Methods for predicting the springback coefficient of microcatheter shaping, as well as shaping methods, equipment, and storage media.

Abstract

This invention discloses a method for predicting the rebound coefficient of microcatheter shaping, along with its shaping method, device, and storage medium. The method includes acquiring historical case data, with each case data point including preoperative images, microcatheter model information, shaping needle shape, intraoperative images, and physician operation parameters. Based on preoperative images, vascular morphology features are extracted to generate an ideal tip shaping shape. Based on intraoperative images, the actual tip shape is constructed, using the shaping needle shape as the pre-rebound shape and the actual shape as the post-rebound shape, to calculate the true rebound coefficient. The vascular morphology features, ideal tip shaping shape, microcatheter model information, and physician operation parameters are combined into a feature vector to construct a training set and train a machine learning model. For new patients, a feature vector is generated based on their preoperative data, input into the model, and a personalized rebound coefficient is output. This invention integrates multidimensional clinical factors to achieve high-precision prediction, overcoming shaping biases caused by reliance on fixed nominal values ​​or subjective experience.

Description

Technical Field

[0001] This invention relates to the field of microcatheter shaping technology, and in particular to a method for predicting the rebound coefficient of microcatheter shaping, as well as the shaping method, equipment and storage medium thereof. Background Technology

[0002] In interventional procedures, microcatheters often require external shaping, pre-bending their tips into a specific shape to facilitate delivery to the target blood vessel. This is particularly crucial in intracranial aneurysm embolization surgery, where microcatheter shaping is a key step in ensuring successful entry into the aneurysm cavity, providing stable support, and safely releasing the coils. The shaping process typically involves using a metal shaping needle: the physician first inserts a straight shaping needle into the microcatheter lumen, then bends the microcatheter and shaping needle together into the desired shape, such as a J-shape, S-shape, or spiral. The microcatheter is then heated to set its shape under the support of the shaping needle; after cooling, the shaping needle is removed, and the microcatheter retains its pre-set shape.

[0003] When a microcatheter is inserted into the blood vessel environment, it undergoes partial deformation due to the elastic recovery of the material and the constraint of the blood vessel wall. The final shape differs from the shape determined in vitro; this phenomenon is called rebound. To quantify this effect, clinicians typically use the nominal rebound coefficient provided by the microcatheter manufacturer as a fixed parameter to guide shaping. Of course, physicians can also make certain adjustments to this rebound coefficient based on their own clinical experience. However, the actual rebound behavior of microcatheters is influenced by a variety of complex and individualized clinical factors. Relying solely on a uniform nominal value or subjective experience for shaping design often leads to the microcatheter tip shape deviating from the expected shape after placement, requiring repeated adjustments or even catheter replacement. This not only reduces surgical efficiency but may also increase the risk of vascular injury. Summary of the Invention

[0004] Based on this, the present invention addresses the above-mentioned technical problems by providing a method for predicting the springback coefficient of microcatheter shaping, as well as a shaping method, device, and storage medium.

[0005] On one hand, the present invention provides a method for predicting the elasticity coefficient of microcatheter shaping, the method comprising:

[0006] Acquire several historical case data, each of which includes preoperative vascular images, microcatheter model information, the shape of the shaping needle used for extracorporeal shaping of the microcatheter, intraoperative vascular images, and physician operation parameters related to extracorporeal shaping and in vivo delivery of the microcatheter.

[0007] Based on preoperative vascular images, vascular morphological features are extracted and the ideal tip shape of the microcatheter in the blood vessel is generated. Based on intraoperative vascular images, the actual tip shape of the microcatheter in the blood vessel is constructed. The shape of the shaping needle is used as the pre-rebound shape and the actual tip shape is used as the post-rebound shape. The true rebound coefficient is calculated.

[0008] The vascular morphology features, the microcatheter shaping features corresponding to the ideal tip shaping shape, the microcatheter model information, and the doctor's operation parameters are combined into a feature vector to generate a training dataset consisting of the feature vector and its corresponding true rebound coefficient.

[0009] A machine learning model is trained based on the training dataset to obtain a well-trained rebound coefficient prediction model.

[0010] For patients to be predicted, vascular morphology features are extracted based on their preoperative vascular images to generate the ideal tip shape of the microcatheter. The feature vector is then combined with the microcatheter model information and the doctor's operation parameters to form a feature vector, which is input into the trained rebound coefficient prediction model and outputs the predicted rebound coefficient.

[0011] In one embodiment, the vascular morphological features include at least one of vascular curvature features, vascular torsion features, vascular size features, vascular angle features, and aneurysm features.

[0012] In one embodiment, the actual tip morphology of the microcatheter constructed within the blood vessel based on intraoperative vascular imaging includes:

[0013] The two intraoperative two-dimensional DSA images from different perspectives after the microcatheter is placed into the aneurysm cavity are referred to as the first DSA image and the second DSA image.

[0014] Microcatheter segmentation was performed on the first DSA image and the second DSA image respectively to obtain the first segmented image and the second segmented image;

[0015] Skeleton extraction is performed on the first segmented image and the second segmented image respectively to obtain the first skeleton curve and the second skeleton curve;

[0016] Using a binocular vision 3D reconstruction algorithm, the first skeleton curve and the second skeleton curve are corrected at three points and restored in three dimensions to obtain the real three-dimensional shape of the microcatheter in the blood vessel, which serves as the actual tip shape.

[0017] In one embodiment, the true rebound coefficient is calculated in the following manner:

[0018] The pre-rebound and post-rebound shapes are length-aligned and sampled at equal intervals. The tangent vectors at the near and far sampling points are calculated, and the angle between the two tangent vectors is obtained to obtain the first angle and the second angle. The ratio of the first angle to the second angle is used as the true rebound coefficient.

[0019] In one embodiment, the microcatheter shaping features include at least one of shaping length features, shaping curvature features, and shaping twist features for an ideal tip shaping morphology.

[0020] In one embodiment, the physician operating parameters related to the in vitro shaping of the microcatheter include at least one of the heating method, heating temperature, heating time, cooling time, and insertion depth of the shaping needle used when heating and shaping the microcatheter in vitro.

[0021] On the other hand, the present invention provides a method for shaping microcatheters, the method comprising:

[0022] The above-described method for predicting the rebound coefficient of microcatheter shaping is applied to obtain the ideal microcatheter tip shaping morphology and the corresponding predicted rebound coefficient.

[0023] Based on the ideal tip shaping shape and the predicted rebound coefficient, the shape of the shaping needle required for in vitro shaping of the microcatheter is calculated in reverse, and the shaping needle is used to shape the microcatheter in vitro.

[0024] In one embodiment, the reverse calculation of the shaping needle shape required for in vitro shaping of the microcatheter includes:

[0025] By equidistantly dispersing the ideal head-end shaping shape, multiple rod elements are obtained;

[0026] Calculate the included angle between adjacent rod elements, and calculate the overmolding angle based on the predicted springback coefficient;

[0027] Determine the rotation axis vectors of adjacent rod elements, and generate a rotation matrix based on the overmolding angle and the rotation axis vectors;

[0028] The shape of the shaping needle is constructed by rotating each rod element sequentially using a rotation matrix.

[0029] In another aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0030] In another aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the steps of the above-described method.

[0031] Compared with existing technologies, this invention constructs a multi-dimensional feature vector that integrates vascular morphology features, ideal tip shaping of the microcatheter, microcatheter model information, and physician operation parameters. Based on the actual rebound coefficient calculated from the actual tip shape constructed by shaping needle shape and intraoperative vascular images in historical cases, the model is trained to achieve high-precision prediction of personalized rebound coefficient for new patients. This effectively overcomes the shaping deviation problem caused by existing technologies relying on fixed rebound coefficients or subjective experience, and provides a reliable basis for precise microcatheter shaping. Attached Figure Description

[0032] Figure 1This is a flowchart illustrating a method for predicting the elasticity coefficient of microcatheter shaping in one embodiment.

[0033] Figure 2 This is a schematic diagram of the microcatheter tip before its rebound in one embodiment.

[0034] Figure 3 This is a schematic diagram of the shape of the microcatheter tip after rebound in one embodiment. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0036] The following detailed description of the embodiments of the present invention uses microcatheter shaping during intracranial aneurysm interventional embolization surgery as an example. Figure 1 As shown, the microcatheter shaping and rebound coefficient prediction method of this embodiment includes the following steps:

[0037] Step S100: Obtain several historical case data. Each case data includes preoperative vascular images, microcatheter model information, the shape of the shaping needle used for extracorporeal shaping of the microcatheter, intraoperative vascular images, and physician operation parameters related to extracorporeal shaping and in vivo delivery of the microcatheter.

[0038] In this step, historical case data from several completed surgeries are acquired and integrated. Each case data includes five key types of information: preoperative vascular imaging, microcatheter model information, the shape of the shaping needle used for in vitro microcatheter shaping, intraoperative vascular imaging, and physician operation parameters closely related to the in vitro microcatheter shaping and in vivo delivery process. Preoperative vascular imaging can be derived from routine clinical imaging modalities such as 3D CTA (3D CT angiography), MRA (magnetic resonance angiography), or DSA (digital subtraction angiography). Based on this image, vascular structures can be automatically segmented using medical image processing technology, and a complete three-dimensional geometric model including the carrier artery and aneurysm can be reconstructed. Furthermore, various vascular morphological features are extracted, including: vascular curvature features (such as maximum curvature and average curvature), vascular torsion features (such as maximum torsion and average torsion), vascular size features (such as average diameter and minimum diameter), vascular angle features (such as aneurysm tilt angle and branch angle), and aneurysm characteristics (such as aneurysm location, maximum diameter of the aneurysm, and neck width). These anatomical parameters collectively reflect the complexity of the path by which the microcatheter enters the target area and its adaptation requirements to vascular compliance, thus directly affecting its rebound behavior.

[0039] The microcatheter model information includes at least one of the following: microcatheter brand, outer diameter, inner diameter, material hardness, plasticity grade, and the manufacturer's nominal resilience coefficient. The manufacturer's nominal resilience coefficient can be used as a reference characteristic.

[0040] Intraoperative vascular imaging typically consists of two 2D DSA images taken from different projection angles after the microcatheter has been successfully positioned in the aneurysm cavity. These images are used to subsequently reconstruct the actual spatial morphology of the microcatheter within the body.

[0041] The physician's operational parameters are divided into two parts: one is the parameters directly related to external shaping, including the heating method (such as steam, hot air, or electric heater), heating temperature, heating time, cooling time, and the insertion depth of the shaping needle; the other is the parameters related to internal delivery, such as the microcatheter advancement speed. These two types of parameters together characterize the physical intervention applied to the microcatheter by the physician during the procedure and are variables that cannot be ignored in affecting the rebound effect.

[0042] Step S200: Based on preoperative vascular images, extract vascular morphological features and generate the ideal tip shaping shape of the microcatheter in the blood vessel. Based on intraoperative vascular images, construct the actual tip shape of the microcatheter in the blood vessel, and use the shape of the shaping needle as the pre-rebound shape and the actual tip shape as the post-rebound shape to calculate the true rebound coefficient.

[0043] This step first involves reconstructing a three-dimensional vascular geometric model based on preoperative vascular images (such as 3D CTA, MRA, or DSA), extracting vascular morphological features, and then using one or more existing microcatheter shaping algorithms based on the three-dimensional geometric model to obtain the optimal microcatheter tip shaping scheme, i.e., the ideal tip shaping shape of the microcatheter within the blood vessel.

[0044] Simultaneously, based on two 2D DSA images acquired during the operation from different perspectives (denoted as the first DSA image and the second DSA image), a three-dimensional morphological reconstruction process for the microcatheter was performed: the first DSA image and the second DSA image were segmented to obtain a first segmented image and a second segmented image; the first segmented image and the second segmented image were respectively subjected to skeleton extraction to obtain a first skeleton curve and a second skeleton curve; using a binocular vision three-dimensional reconstruction algorithm, the first skeleton curve and the second skeleton curve were corrected at three points and three-dimensionally restored to obtain the true three-dimensional morphology of the microcatheter in the blood vessel, which is used as the actual tip morphology.

[0045] Based on this, the shape of the shaping needle used in in vitro shaping is defined as the "pre-rebound morphology," and the actual head shape obtained from the above reconstruction is defined as the "post-rebound morphology," and the true rebound coefficient is calculated accordingly. The specific calculation method is as follows: the pre-rebound morphology and the post-rebound morphology are length-aligned, and then sampled at equal intervals into a three-dimensional spatial coordinate sequence, denoted as Sequence 1 (see...). Figure 2) and sequence 2 (see Figure 3 ); then, as Figure 2 and Figure 3 As shown, the proximal and distal tangent vectors of sequence 1 and sequence 2 are calculated respectively, and the angle between the two tangent vectors is calculated to obtain the first angle, i.e. angle 1 (before rebound) and the second angle, i.e. angle 2 (after rebound). Finally, the true rebound coefficient is defined as the ratio of angle 1 to angle 2 (i.e., true rebound coefficient = angle 1 / angle 2), which objectively quantifies the degree of morphological change of the microcatheter from in vitro shaping to in vivo release.

[0046] Step S300: Combine the vascular morphology features, the microcatheter shaping features corresponding to the ideal tip shaping shape, the microcatheter model information, and the doctor's operation parameters into a feature vector, and generate a training dataset consisting of the feature vector and its corresponding true rebound coefficient.

[0047] In this step, the extracted features are structurally integrated, including vascular morphology features from preoperative images, microcatheter shaping features derived from the ideal tip shaping morphology (such as shaping length, maximum / average shaping curvature, maximum / average shaping torsion, etc.), microcatheter model information, and physician operation parameters. Among these, the microcatheter shaping features reflect the geometric complexity of the required bending, and because polymer materials exhibit nonlinear elastic responses under large deformations, these features significantly influence rebound behavior. All the above features are concatenated in a preset order to form a high-dimensional feature vector; simultaneously, the true rebound coefficient calculated in step S200 is used as the supervision label corresponding to this feature vector. By traversing all historical cases, a training dataset consisting of "feature vector – true rebound coefficient" sample pairs is finally constructed, providing high-quality supervision signals for the machine learning model.

[0048] Step S400: Train a machine learning model based on the training dataset to obtain a trained rebound coefficient prediction model.

[0049] This step employs machine learning methods to model the training dataset constructed in step S300. Considering that clinical data typically has limited sample size and high feature dimensionality, ensemble learning or kernel method models with good generalization ability and interpretability are preferred. Optional models include, but are not limited to, XGBoost, Random Forest, Gradient Boosting Machine, Support Vector Regression, and CatBoost. The model takes feature vectors as input and the true rebound coefficient as the output target, completing parameter learning by optimizing the loss function (such as mean squared error), and finally outputting a trained rebound coefficient prediction model. This model can capture the complex nonlinear coupling relationship between vascular anatomy, device properties, shaping goals, and operating habits.

[0050] Step S500: For the patient to be predicted, extract vascular morphology features based on their preoperative vascular images and generate the ideal tip shaping shape of the microcatheter. Combine the microcatheter model information and the doctor's operation parameters to form a feature vector, input it into the trained rebound coefficient prediction model, and output the predicted rebound coefficient.

[0051] In this step, when facing a new patient to be predicted, the preoperative vascular images are first acquired, and the processing flow in step S200 is repeated: vascular morphological features are extracted to generate the ideal tip shaping shape of the microcatheter; simultaneously, information on the microcatheter model to be used and the planned physician operating parameters are collected; the above information is combined into a new feature vector and input into the rebound coefficient prediction model trained in step S400; the model then outputs the personalized rebound coefficient prediction value for the patient under the current surgical conditions. This prediction result can be directly used to guide the precise bending of the extracorporeal shaping needle, achieving precise interventional operation of "what is shaped is what is obtained".

[0052] In summary, this embodiment constructs a data-driven rebound coefficient prediction framework by deeply integrating multiple factors such as individualized patient anatomical characteristics, ideal microcatheter shaping targets, instrument physical properties, and physician operating habits. This not only overcomes the shaping deviation problems caused by traditional reliance on fixed nominal values ​​or subjective experience, but also provides reliable technical support for intelligent and personalized neurointerventional surgical planning, significantly improving the success rate of one-time microcatheter placement and surgical safety.

[0053] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise explicitly stated in this document, there is no strict order in which these steps are executed; they can be performed in other orders. Furthermore, Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0054] In one embodiment, the present invention provides a microcatheter shaping method, which, based on the aforementioned microcatheter shaping springback coefficient prediction method, obtains the ideal microcatheter tip shaping shape and the corresponding predicted springback coefficient. Based on this ideal tip shaping shape and the predicted springback coefficient, the required shaping needle shape for in vitro microcatheter shaping is calculated in reverse, and the shaping needle is used to shape the microcatheter in vitro. Specifically, the ideal tip shaping shape is equally spaced to obtain multiple rod units; the included angle between adjacent rod units is calculated, and the overmolding angle is calculated based on the predicted springback coefficient; the rotation axis vector of adjacent rod units is determined, and a rotation matrix is ​​generated based on the overmolding angle and the rotation axis vector; the rotation matrix is ​​used to rotate each rod unit sequentially to reconstruct the shape of the shaping needle, thereby achieving an effective transformation from personalized springback prediction to precise physical shaping.

[0055] In one embodiment, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described microcatheter shaping springback coefficient prediction method or microcatheter shaping method.

[0056] In one embodiment, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described microcatheter shaping springback coefficient prediction method or microcatheter shaping method.

[0057] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0058] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0059] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for predicting the elasticity coefficient of microcatheter shaping, characterized in that, The method includes: Acquire several historical case data, each of which includes preoperative vascular images, microcatheter model information, the shape of the shaping needle used for extracorporeal shaping of the microcatheter, intraoperative vascular images, and physician operation parameters related to extracorporeal shaping and in vivo delivery of the microcatheter. Based on preoperative vascular images, vascular morphological features are extracted and the ideal tip shape of the microcatheter in the blood vessel is generated. Based on intraoperative vascular images, the actual tip shape of the microcatheter in the blood vessel is constructed. The shape of the shaping needle is used as the pre-rebound shape and the actual tip shape is used as the post-rebound shape. The true rebound coefficient is calculated. The vascular morphology features, the microcatheter shaping features corresponding to the ideal tip shaping shape, the microcatheter model information, and the doctor's operation parameters are combined into a feature vector to generate a training dataset consisting of the feature vector and its corresponding true rebound coefficient. A machine learning model is trained based on the training dataset to obtain a well-trained rebound coefficient prediction model. For patients to be predicted, vascular morphology features are extracted based on their preoperative vascular images to generate the ideal tip shape of the microcatheter. The feature vector is then combined with the microcatheter model information and the doctor's operation parameters to form a feature vector, which is input into the trained rebound coefficient prediction model and outputs the predicted rebound coefficient.

2. The method for predicting the elasticity coefficient of microcatheter shaping according to claim 1, characterized in that, The vascular morphological features include at least one of the following: vascular curvature features, vascular torsion features, vascular size features, vascular angle features, and aneurysm features.

3. The method for predicting the elasticity coefficient of microcatheter shaping according to claim 1, characterized in that, The actual tip morphology of the microcatheter constructed within the blood vessel based on intraoperative vascular imaging includes: The two intraoperative two-dimensional DSA images from different perspectives after the microcatheter is placed into the aneurysm cavity are referred to as the first DSA image and the second DSA image. Microcatheter segmentation was performed on the first DSA image and the second DSA image respectively to obtain the first segmented image and the second segmented image; Skeleton extraction is performed on the first segmented image and the second segmented image respectively to obtain the first skeleton curve and the second skeleton curve; Using a binocular vision 3D reconstruction algorithm, the first skeleton curve and the second skeleton curve are corrected at three points and restored in three dimensions to obtain the real three-dimensional shape of the microcatheter in the blood vessel, which serves as the actual tip shape.

4. The method for predicting the elasticity coefficient of microcatheter shaping according to claim 1, characterized in that, The true rebound coefficient is calculated in the following way: The pre-rebound and post-rebound shapes are length-aligned and sampled at equal intervals. The tangent vectors at the near and far sampling points are calculated, and the angle between the two tangent vectors is obtained to obtain the first angle and the second angle. The ratio of the first angle to the second angle is used as the true rebound coefficient.

5. The method for predicting the elasticity coefficient of microcatheter shaping according to claim 1, characterized in that, The microcatheter shaping features include at least one of the following: shaping length feature, shaping curvature feature, and shaping twist feature of the ideal tip shaping morphology.

6. The method for predicting the elasticity coefficient of microcatheter shaping according to claim 1, characterized in that, The physician operation parameters related to the in vitro shaping of microcatheters include at least one of the following when heating the microcatheter in vitro, including the heating method, heating temperature, heating time, cooling time, and insertion depth of the shaping needle.

7. A method for shaping microcatheters, characterized in that, The method includes: Perform the method as described in any one of claims 1 to 6 to obtain the ideal tip shaping morphology of the microcatheter and the corresponding predicted resilience coefficient. Based on the ideal tip shaping shape and the predicted rebound coefficient, the shape of the shaping needle required for in vitro shaping of the microcatheter is calculated in reverse, and the shaping needle is used to shape the microcatheter in vitro.

8. The microcatheter shaping method according to claim 7, characterized in that, The reverse calculation of the shaping needle shape required for in vitro shaping of the microcatheter includes: By equidistantly dispersing the ideal head-end shaping shape, multiple rod elements are obtained; Calculate the included angle between adjacent rod elements, and calculate the overmolding angle based on the predicted springback coefficient; Determine the rotation axis vectors of adjacent rod elements, and generate a rotation matrix based on the overmolding angle and the rotation axis vectors; The shape of the shaping needle is constructed by rotating each rod element sequentially using a rotation matrix.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.

Images