantation model based on the results of the simulated interaction; The present invention relates to a method comprising:

[0005] For example, one global risk index can be determined for the entire endoprosthesis. In other examples, the predefined region varies depending on the type of surgical complication. For example, when the surgical complication is a type Ia endoleak, the predefined region is the proximal covered stent. In this disclosure, "endoprosthesis" refers to a "stent graft." The covered stent components of the stent graft are used to secure the top or beginning of the stent graft within the blood vessel. These components provide a sealed, stable connection between the stent graft and the native blood vessel, helping to prevent blood leakage into or around the aneurysm or damaged area. In this disclosure, the term "proximal" refers to upstream of the blood flow; for example, a proximal stent is a stent at the top of the stent graft that is placed closer to the heart. When the surgical complication involves a risk of migration, the predefined region is the proximal uncovered stent. When the surgical complication involves a risk of invagination or occlusion, the predefined region is the entire endoprosthesis. The exact predefined region varies depending on the endoprosthesis manufacturer.

[0006] Thus, this method makes it possible to foresee areas of potential surgical complications by using patient-related data, such as anatomical images, and planning data, such as a numerical model of the endoprosthesis planned to be implanted in the patient's body, while also making it possible to personalize surgical interventions, including the implantation of endoprostheses.

[0007] Advantageously, the surgical complications include at least one of endoleak, endoprosthesis migration, endoprosthesis invagination, endoprosthesis occlusion, thrombosis, and endoprosthesis kinking.

[0008] In some embodiments, the calculations relating to the implantation model include calculating a geometric deformation of a three-dimensional digital model of the endoprosthesis within an aortic model.

[0009] In some embodiments, the calculation of the geometric deformation is performed by finite element analysis.

[0010] Advantageously, the 3D images are preoperative CT scan images. This method therefore makes it possible to predict surgical complications before performing a surgical intervention. This is advantageous, as current techniques are based on postoperative data. This method therefore makes it possible to improve the planning of surgical interventions by allowing the testing of various instruments to be used during the intervention.

[0011] In some embodiments, the interaction is endovascular repair of an aortic aneurysm.

[0012] In some embodiments, the method further comprises, after determining the areas at risk for a surgical complication, determining a risk level for each area at risk for a surgical complication.

[0013] In some embodiments, determining the risk level includes a calculation relating to a crimp index or aspect ratio.

[0014] In some embodiments, each risk level is associated with a corresponding display pattern, and the method further comprises applying the corresponding display pattern to each area at risk for a surgical complication.

[0015] Advantageously, the method further comprises the step of displaying the implantation model on the visualization unit applying a display pattern corresponding to an area at risk of surgical complications.

[0016] Another aspect of the invention relates to a method for planning abdominal aortic aneurysm repair surgery using the areas at risk of surgical complications determined by the method of any of the preceding claims. As already mentioned, the method for simulating the interaction of an endoprosthesis with a portion of a patient's aorta can be used to test different models of endoprostheses and select the model that minimizes the risk of surgical complications.

[0017] Another aspect of the invention relates to a device comprising a processor configured to perform the aforementioned method.

[0018] According to one embodiment, a device for simulating the interaction of a portion of a patient's aorta with a specialized endoprosthesis to be implanted in the portion of the aorta of the patient to determine areas at risk for surgical complications comprises: - at least one input section; - at least one processor; at least one output section; Equipped with the at least one input configured to receive a 3D image of a portion of the patient's aorta; The at least one processor: - reconstructing a deformable three-dimensional digital model of a portion of the patient's aorta, referred to as an aortic model, based on the 3D images; - placing the three-dimensional digital model of the endoprosthesis in its planned position in the aortic model, the three-dimensional model of the endoprosthesis and the aortic model together forming an implantation model; - simulating the interaction based on calculations using the implantation model; - determining at least one risk indicator of a surgical complication in a predefined region in the implantation model based on results of the simulated interaction; The at least one output unit - configured to provide at least one risk indicator of a surgical complication in said predefined region.

[0019] Another aspect of the invention relates to a computer program product comprising instructions that, when executed by a computer, cause the computer to carry out the method described above.

[0020] Another aspect of the invention relates to a surgical method for repairing an abdominal aortic aneurysm in a subject, including the preliminary step of performing the method described above.

[0021] The present disclosure further resides in a non-transitory program storage device readable by a computer and tangibly embodying a program of instructions executable by the computer to perform a simulating method according to the present disclosure.

[0022] Such non-transitory program storage devices may be, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor devices, or any suitable combination of the above. More specific examples are provided below, but it should be noted that this is merely an illustrative and not exhaustive list, as will be readily understood by those skilled in the art: portable computer diskettes, hard disks, ROM, Erasable Programmable ROM (EPROM), or flash memory, portable Compact-Disc ROM (CD-ROM).

[0023] definition In the present invention, the following terms have the following meanings:

[0024] In this disclosure, the terms "adapted" and "configured" are used broadly to encompass the initial configuration of the device, subsequent adaptation or supplementation, or any combination thereof, whether done through material means or software means (including firmware).

[0025] The term "processor" should not be construed as limited to hardware capable of executing software, but generally refers to a processing device that may include, for example, a computer, a microprocessor, an integrated circuit, or a programmable logic device (PLD). A processor may also encompass one or more graphics processing units (GPUs), whether utilized for computer graphics and image processing or other functions. Furthermore, instructions and / or data enabling the execution of the associated and / or resulting functions may be stored on any processor-readable medium, such as, for example, an integrated circuit, a hard disk, an optical disk such as a compact disc (CD) or a digital versatile disc (DVD), a random-access memory (RAM), or a read-only memory (ROM). Instructions may be stored in hardware, software, firmware, or any combination thereof, among others.

[0026] "Machine learning (ML)" traditionally refers to computer algorithms that are automatically improved through experience based on training data, allowing for the tuning of computer model parameters to reduce the gap between the expected output extracted from the training data and the evaluation output calculated by the computer model. [Brief explanation of the drawings]

[0027] [Figure 1] 1 illustrates a device that simulates interaction with a specialized endoprosthesis implanted in a portion of a patient's aorta. [Figure 2] 1 is an example flowchart illustrating a method for simulating interaction with a specialized endoprosthesis implanted in a portion of a patient's aorta. [Figure 3a] FIG. 10 is a diagram showing a cross section of the implementation model in use. [Figure 3b] FIG. 10 is a diagram showing multiple two-dimensional cross sections of an implementation model in use. [Figure 4]FIG. 1 illustrates a created manifest obtained from an original manifest. [Figure 5] 1 is an example flowchart illustrating a method for calculating geometric parameters used to determine at least one risk index. [Figure 6] FIG. 1 shows two curves representing the mean aortic radius and the mean graft radius, respectively, along a portion of the centerline of a patient's aorta. [Figure 7] 1 is an example of an in-use implementation model illustrating risk indicators associated with oversizing of an implanted endoprosthesis. [Figure 8] 1 is an example of an in-use implementation model illustrating risk levels associated with the quality of crimping (ie, crimp index) of an implanted endoprosthesis. [Figure 9] 1 is an example of an in-use implementation model illustrating the risk level associated with the quality of fixation of the deployed endoprosthesis. [Figure 10a] FIG. 10 shows an in-use implementation model corresponding to the first endoprosthesis model, illustrating the risk of endoleak. [Figure 10b] FIG. 10 shows an in-use implementation model corresponding to the second endoprosthesis model, illustrating the risk of endoleak. DETAILED DESCRIPTION OF THE INVENTION

[0028] The present invention relates to a computer-implemented method 100 for simulating the interaction of a portion of a patient's aorta with a specialized endoprosthesis to be implanted in the portion of the patient's aorta to determine areas at risk for surgical complications.

[0029] Examples of surgical complications are endoleak, endoprosthesis migration, endoprosthesis invagination, endoprosthesis occlusion, thrombosis, and endoprosthesis kinking.

[0030] The method 100 may be implemented in a device 200 as shown in FIG.

[0031] FIG. 1 illustrates a device 200 for simulating the interaction of a portion of a patient's aorta with a specialized endoprosthesis to be implanted in the portion of the patient's aorta to determine areas at risk for surgical complications, according to an embodiment of the present invention.

[0032] In these embodiments, device 200 comprises a computer having a memory 201 for storing program instructions loadable into circuitry that, when executed by circuitry 202, are adapted to cause circuitry 202 to perform the steps of the methods of Figures 2 and 3. Memory 201 may also store data and information useful for performing the steps of the present invention, as previously described.

[0033] The circuit 202 may include, for example: a processor or processing unit adapted to interpret instructions in a computer language and comprising, associated with or capable of being attached to a memory containing the instructions, or - a processor / processing unit adapted to interpret instructions in a computer language and associated memory containing said instructions, or an electronic card in which the steps of the invention are written in silicon, or - Programmable electronic chips such as FPGA (Field-Programmable Gate Array) chips, It could be.

[0034] The computer may also include an input interface 203 for receiving input data and an output interface 204 for providing output data. Examples of input and output data are provided below.

[0035] To facilitate interaction with the computer, a screen 205 and keyboard 206 may be provided and connected to the computer circuitry 202 .

[0036] 2 is a flowchart illustrating example steps performed to perform computer-implemented method 100. It is assumed that an endoprosthesis will be implanted in a portion of a patient's aorta during an intervention, and that method 100 determines a potential risk of surgical complications based on a numerical simulation of the implantation of the endoprosthesis in the portion of the patient's aorta. Method 100 is typically performed prior to the intervention. Once the endoprosthesis is implanted in the portion of the patient's aorta, the endoprosthesis interacts with the portion of the patient's aorta. Method 100 includes the step of numerically simulating this interaction, as described below.

[0037] In step S1, a 3D image of a portion of a patient's aorta is received. For example, the 3D image may be a preoperative CT scan image. The 3D image may be input through the input interface 203 of the device 200.

[0038] In step S2, a deformable three-dimensional digital model of a portion of the patient's aorta is reconstructed using the 3D image. For example, the reconstruction is performed by circuitry 202 of device 200. The reconstructed deformable three-dimensional digital model is referred to as an aortic model. For example, when used in numerical simulations, techniques such as those described in International Patent Application WO2018 / 073505A1 can be used to simulate deformation of the aortic model.

[0039] In step S3, a numerical simulation of the implantation of the endoprosthesis into a portion of the patient's aorta is performed as follows.

[0040] In a first substep S31, the three-dimensional digital model of the endoprosthesis is placed in a planned position in the aortic model. The three-dimensional digital model of the endoprosthesis and the aortic model together form an implantation model. For example, the planned position corresponds to a position determined by the medical personnel in charge of the intervention. This placing step is performed by aligning the three-dimensional digital model of the endoprosthesis, preferably the proximal end of the three-dimensional digital model of the endoprosthesis, with the planned position in the aortic model.

[0041] In a second substep S32, the implant model is used to simulate the interaction of the endoprosthesis with a portion of the patient's aorta.

[0042] In a third sub-step S33, based on the results of the simulated interaction, some regions of the patient's aorta are determined to be at risk for surgical complications.

[0043] Substeps S31, S32, and S33 of step S3 may be performed by circuitry 202 of device 200.

[0044] According to some embodiments, in a second substep S32, the interaction is simulated by calculating the contact forces between the endoprosthesis and the aortic model using finite element analysis and then applying the calculated contact forces to a three-dimensional digital model of the endoprosthesis to generate an in-use representation of the endoprosthesis, while the aortic model is deformed accordingly. The generated in-use representation of the endoprosthesis is called the deployed model. The deployed model and the deformed aortic model together form the so-called in-use implant model.

[0045] In some embodiments, the third substep S33 may include calculating at least one risk index. For example, a global risk index may be calculated. In another example, a map of risk indexes is calculated, each risk index corresponding to a location on the implant model in use. The at least one risk index may be calculated based on an index calculated from statistical data from geometric parameters.

[0046] In some embodiments, the geometric parameters are calculated as follows: Two-dimensional cross sections (or slices) of the implanted model in use are calculated. A cutting plane corresponds to each cross section. Such a cutting plane P is illustrated in FIG. 3a. For example, the aortic deformed model comprises a centerline. In this case, two-dimensional cross sections can be calculated along the centerline, as illustrated by curves C1, C2, C3, C4, and C5 in FIG. 3b. These curves represent the intersection of the cutting planes at various points on the centerline of the aortic deformed model with the implanted model in use. For example, the deployment model comprises a stent portion and a graft portion. In a given two-dimensional cross section, a first set of points corresponds to a section of the aortic model at the corresponding cutting plane, a second set of points corresponds to a section of the graft portion at the corresponding cutting plane, and a third set of points corresponds to a section of the stent portion. The distance from the first set of points to the corresponding point on the centerline is represented by RAi, the distance from the second set of points to the corresponding point on the centerline is represented by RGj, and the distance from the third set of points to the corresponding point on the centerline is represented by RSk. 4 shows how the distances RAi and RGj are calculated. The initial diameter of the three-dimensional digital model of the endoprosthesis (unplaced) is represented by RG. The geometric parameters can then be calculated from the two-dimensional cross-sections by performing a calculation procedure 300. An example of the calculation procedure 300 is shown in FIG. 5 and described below. The calculation procedure 300 can be performed, for example, by the circuit 202 of the device 200, and the data used during the calculation procedure 300 can be stored in the memory 201 of the device 200.

[0047] In step 301, a list of distances RAi, RGj and RSk is stored, for example in the memory 201 of the device 200.

[0048] In step 302, the distances RAi, RGj, and RSk are normalized by dividing by RG0 to obtain a list of so-called normalized aortic radii RAnormi, a list of so-called normalized graft radii RGnormj, and a list of so-called normalized stent radii RSnormk.

[0049] In step 303, statistical data is calculated from the lists stored and / or obtained in steps 301 and 302. The statistical data may include an average, a maximum, a minimum, or a standard deviation.

[0050] After the calculation step 300, a list of indices is calculated which are used to calculate at least one risk index.

[0051] For example, a list of indices I1 to I14 that can be derived from the data stored and determined in steps 301 and 302 is as follows: - I1: crimping between the aorta and the endoprosthesis, - I2: aortic, stent, and graft conicity; - I3: Aorta, stent, and graft straightness; - I4: Circularity of the aorta, stents, and grafts; - I5: Variation in diameter of aorta, stent, and graft; - I6: Aorta, stent, and graft stresses; - I7: pressure of the endoprosthesis against the aorta, - I8: cross-sectional area of ​​the aorta and graft, - I9: roughness of the aorta and graft, - I10: Aortic and endoprosthesis angles, - I11: Stent oversizing, - I12: Calcified surface, - I13: Surface of thrombus, - I14: Patient characteristics (age, medical history). For example, the indicator may be calculated from data stored in step 301 and / or obtained by circuitry 202 of device 200 in step 302 .

[0052] In this disclosure, the term "crimping" refers to the fitting (e.g., alignment, positioning, or placement) of the endoprosthesis within the aorta, and more specifically, the contact or diameter difference between the aorta and the graft or the aorta and the stent.

[0053] The at least one risk index may be calculated from the index in a variety of ways.

[0054] For example, at least one risk index can be calculated by linear or quadratic discriminant analysis. Such analyses are statistical analyses that classify a dataset into predefined groups. A scatter plot represents each group in an n-dimensional space, where n is the total number of indexes. The distance between each point and the centroid of each group is used to predict whether a case belongs to a group. Decision surfaces are calculated to separate the different groups. These surfaces can be more or less complex (linear, quadratic, etc.). This method is similar to a learning technique. During the learning process of known cases, various decision surfaces in the n-dimensional space are defined, which make it possible to calculate the probability that a new case belongs to one of the groups.

[0055] In another example, the at least one risk indicator can be calculated by a machine learning classification algorithm. The machine learning classification algorithm can be trained to output the at least one risk indicator based on classifying an unclassified dataset into predefined groups. The machine learning classification algorithm can be trained with a training set of indicators. An example of a machine learning classification algorithm is a support vector machine. In this algorithm, the predefined groups are defined on a parameter space. To delimit the groups, the trained dataset is differentiated in the parameter space. To aid in this differentiation, a kernel function is applied to the training dataset. This kernel function can be a linear, sigmoid, or radial basis function. To separate the different groups, boundaries are calculated in the parameter space. When a new dataset is placed in this space, it is classified into the predefined groups.

[0056] In yet another example, the at least one risk index may be calculated by a weighted average, for example, the at least one risk index may be calculated as a statistical value of a weighted average of several of the indexes.

[0057] A simplified version of this latter weighted average method is to calculate a unique weighted average of all indicators.

[0058] In a typical example where the surgical complication determined by method 100 is a Type Ia endoleak, the relevant index is index I1 (also called the crimp index), i.e., the crimp between the aorta and the endoprosthesis. Crimping can be calculated in various ways. In the first method, the sum of the differences between the graft radius and the aortic radius can be calculated. Figure 6 shows two curves representing the mean aortic radius and the mean graft radius along a portion of the aortic centerline. Here, the average of all values ​​within the cross section corresponding to a point on the aortic centerline is calculated. In the second method, crimping is calculated as the longest area with a lack of crimping. In the third method, crimping is calculated as the length of the crimp that is longer than a given criterion.

[0059] In a second example, an alternative related metric that is calculated is the aspect ratio, which may be calculated as the ratio of the difference between the maximum and minimum normalized radii of the aorta and the maximum and minimum normalized radii of the graft after the simulated placement.

[0060] In some embodiments, results of method 100 for simulating the interaction of a portion of a patient's aorta with a specialized endoprosthesis implanted therein to determine areas at risk for surgical complications can be received, collected, and stored by a web platform. The web platform can enable visualization of the results. For example, the web platform can enable visualization of a deployment model and an aortic deformation model, which together form a so-called in-use implant model, and visualization of areas at risk for surgical complications with corresponding display patterns.

[0061] According to one embodiment, each risk level (for one region / zone) is associated with a corresponding display pattern. The display pattern may be, for example, a color selected from a color scale, or a texture pattern including, for example, dots, stars, or any other marks covering the region / zone. The method includes applying the corresponding display pattern to each region / zone for which a risk of surgical complications has been calculated. Furthermore, the method may further be configured to display the implantation model on a visualization unit, whereby regions for which a risk of surgical complications has been calculated are visualized with the corresponding display pattern, as shown in Figures 7, 8, and 9.

[0062] In some embodiments, the risk index (i.e., risk level) is comprised of a percentage ranging from 0 to 100%, with 100% representing a high risk level. FIG. 7 shows an example of an in-use implant model in which regions with different risk indexes are found. For example, this in-use model can be visualized on the screen 205 of the device 200. In the visualization example of FIG. 7, the risk index (i.e., representing the risk level) is related to the oversizing criteria of the implanted endoprosthesis and is determined by calculating an oversizing factor, such as index I11. For example, region A1 of the stent colored dark blue (i.e., of the corresponding display pattern) corresponds to 0-5% oversizing, region A2 of the stent colored light blue corresponds to 5-10% oversizing, and region A3 of the stent colored yellow corresponds to 20-25% oversizing.

[0063] In one embodiment, determining the risk level of a region includes calculating a crimp index and / or an aspect ratio. In some embodiments, a risk level is assigned to a region depending on the value of a risk index associated with the region of the implant model in use. For example, if the risk index associated with a region is below a first threshold, the region is assigned a low risk level. If the risk index associated with a region is between the first and second thresholds, the region is assigned a medium risk level. If the risk index associated with a region is higher than the second threshold, the region is assigned a high risk level.

[0064] In some embodiments, the risk index can be associated with a particular characteristic regarding the interaction of the endoprosthesis with the portion of the patient's aorta. For example, the risk index can relate to the quality of crimping of the endoprosthesis in the portion of the patient's aorta. FIG. 8 shows an example of an in-use implantation model that can visualize the quality level of crimping of several regions. For example, this in-use model can be visualized on the screen 205 of the device 200. For example, in the region / zone B1 of the stent colored green (i.e., of the corresponding display pattern), the crimping of the endoprosthesis is good. In the region B2 colored red, there is a lack of crimping. In another example, the risk index can relate to the quality of fixation of the endoprosthesis in the portion of the patient's aorta. FIG. 9 shows an example of an in-use implantation model that can visualize the quality level of fixation of the endoprosthesis in several regions. For example, this in-use model can be visualized on the screen 205 of the device 200. For example, the region / zone D1 of the stent colored green (i.e., of the corresponding display pattern) indicates good fixation. The area referenced by D2, colored yellow, shows weak fixation, while the area referenced by D3, colored red, shows no fixation.

[0065] Example The present invention is further illustrated in the following examples. Example 1: Prediction of Type Ia Endoleak After Endovascular Aneurysm Repair (EVAR)

[0066] In this example, a retrospective cohort study was performed. The study population consisted of all patients who underwent implantation of an Endurant® AAA endovascular graft (Medtronic, Dublin, Ireland) at two large aortic centers between January 2012 and September 2017 and who presented with an early or late (up to 5 years) type Ia endoleak. Control cases consisted of all patients without an early or late type Ia endoleak who underwent implantation of the same graft during the same period and had a minimum follow-up of 5 years.

[0067] Preoperative CT scans were analyzed on a dedicated workstation equipped with centerline extraction. 3D and centerline reconstructions were used to measure the maximum diameter of the AAA and proximal neck characteristics. Inadequate neck anatomy was defined as one or more of the following: (1) neck length ≤15 mm (defined as the length at which the aortic diameter falls within 10% of the infrarenal diameter), (2) a conical or tapered neck using the "neck index" introduced by Albertini et al., (3) perineck thrombus or significant calcification, (4) a suprarenal angle ≥45°, and (5) an infrarenal angle ≥60°. The Endurant II / IIs Instructions for Use (IFU) defines the following neck characteristics: (1) a proximal neck length ≥10 mm (defined as the length at which the aortic diameter falls within 10% of the infrarenal diameter), (2) an infrarenal angle <60°, and (3) an aortic neck diameter ranging from 19 to 32 mm. All patients were then classified as hostile and outside the IFU criteria.

[0068] The computational technique was based on finite element analysis, which assessed the deformations caused by device-host interaction and provided predictions of graft behavior and arterial displacement. This study was based on the applicant's proprietary algorithms developed to simulate endovascular device deployment in a patient-specific aortic aneurysm model. Because the purpose of this study focused on EVAR proximal fixation, only main graft deployment was simulated in the model.

[0069] High-resolution CT scans were used to generate patient-specific 3D models of the aorta from the suprarenal aorta to the external iliac arteries, including the renal arteries (aortic segmentation step), and assigned orthotropic elastic behavior. Because aortic wall calcification and intraluminal thrombus were not modeled, the aortic wall, iliac arteries, and renal arteries were considered homogeneous surfaces with constant surfaces of 1.5 mm, 1 mm, and 1 mm, respectively. The unexpanded main Endurant II / IIs® device was digitally reproduced based on the actual characteristic design of the endograft deployed in the corresponding patient. The mechanical properties of the stent graft fabric and stent rings were obtained from the literature or from in-house mechanical testing of samples obtained from the manufacturer. Hyperelastic material behavior was used for the stent rings, and the fabric was modeled as an orthotropic elastic material. Oversizing of the stent rings was taken into account. The commercially available Abaqus / Explicit2018 finite element solver (Dassault Systèmes, Paris, France) was used to perform personalized stent-graft placement into a corresponding model of the aorta. The placement level was achieved in an idealized situation just below the low renal artery, blinded to the actual perioperative EVAR placement. Simulation techniques evaluated the deformations caused by device-host interaction and obtained predictions of stent-graft behavior and arterial displacement. This methodology includes a step called "morphing," presented in International Patent Application WO2018 / 073505A1, during which the aortic surface was numerically deformed until a cylindrical shape was obtained, allowing the stent-graft to be placed inside. Then, reverse deformation of the aorta was achieved while maintaining the stent-graft in place.

[0070] The simulation results measured the mean oversizing and stent graft crimping assessment of the first two covered stents in the main stent graft.

[0071] The aortic and graft segments were measured in 0.5 mm increments at the level of the first two covered stent segments of the stent-graft. For each segment, the normalized representative radius of the graft and aorta was calculated to avoid systematic bias in diameter.

[0072] The "aspect ratio" was evaluated as the ratio of the difference between the maximum and minimum normalized radius of the aorta after the placement simulation to the difference between the maximum and minimum normalized radius of the graft.

[0073] The crimping was evaluated by the difference between the area under the curve of the representative radius of the aorta and the area under the curve of the representative radius of the graft.

[0074] The risk index was calculated based on the aspect ratio index and the crimp coefficient by the previously described method 100. Quadratic discriminant analysis was used to classify the cases into two groups: those at risk of type Ia endoleak and those without.

[0075] Due to small numbers, normality could not be assumed for most of the data. Therefore, descriptive data were presented as medians with interquartile ranges (Q1, Q3). Statistical differences in indicators were assessed using the nonparametric Mann-Whitney U test for continuous variables and the χ2 test or Fisher's exact test for categorical variables. Statistical significance was assumed to be p<0.05, and all reported p-values ​​were two-sided. All statistical analyses were performed with R statistical software, version 4.2.2 (A language and environment for statistical computing, R foundation for Statistical Computing, Vienna, Australia).

[0076] result We analyzed 162 cases with sufficient follow-up, identifying 15 type Ia endoleak cases and 14 control cases. The mean follow-up was 1.9 years (range, 0-8.7 years) in the type Ia endoleak group and 6.8 years (range, 5.2-10.3 years) in the control group. In the type Ia endoleak group, 8 cases (53.3%) had perioperative or early type Ia endoleaks, and 7 cases (46.7%) had late type Ia endoleaks (range, 8 months-5 years). Table 2 shows a comparison of actual observations of type Ia endoleaks with predictions by Method 100 for 15 type Ia endoleak cases and 14 control cases. A black circle indicates the presence of a type Ia endoleak, and an X indicates the absence of a type Ia endoleak.

[0077] [Table 1]

[0078] Example 2: Supporting Endoprosthesis Selection During the Planning Stage By determining areas at risk for surgical complications, method 100 can aid in selecting an appropriate endoprosthesis that minimizes the risk of adverse events. FIG. 10a illustrates an example of an in-use implantation model in which a model of an Endurant II device, type ETBF2516C166EE, a standard-sized device, has been implanted within a portion of a patient's aorta. Region E1 in the in-use implantation model has been determined to be at high risk for endoleak. FIG. 10b illustrates a second example of an in-use implantation model in which another model of an Endurant II device, type ETBF2816C166EE, has been implanted within a portion of the same patient's aorta. Region E2 of this other device has been determined to be at low risk for endoleak.

Claims

1. 1. A computer-implemented method (100) for simulating the interaction of a portion of a patient's aorta with a specialized endoprosthesis to be implanted therein to determine areas at risk for surgical complications, the method comprising: receiving (S1) a 3D image of a portion of the patient's aorta; (S2) reconstructing a deformable three-dimensional digital model of a portion of the patient's aorta based on the 3D image, the reconstructed deformable three-dimensional digital model being referred to as an aortic model; placing (S31) a three-dimensional digital model of the endoprosthesis at its planned position in the aortic model, whereby the three-dimensional model of the endoprosthesis and the aortic model together form an implantation model; a step (S32) of simulating the interaction based on calculations using the implantation model; determining (S33) at least one risk indicator of a surgical complication in a predefined region in the implantation model based on the results of the simulated interaction; providing as output at least one risk indicator of a surgical complication in said predefined region; A method comprising:

2. 10. The computer-implemented method of claim 1, wherein the surgical complication comprises at least one of endoleak, endoprosthesis migration, endoprosthesis invagination, endoprosthesis occlusion, thrombosis, and endoprosthesis kinking.

3. 3. The computer-implemented method (100) of claim 1 or claim 2, wherein the calculations relating to the implantation model include calculations of geometric deformations of a three-dimensional digital model of the endoprosthesis within the aortic model.

4. The computer-implemented method (100) of claim 3, wherein the calculation of the geometric deformation is performed by finite element analysis.

5. 10. The computer-implemented method (100) of any preceding claim, wherein the 3D images are pre-operative CT scan images.

6. 10. The computer-implemented method (100) of any preceding claim, wherein the interaction is endovascular repair of an aortic aneurysm.

7. 10. The computer-implemented method (100) of any preceding claim, further comprising, after determining the areas at risk of a surgical complication, determining a risk level for each area at risk of a surgical complication.

8. 8. The computer-implemented method of claim 7, wherein determining the risk level includes calculating a crimp index or aspect ratio representative of crimping between the aortic model and the three-dimensional digital model of the endoprosthesis.

9. 9. The computer-implemented method (100) of claim 7 or claim 8, wherein each risk level is associated with a corresponding display pattern, the method further comprising applying the corresponding display pattern to each area at risk of surgical complications.

10. 10. The computer-implemented method (100) of claim 9, further comprising displaying the implantation model on the visualization unit along with the areas at risk for surgical complications with the corresponding display pattern applied.

11. A method for planning abdominal aortic aneurysm repair surgery using an area at risk of surgical complications determined by the method of any one of claims 1 to 10.

12. A device comprising a processor configured to perform the method of any one of claims 1 to 10.

13. A computer program product comprising instructions which, when executed by a computer, cause the computer to carry out the method of any one of claims 1 to 10.

Images