Coronary stent implantation planning method and system based on artificial intelligence

Abstract

The invention provides a coronary stent implantation planning method and system based on artificial intelligence, and belongs to the technical field of implantation planning, and the method comprises the steps: S1, determining whether a coronary stent needs to be implanted or not according to a coronary angiography image of a patient; and S2, in response to an instruction for indicating that the coronary stent needs to be implanted, determining an implantation planning scheme for implanting the coronary stent into the diseased region according to the focus characteristics of the target diseased region identified by the coronary angiography image. The method has the beneficial effects that whether the coronary stent needs to be implanted or not is determined according to the coronary angiography image of a patient, the limitation of judgment purely depending on doctor experience is abandoned, and the implantation decision is more scientific and objective; meanwhile, an implantation planning scheme can be determined according to the focus characteristics of the target diseased region, the improper implantation situation caused by insufficient experience or subjective judgment deviation is avoided, such as the problems of stent falling, stent breakage, blood vessel restenosis and the like, and the success rate and the safety of the coronary stent implantation operation are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of implantation planning, and in particular to a coronary stent implantation planning method and system based on artificial intelligence. Background Art

[0002] Coronary angiography stent placement refers to a stent implantation procedure performed to restore blood flow when blood vessel stenosis or blockage is found during coronary angiography.

[0003] Stent implantation is essentially a transradial artery (TRA) puncture. Using a special catheter and guidewire, a balloon is delivered to the narrowed or occluded coronary artery under X-ray fluoroscopy. The balloon then dilates the narrowed or occluded lesion. The stent then expands to further expand the narrowed vessel wall, promoting blood flow. If the radial artery is blocked or inaccessible, the brachial or femoral artery can be used as an alternative.

[0004] Currently, conventional practice is to combine imaging (DSA) and functional instrumentation to assess the fractional flow reserve (FFR) to measure the degree of lesion stenosis and hemodynamics, thereby determining the need for stent implantation. FFR assessment requires the use of vasodilators to induce maximal hyperemia in the myocardial microcirculation, a complex and time-consuming procedure. Furthermore, some patients may experience allergic reactions to drugs such as adenosine or ATP, which can lead to adverse reactions such as dyspnea and hypotension, and can even be life-threatening.

[0005] Moreover, the planning of coronary stent implantation, such as the determination of the implantation path, the selection of the guide catheter, guidewire and coronary stent, also relies on the doctor's experience. However, during the stent implantation process, inappropriate stent implantation may lead to stent detachment, breakage and restenosis of the blood vessel. Summary of the Invention

[0006] In order to solve the above technical problems, the present invention provides a coronary stent implantation planning method based on artificial intelligence; on the other hand, it also provides a coronary stent implantation planning system based on artificial intelligence.

[0007] The technical problem solved by the present invention can be achieved by adopting the following technical solutions:

[0008] In one aspect, a coronary stent implantation planning method based on artificial intelligence is provided, comprising:

[0009] Step S1, determining whether a coronary stent needs to be implanted based on the patient's coronary angiography image;

[0010] Step S2, in response to an instruction indicating the need to implant a coronary stent, determining an implantation plan for implanting the coronary stent in the lesion site based on the lesion characteristics of the target lesion site identified by the coronary angiography image.

[0011] Preferably, the step S1 includes:

[0012] Step S11, inputting the coronary angiography image into a pre-trained lesion recognition model, and the lesion recognition model recognizes the coronary angiography image to obtain a recognition result;

[0013] Step S12, when the recognition result includes at least one lesion site and the lesion feature of the lesion site exceeds a preset threshold, determining the lesion site whose lesion feature exceeds the preset threshold as a target lesion site;

[0014] Step S13: outputting an instruction indicating the need to implant a coronary stent according to the target lesion site.

[0015] Preferably, before step S11, the following steps are further included:

[0016] Step S101, collecting coronary angiography image data and annotating the coronary angiography image data, wherein the annotation includes at least a first annotation and a second annotation, wherein the first annotation is used to indicate the lesion location and corresponding lesion characteristics, and the second annotation is used to indicate whether a coronary stent needs to be implanted;

[0017] Step S102: performing model training, verification and testing based on the annotated coronary angiography image data to obtain the trained lesion recognition model.

[0018] Preferably, step S2 includes:

[0019] Step S21, generating a candidate implantation path based on the lesion characteristics of the target lesion and the candidate access sites;

[0020] Step S22, providing a plurality of device candidates, and generating a plurality of candidate implantation planning schemes based on the candidate implantation paths and the device candidates matching the candidate implantation paths;

[0021] Step S23: In response to an external selection instruction, one of the candidate implantation planning schemes is determined as the implantation planning scheme for implanting the coronary stent into the lesion site.

[0022] Preferably, the device candidates include at least a guide catheter candidate, a guidewire candidate, and a coronary stent candidate;

[0023] The step S22 includes:

[0024] Step S2201, determining a guide catheter candidate and a guide wire candidate that match the candidate implantation path according to the degree of curvature of the candidate implantation path;

[0025] Step S2202, determining a coronary stent candidate based on the target lesion site and corresponding lesion characteristics;

[0026] Step S2203: Generate the candidate implantation planning scheme based on the candidate implantation path and the determined guide catheter candidate, guidewire candidate and coronary stent candidate.

[0027] Preferably, after step S22 and before step S23, the following steps are further included:

[0028] Step S221, for each candidate implantation plan, performing implantation simulation according to the candidate device and the candidate implantation path in the candidate implantation plan;

[0029] Step S222: According to the contact condition between the candidate device and the blood vessel wall during the implantation simulation, the candidate implantation path and the candidate device are modified to obtain a modified implantation planning scheme.

[0030] Preferably, each of the candidate implantation paths comprises a plurality of consecutive sites;

[0031] During the implantation simulation process of step S221 , each site in the candidate implantation path is traversed, and the device candidate is implanted along the tangent direction of each site.

[0032] Preferably, when determining the implantation planning scheme in step S2, the implantation path is displayed simultaneously.

[0033] On the other hand, an artificial intelligence-based coronary stent implantation planning system is also provided, which is used to implement the artificial intelligence-based coronary stent implantation planning method as described above, comprising:

[0034] A stent implantation determination module is used to determine whether a coronary stent needs to be implanted based on the patient's coronary angiography images;

[0035] An implantation planning module is connected to the stent implantation determination module and is used to respond to an instruction indicating the need to implant a coronary stent, and determine an implantation planning scheme for implanting the coronary stent in the lesion site based on the lesion characteristics of the target lesion site identified by the coronary angiography image.

[0036] Preferably, the implant planning module includes:

[0037] An implantation path generation unit, configured to generate a candidate implantation path based on the lesion characteristics of the target lesion and the candidate access site;

[0038] a candidate solution planning unit, connected to the implantation path generation unit, configured to provide a plurality of device candidates and generate a plurality of candidate implantation planning solutions based on the candidate implantation paths and device candidates matching the candidate implantation paths;

[0039] A scheme determination unit is connected to the candidate scheme determination unit and is used to determine one of the candidate implantation planning schemes as the implantation planning scheme for implanting the coronary stent into the lesion site in response to an external selection instruction.

[0040] The advantages or beneficial effects of the technical solution of the present invention are:

[0041] The present invention determines whether a coronary stent needs to be implanted based on the patient's coronary angiography images, eliminating the limitation of relying solely on the doctor's experience to make judgments, making implantation decisions more scientific and objective. At the same time, after receiving an instruction indicating the need for coronary stent implantation, the present invention can determine the implantation plan for the coronary stent at the lesion site based on the lesion characteristics of the target lesion site identified by the coronary angiography images, avoiding inappropriate implantation caused by lack of experience or subjective judgment deviation, such as stent detachment, breakage, and restenosis of the blood vessel, thereby improving the success rate and safety of coronary stent implantation surgery. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a flow chart of a coronary stent implantation planning method based on artificial intelligence in a preferred embodiment of the present invention;

[0043] Figure 2 This is a flow chart of step S1 in a preferred embodiment of the present invention;

[0044] Figure 3 A schematic diagram of the process of training a lesion recognition model in a preferred embodiment of the present invention;

[0045] Figure 4 This is a flow chart of step S2 in a preferred embodiment of the present invention;

[0046] Figure 5 This is a flow chart of step S22 in a preferred embodiment of the present invention;

[0047] Figure 6 This is a flow chart after step S22 and before step S23 in a preferred embodiment of the present invention;

[0048] Figure 7 This is a structural block diagram of a coronary stent implantation planning system based on artificial intelligence in a preferred embodiment of the present invention;

[0049] Figure 8This is a structural block diagram of the implant planning module in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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 making any creative efforts shall fall within the scope of protection of the present invention.

[0051] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0052] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0053] In a preferred embodiment of the present invention, based on the above problems existing in the prior art, a coronary stent implantation planning method based on artificial intelligence is provided. Figure 1 As shown, including:

[0054] Step S1, determining whether a coronary stent needs to be implanted based on the patient's coronary angiography image;

[0055] Step S2, in response to an instruction indicating the need to implant a coronary stent, determining an implantation plan for the coronary stent at the lesion site based on the lesion characteristics of the target lesion site identified by the coronary angiography image.

[0056] Specifically, coronary angiography involves inserting a catheter through a peripheral artery (usually the radial or femoral artery), pushing it along the artery to the opening of the coronary artery, and then injecting a contrast agent through the catheter. The contrast agent flows within the coronary arteries and creates an image. Coronary angiography provides coronary angiographic images, which consist of multiple consecutive images of the coronary arteries. These images can show the course and branches of the coronary arteries, as well as the presence of lesions such as stenosis, blockage, and dilation.

[0057] In this embodiment, coronary angiography images of the patient are obtained and analyzed for lesion locations (e.g., locations of stenosis or blockage) and corresponding lesion characteristics. These characteristics include, but are not limited to, lesion length, morphology (e.g., diffuse or localized stenosis), degree of stenosis, and lesion location. Based on these lesion characteristics, an appropriate treatment plan is selected.

[0058] The degree of stenosis is generally assessed based on the percentage of reduction in vessel diameter. For example, when the vessel diameter decreases by more than a certain percentage (e.g., 70%), a coronary stent is needed to improve blood flow. For stenosis less than 70%, medication is generally used. Coronary stent implantation is suitable for short-segment lesions, but long, multiple lesions are less suitable for stent therapy.

[0059] That is, when there is a lesion with a stenosis degree exceeding 70%, an instruction to implant a coronary stent is output, and the instruction includes a target lesion. The target lesion is the lesion with a stenosis degree exceeding 70%.

[0060] In other cases, no instructions for coronary stent implantation will be output.

[0061] When an instruction to implant a coronary stent is received, an implantation plan is formulated based on the lesion characteristics of the target lesion site, which at least includes the implantation path and the instrument used for implantation.

[0062] The implantation pathway is the path taken from the access site to the target lesion. The access site is usually the radial artery or femoral artery.

[0063] In addition, the path may also include the implantation angle, which refers to the inclination angle of the device relative to the axis of the blood vessel when it enters the human blood vessel during the implantation process.

[0064] The implantation path consists of multiple consecutive sites. The implantation angle is the inclination angle of the tangent direction of each site relative to the vascular axis.

[0065] The devices used for implantation include at least a guide catheter, a guide wire, and a coronary stent, and generally may also include a balloon.

[0066] The embodiments of the present invention eliminate the limitations of relying solely on the doctor's experience to make judgments, avoid inappropriate implantation situations caused by lack of experience or subjective judgment bias, such as stent detachment, breakage, and restenosis of blood vessels, making implantation decisions more scientific and objective, and improving the success rate and safety of coronary stent implantation surgery.

[0067] As a preferred embodiment, wherein Figure 2 As shown, step S1 includes:

[0068] Step S11, inputting the coronary angiography image into a pre-trained lesion recognition model, and the lesion recognition model recognizes the coronary angiography image to obtain a recognition result;

[0069] Step S12, when the recognition result includes at least one lesion site and the lesion feature of the lesion site exceeds a preset threshold, determining the lesion site whose lesion feature exceeds the preset threshold as a target lesion site;

[0070] Step S13: outputting an instruction indicating the need to implant a coronary stent according to the target lesion site.

[0071] Specifically, in this embodiment, the lesion recognition model is trained based on a large amount of coronary angiography data. Using deep learning technology, it learns the characteristics of lesions with different degrees of stenosis. The lesion recognition model can identify lesions and extract lesion characteristics.

[0072] The coronary angiography image to be processed is fed into the trained lesion recognition model. Upon receiving the coronary angiography image, the lesion recognition model analyzes and processes the image pixel by pixel, generating a recognition result. This result indicates whether the image contains a lesion and, if so, may further include the lesion characteristics of each lesion.

[0073] This lesion feature is usually used to characterize the degree of stenosis, that is, the ratio of the lesion diameter to the blood vessel diameter.

[0074] If the identification result includes at least one lesion, then for each lesion, it is determined whether the lesion characteristics exceed a preset threshold. The preset threshold can be determined based on clinical experience and research, and is generally set at 70%.

[0075] If the lesion characteristics of a certain lesion site exceed a preset threshold, the lesion site is set as a target lesion site and stent implantation is performed on the target lesion site; otherwise, the lesion site does not require stent implantation.

[0076] If the lesion characteristics of a single lesion exceed a preset threshold, stent implantation is indicated. A command indicating the need for coronary stent implantation is output, which includes all target lesions for stent implantation. This command can include one or more target lesions.

[0077] If the lesion characteristics of all lesions do not exceed the preset threshold, stent implantation is not required.

[0078] As a preferred embodiment, wherein Figure 3 As shown, before step S11, the following steps are also included:

[0079] Step S101: collecting coronary angiography image data and annotating the coronary angiography image data, wherein the annotation includes at least a first annotation and a second annotation, wherein the first annotation is used to indicate the lesion location and corresponding lesion characteristics, and the second annotation is used to indicate whether a coronary stent needs to be implanted;

[0080] Step S102 : performing model training, verification, and testing based on the annotated coronary angiography image data to obtain a trained lesion recognition model.

[0081] Specifically, when training the lesion recognition model, we first collected a large amount of coronary angiography data from clinical practice. This data covers coronary angiography images of different patients and different degrees of coronary artery stenosis or obstruction to ensure the model's generalization ability.

[0082] Then, the location of the lesion and the characteristics of the lesion corresponding to the lesion location are marked in the coronary angiography image data, such as the length along the extending direction of the blood vessel and the width along the cross section of the blood vessel.

[0083] At the same time, for each lesion site, if implantation is required, the second annotation is marked as "yes"; if implantation is not required, the second annotation is marked as "no".

[0084] Labeled coronary angiography data is fed into a deep learning model architecture, such as a convolutional neural network (CNN). During training, the model gradually adjusts its weights and parameters based on the labeling information to learn how to identify lesion locations and lesion-related features, and also learns how to determine the need for coronary stent implantation based on lesion characteristics. Through extensive training data and repeated iterative optimization, the model gradually improves the accuracy and reliability of lesion identification.

[0085] After training is completed, the trained model is verified using coronary angiography image data that has not participated in model training. The performance of the model is evaluated based on indicators such as accuracy, recall, and specificity.

[0086] After the verification is completed, the verified model is tested using coronary angiography image data that has not participated in model training and verification to evaluate the performance of the model in actual application scenarios.

[0087] After the test is passed, the final lesion recognition model can be obtained.

[0088] During the application of the model, new coronary angiography imaging data can be continuously collected to achieve continuous updating of the model.

[0089] As a preferred embodiment, wherein Figure 4 As shown, step S2 includes:

[0090] Step S21, generating candidate implantation paths based on the lesion characteristics of the target lesion and the candidate access sites;

[0091] During implant path generation, the candidate entry point is used as the implant starting point, and the target lesion location is used as the implant endpoint. Based on the curvature of the vessel, a path is planned that initially follows the main vessel and then gradually turns toward the branch vessel where the target lesion is located.

[0092] Candidate access sites include the radial artery and the femoral artery. The access site for stent implantation is generally the same as that used during coronary angiography.

[0093] Step S22, providing multiple device candidates, and generating multiple candidate implantation planning schemes based on the candidate implantation paths and the device candidates that match the candidate implantation paths;

[0094] Among them, device candidates include at least guide catheter candidates, guidewire candidates and coronary stent candidates;

[0095] like Figure 5 As shown, step S22 includes:

[0096] Step S2201, determining a guide catheter candidate and a guide wire candidate that match the candidate implantation path according to the degree of curvature of the candidate implantation path;

[0097] The guide catheter candidates include guide catheters of different models and different inner diameters. For example, the inner diameter of the guide catheter includes but is not limited to 5 French (5F), 6F, or 7F.

[0098] The system can provide most types of guide catheters on the market, and some guide catheters can be selected based on the actual purchasing situation of our hospital.

[0099] During the matching process of the guide catheter candidates, the candidate implantation path is matched with the selected guide catheter according to the degree of curvature corresponding to the candidate implantation path and the diameter of the coronary artery, and the matching guide catheter candidate is output.

[0100] The system can also provide most guidewire types on the market, and some guidewires can be selected based on the actual purchasing situation of our hospital.

[0101] During the guidewire candidate matching process, the candidate implantation path is matched with the selected guidewire according to the degree of curvature of the coronary artery, and the matching guidewire candidate is output.

[0102] In addition to guide catheter and guidewire candidates, device candidates may also include balloon candidates. The balloon length is selected based on the target lesion length, and the balloon diameter is selected based on the remaining vessel diameter at the lesion site. The balloon diameter should be slightly larger than the vessel diameter at the lesion site to ensure adequate expansion of the lesion, but not too large to prevent complications such as vessel rupture. Typically, a balloon 0.5 mm to 1 mm larger than the remaining vessel diameter is selected.

[0103] Step S2202, determining a coronary stent candidate based on the target lesion site and corresponding lesion characteristics;

[0104] The diameter of the candidate coronary stent should match the diameter of the blood vessel at the lesion site. Usually, a stent that is 0.5mm-1mm larger than the blood vessel diameter is selected to ensure that the stent can adhere well to the wall of the blood vessel and will not expand excessively and cause blood vessel damage.

[0105] The length of the stent should be slightly longer than the lesion to ensure that the lesion can be completely covered, but it should not be too long to avoid excessive expansion of normal blood vessels at both ends of the stent.

[0106] The number of guide catheter candidates, guidewire candidates, and coronary stent candidates determined after matching may be one or more. Due to the actual equipment procurement situation of each hospital, the number of device candidates obtained after matching will not be too large.

[0107] Step S2203: Generate a candidate implantation planning scheme based on the candidate implantation paths and the determined guide catheter candidates, guidewire candidates, and coronary stent candidates.

[0108] Specifically, candidate implantation paths are combined with determined guide catheter candidates, guidewire candidates, and coronary stent candidates in different ways to form multiple candidate implantation planning schemes.

[0109] For example, the determined guide catheter candidates include guide catheter A; the determined guidewire candidates include guidewire B; and the determined coronary stent candidates include coronary stent C and coronary stent D. The combined candidate implantation plans are as follows:

[0110] Option 1: Use guide catheter A, guide wire B and coronary stent C.

[0111] Option 2: Use guide catheter A, guide wire B and coronary stent D.

[0112] Furthermore, after a plurality of candidate implantation planning schemes are generated, the generated plurality of candidate implantation planning schemes are displayed on an operation interface for selection.

[0113] Furthermore, a personalized label may be provided for each candidate implantation planning scheme, and the display order of the multiple candidate implantation planning schemes may be adjusted according to the personalized label.

[0114] For example, personalized labels include but are not limited to economical and performance types. In actual medical operations, appropriate candidate implant planning plans can be selected more conveniently based on specific circumstances and needs.

[0115] Step S23: In response to an external selection instruction, one of the candidate implantation plans is determined as the implantation plan for the coronary stent implanted in the lesion site.

[0116] Specifically, in this embodiment, the external selection instruction can come from the doctor's operation interface, and one of the displayed candidate implant plans can be selected by clicking, dragging, etc. The selection operation can also be achieved through voice instructions, gesture recognition, and other technologies.

[0117] A selection instruction is generated based on the selected candidate implantation planning scheme.

[0118] According to the selection instruction, the selected candidate implantation plan is determined as the final implantation plan for the coronary stent implantation in the lesion site and displayed on the operation interface to assist in completing the instrument operation in the coronary stent implantation surgery.

[0119] As a preferred embodiment, wherein Figure 6 As shown, after step S22 and before step S23, the following steps are further included:

[0120] Step S221, for each candidate implantation plan, performing implantation simulation according to the device candidate and the candidate implantation path in the candidate implantation plan;

[0121] Each candidate implantation path includes multiple consecutive sites;

[0122] During the implantation simulation, each site in the candidate implantation path is traversed, and the device candidate is implanted along the tangent direction of each site.

[0123] Step S222: According to the contact between the device candidate and the blood vessel wall during the implantation simulation, the candidate implantation path and the device candidate are modified to obtain a modified implantation plan.

[0124] Specifically, the implantation of the coronary stent needs to be simulated before the stent is implanted into the blood vessel, so as to ensure the accuracy of the coronary stent implantation process.

[0125] In this embodiment, the implantation simulation can be performed in a three-dimensional coronary artery model. During the simulation, the determined guide catheter candidate, guidewire candidate, and coronary stent candidate are implanted along the tangent direction of each site in the candidate implantation path.

[0126] According to the contact between the front ends of the guide catheter candidate, guidewire candidate and coronary stent candidate and the blood vessel wall during the implantation simulation, the candidate implantation paths, guide catheter candidate, guidewire candidate and coronary stent candidate are modified.

[0127] More specifically, the correction in step S222 includes:

[0128] Adjust the position of the site; and / or

[0129] Adjusting the curvature of the site; and / or

[0130] Change of guide catheter candidate; and / or

[0131] Candidate for guidewire replacement; and / or

[0132] Candidates for coronary stent replacement.

[0133] Specifically, if the tip of the device candidate is excessively in contact with the vessel wall at a certain location, the position or curvature of that location is adjusted. For example, the coordinates of that location can be adjusted to move it a certain distance along the vessel's radial direction to improve the fit between the device and the vessel wall. Another example is adjusting the deflection direction of that location to offset it by a certain angle to improve the fit between the device and the vessel wall.

[0134] If the device candidate has excessive contact with the vessel wall within the position or curvature adjustment range of the site, the device candidate is replaced.

[0135] As a preferred embodiment, in step S2, when the implantation planning scheme is determined, the implantation path is displayed simultaneously.

[0136] During the operation, the implantation path in the final implantation plan will be displayed on the operation interface to assist in completing the coronary stent implantation surgery.

[0137] The steps of the various methods above are divided only for the purpose of clear description. During implementation, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this patent. Adding insignificant modifications or introducing insignificant designs to the algorithm or process without changing the core design of the algorithm and process are all within the scope of protection of this patent.

[0138] On the other hand, a coronary stent implantation planning system based on artificial intelligence is also provided for implementing the above-mentioned coronary stent implantation planning method based on artificial intelligence, such as Figure 7 Shown, including:

[0139] Stent implantation determination module 1, used to determine whether a coronary stent needs to be implanted based on the patient's coronary angiography images;

[0140] The implantation planning module 2 is connected to the stent implantation determination module 1 and is used to respond to an instruction indicating the need to implant a coronary stent, and determine an implantation planning scheme for the coronary stent implantation in the lesion site based on the target lesion location and lesion characteristics identified by the coronary angiography image.

[0141] As a preferred embodiment, wherein Figure 8As shown, the implant planning module 2 includes:

[0142] An implantation path generation unit 21 is configured to generate a candidate implantation path based on the lesion characteristics of the target lesion and the candidate access sites;

[0143] The candidate solution planning unit 22 is connected to the implantation path generating unit 21 and is used to provide a plurality of device candidates and generate a plurality of candidate implantation planning solutions according to the candidate implantation paths and the device candidates matching the candidate implantation paths;

[0144] The plan determination unit 23 is connected to the candidate plan determination unit 22 and is used to determine one of the candidate implantation plan plans as the implantation plan plan for implanting the coronary stent into the lesion site in response to an external selection instruction.

[0145] More implementation details of the system and the method have been disclosed and will not be repeated in this embodiment.

[0146] For the convenience of description, the above system is described as being divided into various modules according to their functions. Of course, when implementing the present invention, the functions of each unit or module can be implemented in the same or multiple software and / or hardware.

[0147] The advantages or beneficial effects of adopting the above technical solution are: the present invention determines whether a coronary stent needs to be implanted based on the patient's coronary angiography images, thereby abandoning the limitations of relying solely on the doctor's experience to make judgments, making the implantation decision more scientific and objective; at the same time, after receiving an instruction indicating the need to implant a coronary stent, it can determine the implantation plan for the coronary stent at the lesion site based on the lesion characteristics of the target lesion site identified by the coronary angiography image, avoiding inappropriate implantation caused by lack of experience or subjective judgment deviation, such as stent detachment, breakage, and restenosis of the blood vessel, thereby improving the success rate and safety of coronary stent implantation surgery.

[0148] The above are only preferred embodiments of the present invention and do not limit the implementation mode and protection scope of the present invention. For those skilled in the art, it should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the contents of this specification and illustrations should be included in the protection scope of the present invention.

Claims

1. A coronary stent implantation planning method based on artificial intelligence, characterized in that: include: Step S1, determining whether a coronary stent needs to be implanted based on the patient's coronary angiography image; Step S2, in response to an instruction indicating the need to implant a coronary stent, determining an implantation plan for implanting the coronary stent in the lesion site based on the lesion characteristics of the target lesion site identified by the coronary angiography image.

2. The coronary stent implantation planning method according to claim 1, characterized in that: The step S1 comprises: Step S11, inputting the coronary angiography image into a pre-trained lesion recognition model, and the lesion recognition model recognizes the coronary angiography image to obtain a recognition result; Step S12, when the recognition result includes at least one lesion site and the lesion feature of the lesion site exceeds a preset threshold, determining the lesion site whose lesion feature exceeds the preset threshold as a target lesion site; Step S13: outputting an instruction indicating the need to implant a coronary stent according to the target lesion site.

3. The coronary stent implantation planning method according to claim 2, characterized in that: Before step S11, the following steps are also included: Step S101, collecting coronary angiography image data and annotating the coronary angiography image data, wherein the annotation includes at least a first annotation and a second annotation, wherein the first annotation is used to indicate the lesion location and corresponding lesion characteristics, and the second annotation is used to indicate whether a coronary stent needs to be implanted; Step S102: performing model training, verification and testing based on the annotated coronary angiography image data to obtain the trained lesion recognition model.

4. The coronary stent implantation planning method according to claim 1, characterized in that: The step S2 comprises: Step S21, generating a candidate implantation path based on the lesion characteristics of the target lesion and the candidate access sites; Step S22, providing a plurality of device candidates, and generating a plurality of candidate implantation planning schemes based on the candidate implantation paths and the device candidates matching the candidate implantation paths; Step S23: In response to an external selection instruction, one of the candidate implantation planning schemes is determined as the implantation planning scheme for implanting the coronary stent into the lesion site.

5. The coronary stent implantation planning method according to claim 4, characterized in that: The device candidates include at least a guide catheter candidate, a guidewire candidate and a coronary stent candidate; The step S22 includes: Step S2201, determining a guide catheter candidate and a guide wire candidate that match the candidate implantation path according to the degree of curvature of the candidate implantation path; Step S2202, determining a coronary stent candidate based on the target lesion site and corresponding lesion characteristics; Step S2203: Generate the candidate implantation planning scheme based on the candidate implantation path and the determined guide catheter candidate, guidewire candidate and coronary stent candidate.

6. The coronary stent implantation planning method according to claim 4, characterized in that: After step S22 and before step S23, the following steps are further included: Step S221, for each candidate implantation plan, performing implantation simulation according to the candidate device and the candidate implantation path in the candidate implantation plan; Step S222: According to the contact condition between the candidate device and the blood vessel wall during the implantation simulation, the candidate implantation path and the candidate device are modified to obtain a modified implantation planning scheme.

7. The coronary stent implantation planning method according to claim 6, characterized in that: Each of the candidate implantation paths comprises a plurality of continuous sites; During the implantation simulation process of step S221 , each site in the candidate implantation path is traversed, and the device candidate is implanted along the tangent direction of each site.

8. The coronary stent implantation planning method according to claim 1, characterized in that: In step S2, when the implantation planning scheme is determined, the implantation path is displayed simultaneously.

9. An artificial intelligence-based coronary stent implantation planning system, characterized in that: A method for implementing the artificial intelligence-based coronary stent implantation planning method according to any one of claims 1 to 8, comprising: A stent implantation determination module is used to determine whether a coronary stent needs to be implanted based on the patient's coronary angiography images; An implantation planning module is connected to the stent implantation determination module and is used to respond to an instruction indicating the need to implant a coronary stent, and determine an implantation planning scheme for implanting the coronary stent in the lesion site based on the lesion characteristics of the target lesion site identified by the coronary angiography image.

10. The coronary stent implantation planning system according to claim 9, characterized in that: The implantation planning module includes: An implantation path generation unit, configured to generate a candidate implantation path based on the lesion characteristics of the target lesion and the candidate access site; a candidate solution planning unit, connected to the implantation path generation unit, configured to provide a plurality of device candidates and generate a plurality of candidate implantation planning solutions based on the candidate implantation paths and device candidates matching the candidate implantation paths; A scheme determination unit is connected to the candidate scheme determination unit and is used to determine one of the candidate implantation planning schemes as the implantation planning scheme for implanting the coronary stent into the lesion site in response to an external selection instruction.

Images