Brain puncture path planning method and apparatus

Abstract

The present application relates to the field of computer image processing technology, and in particular, to a brain puncture path planning method and apparatus. An MRI-compatible robot, by acquiring pre-operative and intra-operative brain MRI images and registering the pre-operative and intra-operative images, can determine a tissue drift in the target tissue of brain puncture and registration errors caused by registering different brain MRI images. Simultaneously, by combining the three-dimensional reconstruction error that would be generated by three-dimensional reconstruction of the brain MRI images and the positioning error of the puncture needle, the MRI-compatible robot performs robustness evaluation on a plurality of brain puncture paths obtained from pre-operative planning. The brain puncture path with higher robustness is selected as the target brain puncture path. At this time, the target brain puncture path has higher safety and lower risk, is more reliable even in the presence of the aforementioned errors, and can reduce the possibility of damage to the patient's intracranial nerves, blood vessels, and functional areas due to errors.

Description

A method and device for brain puncture path planning Technical Field

[0001] This application relates to the field of computer image processing technology, and in particular to a method and apparatus for brain puncture path planning. Background Technology

[0002] The number of patients with brain functional disorders (such as Parkinson's disease, stroke, and Alzheimer's disease) and brain tumors has also increased significantly. The pathological mechanisms of brain functional disorders and brain tumors are complex and highly dangerous, and they also face high clinical and social costs.

[0003] Brain biopsy establishes a pathway from outside the patient to the intracranial target site, making it an important minimally invasive surgical procedure for treating brain lesions. Current research focuses on designing MRI-compatible robots to perform these procedures, but research on MRI-compatible robots in brain biopsy remains incomplete.

[0004] The basis for puncture path planning in MRI-compatible robots is usually the patient's preoperative MRI images, that is, puncture path planning is based on image navigation. However, the above-mentioned puncture path planning method has the problem of inaccurate path planning. Summary of the Invention

[0005] In view of this, this application provides a brain puncture path planning method and device, thereby enabling the application of MRI-compatible robots in brain puncture surgery, while improving the safety and reliability of brain puncture surgery.

[0006] This application provides a brain puncture path planning method, applied to MRI-compatible robots, implemented in the following manner:

[0007] The first brain MRI image acquired before the brain puncture surgery was obtained, and the first brain MRI image was reconstructed in three dimensions to obtain the first target region; the first target region includes: the target brain puncture tissue, the target brain functional area, and the areas where blood vessels and nerves are located in the first brain MRI image;

[0008] Based on the first target region, multiple brain puncture pathways are obtained through path planning.

[0009] A second brain MRI image was acquired during a brain puncture surgery, and the second brain MRI image was registered with the first brain MRI image to obtain the registration error.

[0010] Three-dimensional reconstruction was performed on the second brain MRI image to obtain the second target region, and the deformation drift of the second target region relative to the first target region was calculated; the second target region includes: the areas where the target brain puncture tissue, target brain functional areas, blood vessels and nerves are located in the second brain MRI image;

[0011] Based on the puncture needle positioning error, three-dimensional reconstruction error, registration error and deformation drift, the robustness of multiple brain puncture paths is evaluated, and the robustness target values ​​of multiple brain puncture paths are obtained.

[0012] The brain puncture path with a robust target value less than a first threshold among multiple brain puncture paths is selected as the target brain puncture path.

[0013] Optionally, based on the puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift, robustness assessments are performed on multiple brain puncture paths to obtain robustness target values ​​for multiple brain puncture paths, including:

[0014] The range of the first interval is determined based on the 3D reconstruction error, registration error, and deformation drift.

[0015] For each path point in multiple brain puncture pathways, multiple perturbation solutions are generated within the first interval.

[0016] Based on the target brain functional area and the location of the blood vessels and nerves closest to the perturbation solution, the cumulative danger distance of the perturbation solution within the puncture needle positioning error is calculated.

[0017] Based on the cumulative danger distance of each perturbation solution in multiple perturbation solutions, the robustness of multiple brain puncture paths is evaluated, and the robustness target value of multiple brain puncture paths is obtained.

[0018] Optionally, based on the first target region, path planning is performed to obtain multiple brain puncture paths, including:

[0019] Obtain the coordinate range of the needle insertion point and the coordinate range of the target point; the coordinate range of the needle insertion point and the coordinate range of the target point are determined based on the first target region;

[0020] For each of the multiple initial paths, multiple perturbation solutions are generated for each path point in the initial path within the second interval. The multiple initial paths are obtained based on the coordinate range of the needle entry point and the coordinate range of the target point. The second interval is determined based on a safety threshold.

[0021] Based on the target brain functional area and the location of the blood vessels and nerves closest to the perturbation solution, the cumulative danger distance of the perturbation solution within the puncture needle positioning error is calculated.

[0022] Based on the cumulative danger distance of each perturbation solution in multiple perturbation solutions, the robustness of multiple initial paths is evaluated to obtain the robustness target value of multiple initial paths;

[0023] The first preset number of initial paths whose robust target value is less than the second threshold is selected as the brain puncture path; the second preset number is greater than the first preset number.

[0024] Optionally, based on the cumulative danger distance of each of the multiple perturbation solutions, a robustness assessment is performed on the multiple initial paths to obtain the robustness target value of the multiple initial paths, which is achieved through the following formula:

[0025] Among them, f eff (x bs ,x be x is the robustness target value for the initial path. bs x is the needle entry point. be For the target point, d(x) bs ,x be q(y) represents the path length from the needle insertion point to the target point, K represents the number of path points in the initial path, N represents the number of perturbation solutions corresponding to each path point, and q(y) represents the number of perturbation solutions corresponding to each path point. k,i ) represents the path point y k The cumulative danger distance of the i-th perturbation solution.

[0026] Optionally, after selecting a first preset number of brain puncture paths with a robust target value less than a first threshold as the target brain puncture path, the method further includes:

[0027] Based on the length and cumulative danger distance of the target brain puncture path, a topographic map of the target brain puncture path is fitted and visualized.

[0028] Optionally, after fitting and visualizing the path topography map corresponding to the target brain puncture path, the process also includes:

[0029] In response to the replanning instruction, the first and second brain MRI images are fused to obtain the third brain MRI image;

[0030] Three-dimensional reconstruction was performed on the third brain MRI image to obtain the third target region. Based on the third target region, the path was replanned to obtain multiple replanned brain puncture paths. The replanned brain puncture paths were then used to replace the original brain puncture paths. The third target region includes: the target brain puncture tissue, target brain functional areas, and the areas where blood vessels and nerves are located in the third brain MRI image.

[0031] Based on the third and first brain MRI images, the registration error and deformation drift were adjusted.

[0032] Based on the adjusted registration error and the adjusted deformation drift, the following steps are repeated: robustness assessment is performed on multiple brain puncture paths to obtain robust target values ​​for multiple brain puncture paths, and a first preset number of brain puncture paths whose robust target values ​​are less than a first threshold are selected as target brain puncture paths.

[0033] Optionally, after fitting and visualizing the path topography map corresponding to the target brain puncture path, the process also includes:

[0034] In response to the selection command, the execution path is determined from the target brain puncture path;

[0035] Obtain the real-time brain MRI image corresponding to the execution path, register and fuse the real-time brain MRI image with the first brain MRI image to obtain the fourth brain MRI image and output it.

[0036] Three-dimensional modeling was performed on the fourth brain MRI image, and the real-time risk of the execution path was assessed.

[0037] If the real-time risk of the execution path is higher than the risk threshold, a risk warning signal will be output and the path will be replanned.

[0038] Optionally, based on the puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift, robustness assessments are performed on multiple brain puncture paths to obtain robustness target values ​​for multiple brain puncture paths, including:

[0039] A first brain puncture pathways were selected as candidate pathways from multiple brain puncture pathways.

[0040] Based on the puncture needle positioning error, 3D reconstruction error, registration error and deformation drift, the robustness of the candidate path is evaluated, and the robustness target value of the candidate path is obtained.

[0041] Record B candidate paths from A candidate paths whose robust target value is less than the third threshold to obtain the robust path; where A and B are both positive integers, and B is less than or equal to A;

[0042] A second brain puncture path is selected from multiple brain puncture paths to replace the candidate path, and the robustness evaluation of the candidate path is re-executed based on puncture needle positioning error, three-dimensional reconstruction error, registration error and deformation drift to obtain the robustness target value of the candidate path.

[0043] If the robust target value of the second brain puncture path is less than the robust target value of the robust path, then the robust path is replaced by the second brain puncture path.

[0044] This application also provides a brain puncture path planning device for use in MRI-compatible robots, the device comprising:

[0045] The three-dimensional reconstruction module is used to acquire the first brain MRI image taken before the brain puncture surgery, and to perform three-dimensional reconstruction on the first brain MRI image to obtain the first target region; the first target region includes: the target brain puncture tissue, the target brain functional area, and the areas where blood vessels and nerves are located in the first brain MRI image;

[0046] The path planning module is used to plan the path based on the first target region to obtain multiple brain puncture paths;

[0047] The registration module is used to acquire the second brain MRI image acquired during brain puncture surgery, and to register the second brain MRI image with the first brain MRI image to obtain the registration error;

[0048] The three-dimensional reconstruction module is also used to perform three-dimensional reconstruction of the second brain MRI image to obtain the second target region and calculate the deformation drift of the second target region relative to the first target region; the second target region includes: the areas where the target brain puncture tissue, target brain functional areas, blood vessels and nerves are located in the second brain MRI image;

[0049] The evaluation module is used to evaluate the robustness of multiple brain puncture paths based on puncture needle positioning error, 3D reconstruction error, registration error and deformation drift, and obtain the robustness target value of multiple brain puncture paths.

[0050] The evaluation module is also used to select a first preset number of brain puncture paths among multiple brain puncture paths whose robust target value is less than a first threshold as target brain puncture paths.

[0051] This application also provides an NMR-compatible robot, comprising: a processor coupled to a memory, the memory storing at least one computer program instruction, the at least one computer program instruction being loaded and executed by the processor to enable the NMR-compatible robot to implement the above-described method.

[0052] Therefore, the beneficial effects of this application are: by acquiring brain MRI images of the patient before and during the operation, and registering the preoperative and intraoperative images, the MRI-compatible robot can determine the tissue drift caused by the target brain puncture tissue and the registration error caused by registering different brain MRI images. At the same time, by combining the three-dimensional reconstruction error caused by the three-dimensional reconstruction of brain MRI images and the positioning error of the puncture needle, the robustness of multiple brain puncture paths planned before the operation can be evaluated, and the brain puncture path with higher robustness can be used as the target brain puncture path. At this time, the target brain puncture path has higher safety and lower risk, and is more reliable even in the presence of the above errors. It can reduce the possibility of damage to the nerves, blood vessels and functional areas in the patient's brain caused by errors. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings.

[0054] Figure 1 is a flowchart of the first embodiment of this application;

[0055] Figure 2 is a flowchart of the second embodiment of this application;

[0056] Figure 3 is a flowchart of the third embodiment of this application;

[0057] Figure 4 is a flowchart of steps S401 to S405 in the fourth embodiment of this application;

[0058] Figure 5 is a flowchart of steps S406 to S410 in the fourth embodiment of this application;

[0059] Figure 6 is a schematic diagram of a path planning method provided in an embodiment of this application;

[0060] Figure 7 is a schematic diagram of a visual display path provided in an embodiment of this application;

[0061] Figure 8 is a schematic diagram of a brain puncture path planning device according to this application;

[0062] Figure 9 is a schematic diagram of a nuclear magnetic resonance compatible robot according to this application. Detailed Implementation

[0063] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0064] Since the human brain is soft tissue, brain tissue drift may occur during surgery, making the pre-planned brain puncture path unusable. Therefore, this application adopts a method that combines the patient's pre-operative and intra-operative brain MRI images and takes into account the possible registration errors between the pre-operative and intra-operative brain MRI images to plan the brain puncture path, thereby improving the safety and reliability of the brain puncture path.

[0065] Please refer to Figure 1. The specific steps of the first embodiment of this application are as follows:

[0066] S101: The MRI-compatible robot acquires the first brain MRI image taken before the brain puncture surgery, performs three-dimensional reconstruction on the first brain MRI image, and obtains the first target region.

[0067] It should be noted that MRI-compatible robots can be used in conjunction with MRI machines without interfering with brain MRI images.

[0068] The first brain MRI image is a clear and complete image of brain tissue acquired before a brain puncture surgery. It is mainly used for clinical diagnosis, treatment planning, surgical planning, and preoperative pathway planning, and is used as key preoperative information in the brain puncture surgery.

[0069] The primary target area includes: the target brain puncture tissue, target brain functional areas, and the regions containing blood vessels and nerves as seen in the first brain MRI image. The target brain puncture tissue is the tissue containing the target point of the brain puncture surgery. The target brain puncture tissue can be different tissues depending on the specific disease condition, such as brain tumors or thalamic nuclei.

[0070] It should be noted that when performing three-dimensional reconstruction of brain MRI images, in addition to using the target brain puncture tissue as the target area, target brain functional areas, blood vessels, and nerves can also be used as obstacles. The target brain functional area can be any one of the frontal lobe, temporal lobe, parietal lobe, occipital lobe, hypothalamus, cerebellum, and brainstem; the specific target brain functional area should be determined based on the location of the target brain puncture tissue.

[0071] S102: The MRI-compatible robot performs path planning based on the first target area to obtain multiple brain puncture paths.

[0072] The starting point of the brain puncture path is the outer surface of the brain, and the ending point is the target brain tissue inside the brain. Based on preoperative brain MRI images, the MRI-compatible robot can first perform path planning to obtain a preliminary brain puncture path.

[0073] Based on the three-dimensional reconstruction of the first brain MRI image, the specific location range of the target brain puncture tissue can be determined, and the location information of obstacles such as target brain functional areas, blood vessels and nerves can also be obtained. At this time, based on the above specific location range and location information, the needle insertion point and target point can be set in the three-dimensional reconstruction image.

[0074] In some implementations, the coordinate range of the needle insertion point and the target point can be set based on the first target area. At this time, the NMR-compatible robot can perform path planning based on the set results. Specifically, the process of "an MRI-compatible robot executing a path planning algorithm based on a first target region to obtain multiple brain puncture paths" can be achieved as follows: The MRI-compatible robot obtains the coordinate range of the needle insertion point and the coordinate range of the target point; the coordinate range of the needle insertion point and the coordinate range of the target point are determined based on the first target region; for each initial path, multiple perturbation solutions are generated for each path point within a second interval; the multiple initial paths are obtained based on the coordinate range of the needle insertion point and the coordinate range of the target point; the second interval is determined based on a safety threshold; the cumulative danger distance of the perturbation solution within the puncture needle positioning error is calculated based on the target brain functional area and the location of the blood vessels and nerves closest to the perturbation solution; the robustness of the multiple initial paths is evaluated based on the cumulative danger distance of each perturbation solution, and a robustness target value is obtained for the multiple initial paths; a second preset number of initial paths with robustness target values ​​less than the second threshold are selected as brain puncture paths. The second threshold can be set according to actual needs. Specifically, the robust target values ​​of multiple initial paths can be sorted in ascending order. In this case, the robust target value of the initial path with the sequence number of the second preset number plus 1 can be used as the second threshold.

[0075] It should be noted that, due to differences in the patient's individual information, such as the surgical procedure, the size and location of the tumor, and the brain functional areas near the tumor, the range of coordinate values ​​for the needle insertion point and the range of coordinate values ​​for the target point can be set to different values.

[0076] The initial path only needs to ensure that both the needle insertion point and the target point are within the set coordinate range. However, if the initial path passes through the target brain functional area, blood vessels, and nerves, it will cause unnecessary harm to the patient.

[0077] Currently, the objective function for path planning can be defined by the following formula: f(x) bs ,x be )=d(x bs ,x be )+H(x bs ,x be stg(x) bs ,x be ) = 0

[0078] Where, x bs x is the needle entry point. be For the target point, d(x) bs ,x beH(x) represents the path length from the needle insertion point to the target point. bs ,x be ) represents the cumulative danger distance between the pathway and blood vessels, nerves, and target brain functional areas, g(x) bs ,x be This is used to constrain the path away from the target brain functional area.

[0079] Since the target value obtained by the objective function is calculated by summing the length of the initial path and the cumulative danger distance of the initial path to blood vessels, nerves, and target brain functional areas, the magnitude of the target value can determine the level of risk of the initial path; a lower target value indicates lower risk, and vice versa. However, due to some errors, the MRI-compatible robot may still pass through the target brain functional areas, blood vessels, and nerves when executing the path. Therefore, by setting a safety threshold to perturb the initial path and combining it with the puncture needle positioning error, the robustness of the initial path can be evaluated, resulting in a more robust brain puncture path.

[0080] The safety threshold can be set according to actual needs. For example, the safety threshold can be set to 2-4mm. Taking a safety threshold of 2mm as an example, the second interval range can be [-2mm, +2mm].

[0081] Specifically, "based on the cumulative danger distance of each perturbation solution in multiple perturbation solutions, the robustness assessment of multiple initial paths is performed to obtain the robustness target value of multiple initial paths" can be achieved through the following formula:

[0082] Among them, f eff (x bs ,x be x is the robustness target value for the initial path. bs x is the needle entry point. be For the target point, d(x) bs ,x be q(y) represents the path length from the needle insertion point to the target point, K represents the number of path points in the initial path, N represents the number of perturbation solutions corresponding to each path point, and q(y) represents the number of perturbation solutions corresponding to each path point. k,i ) represents the path point y k The cumulative danger distance of the i-th perturbation solution.

[0083] By adding perturbations to the initial path, it is possible to determine whether the initial path is reliable in the presence of errors. Therefore, the robustness evaluation method using this formula is more reliable than the aforementioned evaluation method using an objective function.

[0084] The process of obtaining the brain puncture path described above can be achieved by performing robustness assessments on each initial path and selecting the few most robust initial paths as the brain puncture path based on their robustness target values; alternatively, it can be achieved through iteration, comparing the robustness target values ​​of the initial paths in multiple iterations and replacing the recorded less robust initial paths with the initial paths having smaller robustness target values; alternatively, the process can be achieved by evaluating the initial paths using the aforementioned objective function, selecting a few initial paths with smaller target values, performing robustness assessments on them, and selecting a few more paths with smaller robustness target values ​​as the brain puncture path; or alternatively, by using all initial paths as brain puncture paths. It should be noted that there are multiple ways to obtain the brain puncture path, and this application does not limit the specific method used.

[0085] S103: The MRI-compatible robot acquires the second brain MRI image obtained during brain puncture surgery, and registers the second brain MRI image with the first brain MRI image to obtain the registration error.

[0086] It should be noted that brain puncture surgery consists of two stages. The first stage is when the patient is in the operating room and on the operating table, but the puncture needle has not yet begun to puncture the brain. The second stage is when an MRI-compatible robot controls the puncture needle to perform brain puncture.

[0087] At this point, the second brain MRI image is a brain tissue image acquired in the first stage of the brain puncture surgery. The clarity of the second brain MRI image can be lower than that of the first brain MRI image, thus allowing brain MRI images to be obtained more quickly in the operating room.

[0088] Due to differences in acquisition conditions, image clarity, and deformation and drift of the target brain puncture tissue, the registration results may contain certain errors. By taking these registration errors into account, the originally planned brain puncture path can be evaluated.

[0089] S104: The MRI-compatible robot performs three-dimensional reconstruction of the second brain MRI image to obtain the second target region and calculates the deformation drift of the second target region relative to the first target region.

[0090] The second target area includes: the target brain puncture tissue, target brain functional areas, and the areas where blood vessels and nerves are located in the second brain MRI image.

[0091] Since brain tissue is soft tissue, the area where the target brain puncture tissue is located may differ from the first target area in the first brain MRI image due to deformation drift over time. In this case, by calculating the deformation drift of the second target area relative to the first target area, the originally planned brain puncture path can be evaluated.

[0092] It should be noted that the NMR-compatible robot can determine the registration error first and then the deformation drift, or it can determine the deformation drift first and then the registration error, or it can determine the registration error and deformation drift simultaneously. This application does not limit the order in which the registration error and deformation drift are obtained.

[0093] S105: The MRI-compatible robot performs robustness assessments on multiple brain puncture paths based on puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift, and obtains robust target values ​​for multiple brain puncture paths.

[0094] MRI-compatible robots can automatically acquire the puncture needle positioning error based on their own hardware parameters. By combining this error with the actual positioning error, the system can account for the errors that occur during the puncture process, thereby assessing the robustness of the brain puncture path. Furthermore, since certain errors also occur during 3D reconstruction, combining these errors can also more accurately evaluate the robustness of each brain puncture path.

[0095] The robustness of a brain puncture path can be assessed by adding perturbations and judging based on the target robustness value of the path. If the target robustness value of the brain puncture path is small after adding perturbations, then the path will not be significantly affected by the aforementioned errors, and the path can be considered robust.

[0096] Currently, the robustness assessment of brain puncture pathways is typically performed using the following formula:

[0097] Where x′ represents any point on the path, and y is the value of x′ in the range space. The solution after perturbation, as shown in the above formula, means that the solution around x′ is... The objective function of all solutions in the space is averaged and used as the robust objective value of solution x′. If f eff If the difference between (x′) and f(x′) is large, then the solution x′ is not robust, and vice versa.

[0098] However, since calculating the integral consumes a lot of computational resources, and considering the existence of various errors such as puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift, f can be obtained by sampling averaging. eff The estimated value of (x′).

[0099] In some implementations, "an MRI-compatible robot performs robustness assessments on multiple brain puncture paths based on puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift, obtaining robust target values ​​for multiple brain puncture paths" can be achieved as follows: The MRI-compatible robot determines a first interval range based on 3D reconstruction error, registration error, and deformation drift; for each path point in the multiple brain puncture paths, multiple perturbation solutions are generated within the first interval range; based on the target brain functional area and the location of the blood vessels and nerves closest to the perturbation solutions, the cumulative danger distance of the perturbation solutions within the puncture needle positioning error is calculated; based on the cumulative danger distance of each perturbation solution, the robustness assessment of the multiple brain puncture paths is performed, obtaining robust target values ​​for the multiple brain puncture paths.

[0100] It should be noted that 3D reconstruction error, registration error, and deformation drift can be collectively referred to as brain tissue position error. Positive and negative floating based on this brain tissue position error yields a range within which multiple perturbation solutions can be obtained for any point on the brain puncture path. Simultaneously, errors may also occur during the puncture process. In this case, for each perturbation solution, we can consider whether the superposition of the puncture needle positioning error will damage the target brain functional area, blood vessels, and nerves, thus enabling a robust assessment of the brain puncture path.

[0101] Specifically, "based on the cumulative danger distance of each perturbation solution in multiple perturbation solutions, the robustness assessment of multiple brain puncture paths is performed to obtain the robustness target value of multiple brain puncture paths" can be achieved through the following formula:

[0102] Among them, f eff (x bs ,x be x is the robust target value for the brain puncture path. bs x is the needle entry point. be For the target point, d(x) bs ,x be q(y) represents the path length from the needle insertion point to the target point, K represents the number of path points in the brain puncture path, N represents the number of perturbation solutions corresponding to each path point, and q(y) represents the path length from the needle insertion point to the target point. k,i ) represents the path point y k The cumulative danger distance of the i-th perturbation solution.

[0103] In this implementation, by performing the above method on each brain puncture path, robust evaluation of multiple brain puncture paths can be achieved.

[0104] In practical implementation, robustness assessments can be performed on multiple brain puncture paths simultaneously, and robustness target values ​​for multiple brain puncture paths can be obtained through iterative optimization. Therefore, in some other implementations, "the MRI-compatible robot performs robustness assessments on multiple brain puncture paths based on puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift to obtain robustness target values ​​for multiple brain puncture paths" can also be achieved as follows: Select A first brain puncture paths as candidate paths from multiple brain puncture paths; perform robustness assessments on the candidate paths based on puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift to obtain robustness target values ​​for the candidate paths; record the A candidate paths... B candidate paths with robust target values ​​less than the third threshold are selected as robust paths; where A and B are both positive integers, and B is less than or equal to A. A second brain puncture paths are selected from multiple brain puncture paths to replace the candidate paths, and the robustness of the candidate paths is re-evaluated based on puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift to obtain the robust target value of the candidate paths. If the robust target value of the second brain puncture path is less than the robust target value of the robust path, the second brain puncture path is used to replace the robust path.

[0105] It should be noted that candidate paths can be repeatedly replaced and subsequent steps executed until the iterative selection reaches the termination condition. Specifically, the termination condition can be any of the following: 1. The number of iterations exceeds a threshold; 2. The running time of the iterative program exceeds a time threshold. The termination condition can also be set according to actual needs and is not limited to the examples above.

[0106] The method for evaluating the robustness of candidate paths and obtaining their robustness target values ​​can be found in the previous implementation description, and will not be repeated here.

[0107] S106: The MRI-compatible robot selects a first preset number of brain puncture paths from among multiple brain puncture paths whose robust target value is less than a first threshold as the target brain puncture path.

[0108] Since the robustness target value represents the robustness of the brain puncture path, a brain puncture path with a smaller robustness target value can be used as the target brain puncture path to obtain a brain puncture path with lower risk. The first preset number can be set according to actual needs; this application does not limit the specific value of the first preset number, only requiring that the second preset number is greater than the first preset number.

[0109] The first threshold can be set according to actual needs. Specifically, it can be achieved by sorting the robust target values ​​of multiple brain puncture paths in ascending order, and then using the robust target value of the brain puncture path with the sequence number being the first preset number plus 1 as the first threshold.

[0110] In some implementations, the MRI-compatible robot selects a first preset number of brain puncture paths with robustness target values ​​less than a first threshold as target brain puncture paths. It can then fit and visualize a path topography map corresponding to the target brain puncture path based on its length and cumulative danger distance. Visualizing the target brain puncture path provides a more intuitive understanding of the characteristics of different paths, allowing surgeons to select the appropriate path for the brain puncture procedure. It should be noted that if no usable path is found among the target brain puncture paths, the MRI-compatible robot can perform path replanning to obtain a replanned brain puncture path.

[0111] In the first embodiment of this application, the MRI-compatible robot acquires brain MRI images of the patient before and during the operation, and registers the preoperative and intraoperative images. This allows it to determine tissue drift caused by the target brain puncture tissue and registration errors caused by registering different brain MRI images. Simultaneously, it combines the three-dimensional reconstruction errors caused by three-dimensional reconstruction of brain MRI images and the positioning errors of the puncture needle to conduct a robustness assessment of multiple brain puncture paths planned before the operation. The brain puncture path with higher robustness is selected as the target brain puncture path. At this time, the target brain puncture path has higher safety and lower risk, and is more reliable even in the presence of the above-mentioned errors. This can reduce the possibility of damage to the nerves, blood vessels and functional areas in the patient's brain caused by errors.

[0112] The following section introduces the replanning of the brain puncture route.

[0113] Please refer to Figure 2. The specific steps of the second embodiment of this application are as follows:

[0114] S201: In response to a replanning command, the MRI-compatible robot fuses the first and second brain MRI images to obtain a third brain MRI image.

[0115] After the MRI-compatible robot visualizes and outputs the topographic map of the target brain puncture path, the doctor can determine whether there is a suitable execution path for the patient to use during the operation based on the length of each target brain puncture path and the cumulative danger distance. If the doctor believes that there is no usable path, he can input a replanning command to the MRI-compatible robot.

[0116] By fusing the first and second brain MRI images, a clearer third brain MRI image containing information from the second brain MRI image can be obtained. Based on the third brain MRI image, the brain puncture path can be replanned.

[0117] S202: The MRI-compatible robot performs three-dimensional reconstruction on the third brain MRI image to obtain the third target region, and re-plans the path based on the third target region to obtain multiple replanned brain puncture paths, and uses the replanned brain puncture paths to replace the original brain puncture path.

[0118] The third target area includes: the target brain puncture tissue, target brain functional areas, and the areas where blood vessels and nerves are located in the third brain MRI image.

[0119] It should be noted that, based on the third target region, the coordinate range of the needle insertion point and the coordinate range of the target point can be redefined. The MRI-compatible robot can then re-plan the path based on the redefined coordinate range of the needle insertion point and the target point, resulting in multiple replanned brain puncture paths.

[0120] S203: MRI-compatible robots adjust registration errors and deformation drift based on third-brain MRI images and first-brain MRI images.

[0121] It should be noted that by registering the third brain MRI image with the first brain MRI image, the adjusted registration error can be obtained. By calculating the third target region and the first target region, the adjusted deformation drift can be obtained.

[0122] S204: Based on the adjusted registration error and the adjusted deformation drift, the MRI-compatible robot re-executes the following steps: robustness assessment of multiple brain puncture paths, obtaining robust target values ​​for multiple brain puncture paths, and selecting a first preset number of brain puncture paths whose robust target values ​​are less than a first threshold as target brain puncture paths.

[0123] Based on the adjusted registration error and the adjusted deformation drift, the robustness of the reprogrammed brain puncture path can be evaluated, and the target brain puncture path can be obtained.

[0124] In some implementations, MRI-compatible robots can replan brain puncture paths using iterative optimization. Specifically, the MRI-compatible robot can perform the following steps: Step 1: Obtain the 3D reconstruction error, puncture needle positioning error, the redefined coordinate range of the needle insertion point and the target point, the adjusted registration error, and the adjusted deformation drift; Step 2: Based on the redefined coordinate range of the needle insertion point and the target point, perform path planning to obtain C brain puncture paths; Step 3: Based on the 3D reconstruction error, puncture needle positioning error, adjusted registration error, and adjusted deformation drift, perform robustness evaluation on the C brain puncture paths and record the C brain puncture paths. Step 4: Continue path planning, update the C brain puncture paths, and evaluate the robustness of the updated C brain puncture paths. If the robustness target value of brain puncture path A in the updated C brain puncture paths is less than the robustness target value of brain puncture path B in the D recorded brain puncture paths, then brain puncture path A is used to replace brain puncture path B. Step 5: Determine whether the termination condition has been met. If the termination condition has not been met, repeat steps 2 to 5. If the termination condition has been met, input the currently recorded D brain puncture paths as the target brain puncture paths.

[0125] In this case, both C and D are positive numbers, and D is less than or equal to C.

[0126] In the second embodiment of this application, by replanning the brain puncture path, a new brain puncture path can be generated for doctors to choose from when no path is currently available, thereby avoiding the situation where no path is available.

[0127] The following section describes the real-time evaluation process of an MRI-compatible robot during brain puncture surgery, assuming a selected execution path.

[0128] Please refer to Figure 3. The specific steps of the third embodiment of this application are as follows:

[0129] S301: The MRI-compatible robot responds to a selection command and determines the execution path from the target brain puncture path.

[0130] After the doctor visualizes the topographic map of the target brain puncture path from the MRI-compatible robot, if an executable path is found in the target brain puncture path, the doctor can select the executable path. At this time, the MRI-compatible robot can perform brain puncture surgery according to the selected executable path.

[0131] S302: The MRI-compatible robot acquires the real-time brain MRI image corresponding to the execution path, registers and fuses the real-time brain MRI image with the first brain MRI image to obtain the fourth brain MRI image and outputs it.

[0132] During brain puncture surgery, the MRI-compatible robot can also assess the risk of the execution path in real time based on the corresponding real-time brain MRI images, thereby determining whether the execution path has a high risk.

[0133] To obtain images more quickly during surgery, the resolution of real-time brain MRI images can be lower than that of the first-line brain MRI images. In this case, by registering and fusing the real-time brain MRI images with the first-line brain MRI images, a clearer brain MRI image can be obtained.

[0134] The MRI-compatible robot can also facilitate doctors to obtain information on the implementation of brain puncture surgery by outputting fourth brain MRI images in real time.

[0135] S303: MRI-compatible robots perform 3D modeling of fourth brain MRI images and assess the real-time hazards of the execution path.

[0136] By performing three-dimensional modeling on the fourth brain MRI image, the location information of the target brain functional area, blood vessels and nerves can be obtained. At the same time, the actual execution status of the MRI-compatible robot on the execution path can be obtained, and the real-time danger of the execution path can be assessed based on the actual execution status.

[0137] Specifically, the real-time risk of the execution path can be assessed by calculating the cumulative risk distance between each point on the execution path and blood vessels, nerves, and target brain functional areas.

[0138] S304: If an NMR-compatible robot determines that the real-time hazard of the execution path is higher than the hazard threshold, it will output a hazard warning signal and re-plan the path.

[0139] The specific value of the risk threshold can be set according to actual needs. When the real-time risk exceeds the risk threshold, the execution path can be considered to have a significant risk and may cause unnecessary harm to the patient, thus requiring re-planning of the path.

[0140] It should be noted that the re-path planning can be performed based on the principle in the second embodiment, using the fourth brain MRI image (i.e., the brain MRI image obtained by registering and fusing the real-time brain MRI image with the first brain MRI image).

[0141] In the third embodiment of this application, by conducting real-time risk assessment during brain puncture surgery, the risk of the execution path can be controlled, thereby reducing the potential harm to the patient caused by the high risk of the execution path.

[0142] The specific implementation of this application will be explained below using a specific application scenario.

[0143] Please refer to Figure 4. The specific steps of the fourth embodiment of this application are as follows:

[0144] S401: The MRI-compatible robot acquires the first brain MRI image taken before the brain puncture surgery, performs three-dimensional reconstruction on the first brain MRI image, and obtains the first target region.

[0145] In this embodiment, the first brain MRI image can be acquired using a 3.0T MRI scanner with a slice spacing of 1mm. It should be noted that the slice spacing can also be set to other values, as long as the requirement for clear imaging is met.

[0146] S402: The MRI-compatible robot performs path planning based on the first target area, resulting in multiple brain puncture paths.

[0147] It should be noted that for a detailed explanation of step S402, please refer to the explanation of step S102, which will not be repeated here.

[0148] Please refer to Figure 6, which is a schematic diagram of a path planning method provided in this embodiment. It should be noted that the feasible path in Figure 6 refers to the brain puncture path, and the infeasible path refers to any path in the initial path other than the brain puncture path. The feasible path here does not mean that it can be used as the execution path for brain puncture surgery.

[0149] S403: The MRI-compatible robot acquires the second brain MRI image obtained during brain puncture surgery, and registers the second brain MRI image with the first brain MRI image to obtain the registration error.

[0150] In this embodiment, the second brain MRI image can be acquired by a 3.0T MRI scanner with a slice spacing of 3mm. It should be noted that the slice spacing can also be set to other values, such as 2mm, as long as the requirement for fast imaging is met.

[0151] It should be noted that in the third embodiment, the acquisition conditions of real-time brain MRI images are based on the same principle as those of the second brain MRI images. Therefore, real-time brain MRI images can also be acquired by a 3.0T MRI scanner with a slice spacing of 3mm or 2mm.

[0152] S404: The MRI-compatible robot performs three-dimensional reconstruction of the second brain MRI image to obtain the second target region and calculates the deformation drift of the second target region relative to the first target region.

[0153] S405: The NMR-compatible robot determines the range of the first interval based on the 3D reconstruction error, registration error, and deformation drift.

[0154] It should be noted that the first interval range is the disturbance space B shown in Figure 5. δ .

[0155] Three-dimensional reconstruction error, registration error, and deformation drift are collectively referred to as brain tissue position error. Assuming the threshold range of brain tissue position error is 0 to 0.5 mm, the first interval range can be [-0.5 mm, +0.5 mm]. It should be noted that the threshold range of brain tissue position error can be obtained based on actual conditions and is not limited to the example above.

[0156] S406: For each path point in multiple brain puncture paths, the MRI-compatible robot generates multiple perturbation solutions within a first interval.

[0157] At this point, for each path point, multiple perturbation solutions can be generated within the interval [-0.5mm, +0.5mm]. Each perturbation solution can be represented as y i .

[0158] S407: The MRI-compatible robot calculates the cumulative danger distance of the perturbation solution within the puncture needle positioning error based on the target brain functional area and the location of the blood vessels and nerves closest to the perturbation solution.

[0159] Assuming the puncture needle positioning error is 0 to 0.4 mm, it is necessary to calculate the cumulative danger distance q(y) between the perturbation solution and the target brain functional area, blood vessels, and nerves within an error range of 0.4 mm. i Specifically, in order to calculate q(y) i The error range of 0.4mm can be divided into M elements. The number of elements with respect to y can be calculated. i There are intersection points, thus the cumulative danger distance can be calculated. If two out of M elements intersect with y... i If there is an intersection, then the cumulative danger distance is: M is used to perform grid division to solve the cumulative danger distance, and the specific value of M can be set according to actual needs.

[0160] S408: The MRI-compatible robot performs robustness assessment on multiple brain puncture paths based on the cumulative danger distance of each perturbation solution in multiple perturbation solutions, and obtains the robustness target value of multiple brain puncture paths.

[0161] It should be noted that for a detailed explanation of step S408, please refer to the explanation of step S105, which will not be repeated here.

[0162] S409: The MRI-compatible robot selects a first preset number of brain puncture paths from among multiple brain puncture paths whose robust target value is less than a first threshold as the target brain puncture path.

[0163] In this embodiment, multiple brain puncture paths can be sorted from low to high according to the robustness target value, and the first preset number of brain puncture paths can be used as the target brain puncture path.

[0164] S410: The MRI-compatible robot fits and visualizes the topographic map of the target brain puncture path based on the length and cumulative danger distance of the target brain puncture path.

[0165] Please refer to Figure 7, which is a visualization diagram provided in this embodiment. Since the target brain puncture path is a brain puncture path with a small robust target value, it can also be called a robust path. Figure 7 shows the length and cumulative danger distance (i.e., risk value) of each robust path in the path topography map.

[0166] It should be noted that after step S410, there may be two other situations: there is an executable path in the target brain puncture path, and there is no executable path in the target brain puncture path. These will not be elaborated here. For details, please refer to the second and third embodiments.

[0167] In the fourth embodiment of this application, by adding a safety threshold to the path planning algorithm, the safety and risk of each planned brain puncture path are improved, thereby enhancing the safety and reliability of brain puncture surgery.

[0168] Please refer to Figure 8. This application provides a brain puncture path planning device 800, which is applied to an MRI-compatible robot. The device includes: a three-dimensional reconstruction module 801, a path planning module 802, a registration module 803, and an evaluation module 804.

[0169] 3D Reconstruction Module 801: Used to acquire the first brain MRI image collected before brain puncture surgery, and to perform 3D reconstruction on the first brain MRI image to obtain the first target region; the first target region includes: the area where the target brain puncture tissue, target brain functional area, blood vessels and nerves are located in the first brain MRI image.

[0170] Path planning module 802: Used to plan paths based on the first target region to obtain multiple brain puncture paths.

[0171] Registration module 803: used to acquire the second brain MRI image acquired during brain puncture surgery, and to register the second brain MRI image with the first brain MRI image to obtain the registration error.

[0172] The three-dimensional reconstruction module 801 is also used to perform three-dimensional reconstruction on the second brain MRI image to obtain the second target region and calculate the deformation drift of the second target region relative to the first target region; the second target region includes: the area where the target brain puncture tissue, target brain functional area, blood vessels and nerves are located in the second brain MRI image.

[0173] Evaluation module 804: Used to perform robustness evaluation on multiple brain puncture paths based on puncture needle positioning error, three-dimensional reconstruction error, registration error and deformation drift, and obtain robust target values ​​for multiple brain puncture paths.

[0174] Evaluation module 804: is also used to select a first preset number of brain puncture paths among multiple brain puncture paths whose robust target value is less than a first threshold as target brain puncture paths.

[0175] Optionally, the evaluation module 804 includes: a determination unit, a generation unit, a calculation unit, and an evaluation unit.

[0176] Determining unit: used to determine the range of the first interval based on 3D reconstruction error, registration error and deformation drift.

[0177] Generation unit: Used to generate multiple perturbation solutions for each path point in multiple brain puncture paths within a first interval.

[0178] Calculation unit: used to calculate the cumulative danger distance of the perturbation solution within the puncture needle positioning error, based on the target brain functional area and the location of the nearest blood vessels and nerves to the perturbation solution.

[0179] Evaluation unit: Used to evaluate the robustness of multiple brain puncture paths based on the cumulative danger distance of each perturbation solution in multiple perturbation solutions, and obtain the robustness target value of multiple brain puncture paths.

[0180] Optionally, the path planning module 802 includes: an acquisition unit, a generation unit, a calculation unit, and an evaluation unit.

[0181] The acquisition unit is used to acquire the coordinate range of the needle insertion point and the coordinate range of the target point; the coordinate range of the needle insertion point and the coordinate range of the target point are determined based on the first target area.

[0182] The generation unit is used to generate multiple perturbation solutions for each path point in the initial path among multiple initial paths, within a second interval. The multiple initial paths are obtained based on the coordinate range of the needle entry point and the coordinate range of the target point. The second interval is determined based on a safety threshold.

[0183] The calculation unit is used to calculate the cumulative danger distance of the perturbation solution within the puncture needle positioning error, based on the target brain functional area and the location of the nearest blood vessels and nerves to the perturbation solution.

[0184] The evaluation unit is used to evaluate the robustness of multiple initial paths based on the cumulative danger distance of each perturbation solution in multiple perturbation solutions, and obtain the robustness target value of multiple initial paths.

[0185] The evaluation unit is also used to select a second preset number of initial paths with a robust target value less than a second threshold as brain puncture paths; the second preset number is greater than the first preset number.

[0186] Optionally, the evaluation unit is specifically implemented using the following formula:

[0187] Among them, f eff (x bs ,x be x is the robustness target value for the initial path. bs x is the needle entry point. be For the target point, d(x) bs ,x be q(y) represents the path length from the needle insertion point to the target point, K represents the number of path points in the initial path, N represents the number of perturbation solutions corresponding to each path point, and q(y) represents the number of perturbation solutions corresponding to each path point. k,i ) represents the path point y k The cumulative danger distance of the i-th perturbation solution.

[0188] Optionally, a brain puncture path planning device 800 further includes a visualization module 805.

[0189] Visualization Unit 805: Used to fit and visualize the topographic map of the target brain puncture path based on the length and cumulative danger distance of the target brain puncture path.

[0190] Optionally, a brain puncture path planning device 800 further includes a fusion module 806 and an adjustment module 807.

[0191] Fusion module 806: In response to the replanning command, it fuses the first brain MRI image and the second brain MRI image to obtain the third brain MRI image.

[0192] 3D Reconstruction Module 801: Also used to perform 3D reconstruction of the third brain MRI image to obtain the third target region.

[0193] Path planning module 802: It is also used to replan the path based on the third target region to obtain multiple replanned brain puncture paths, and replace the brain puncture path with the replanned brain puncture path; the third target region includes: the area where the target brain puncture tissue, target brain functional area, blood vessels and nerves are located in the third brain MRI image.

[0194] Adjustment module 807: used to adjust registration error and deformation drift based on the third brain MRI image and the first brain MRI image;

[0195] Evaluation module 804: It is also used to re-execute the following steps based on the adjusted registration error and the adjusted deformation drift: to perform robustness evaluation on multiple brain puncture paths, obtain robust target values ​​for multiple brain puncture paths, and select a first preset number of brain puncture paths with robust target values ​​less than a first threshold as target brain puncture paths.

[0196] Optionally, a brain puncture path planning device 800 further includes: a determination module 808, a fusion module 806, and...

[0197] Determine module 808: In response to a selection command, determine the execution path from the target brain puncture path.

[0198] Fusion module 806: Used to acquire the real-time brain MRI image corresponding to the execution path, register and fuse the real-time brain MRI image with the first brain MRI image to obtain the fourth brain MRI image and output it.

[0199] Assessment module 804: It is also used to perform three-dimensional modeling of the fourth brain MRI image and assess the real-time risk of the execution path.

[0200] Evaluation module 804: It is also used to output a danger warning signal and re-plan the path if the real-time danger of the execution path is higher than the danger threshold.

[0201] Optionally, the evaluation module 804 includes: a selection unit, an evaluation unit, a recording unit, an iteration unit, and a replacement unit.

[0202] Selection unit: used to select A first brain puncture paths as candidate paths from multiple brain puncture paths.

[0203] Evaluation unit: Used to evaluate the robustness of candidate paths based on puncture needle positioning error, 3D reconstruction error, registration error and deformation drift, and obtain the robustness target value of the candidate paths.

[0204] Recording unit: Used to record B candidate paths out of A candidate paths whose robust target value is less than the third threshold, thus obtaining the robust path. Here, A and B are both positive integers, and B is less than or equal to A.

[0205] Iteration unit: used to select A second brain puncture paths from multiple brain puncture paths to replace the candidate paths, and re-execute the robustness evaluation of the candidate paths based on puncture needle positioning error, 3D reconstruction error, registration error and deformation drift to obtain the robustness target value of the candidate paths.

[0206] Replacement unit: Used to replace the robust path with the second brain puncture path if the robust target value of the second brain puncture path is less than the robust target value of the robust path.

[0207] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0208] It should be noted that the brain puncture path planning device provided in the above embodiments is only illustrated by the division of the above functional modules when realizing the brain puncture path planning function. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the brain puncture path planning device can be divided into different functional modules to complete all or part of the functions described above. In addition, the brain puncture path planning device and the brain puncture path planning method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process is detailed in the method embodiments, which will not be repeated here.

[0209] Please refer to Figure 9. This application also provides an NMR-compatible robot 900, including a processor 901 and a memory 902.

[0210] The processor 901 is coupled to the memory 902, which stores at least one computer program instruction. The processor 901 loads and executes the at least one computer program instruction to enable the computer device to implement the above-mentioned brain puncture path planning method.

[0211] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0212] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for brain puncture path planning, characterized in that, The method, applied to NMR-compatible robots, includes: A first brain MRI image acquired before a brain puncture surgery is obtained, and the first brain MRI image is reconstructed in three dimensions to obtain a first target region; the first target region includes: the area where the target brain puncture tissue, target brain functional area, blood vessels and nerves are located in the first brain MRI image; Based on the first target region, multiple brain puncture paths are obtained through path planning. A second brain MRI image acquired during a brain puncture surgery was obtained, and the second brain MRI image was registered with the first brain MRI image to obtain the registration error. The second brain MRI image is reconstructed in three dimensions to obtain a second target region, and the deformation drift of the second target region relative to the first target region is calculated; the second target region includes: the area where the target brain puncture tissue, target brain functional area, blood vessels and nerves are located in the second brain MRI image; Based on the puncture needle positioning error, three-dimensional reconstruction error, registration error, and deformation drift, the robustness of the multiple brain puncture paths is evaluated to obtain the robustness target value of the multiple brain puncture paths. The brain puncture path with a robust target value less than a first threshold among the multiple brain puncture paths is selected as the target brain puncture path.

2. The method according to claim 1, characterized in that, The robustness assessment of the multiple brain puncture paths is performed based on the puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift to obtain robustness target values ​​for the multiple brain puncture paths, including: The first interval range is determined based on the 3D reconstruction error, the registration error, and the deformation drift; For each path point in the plurality of brain puncture paths, multiple perturbation solutions are generated within the first interval. Based on the target brain functional area and the location of the blood vessels and nerves closest to the perturbation solution, calculate the cumulative danger distance of the perturbation solution within the puncture needle positioning error; Based on the cumulative danger distance of each of the multiple perturbation solutions, the robustness of the multiple brain puncture paths is evaluated to obtain the robustness target value of the multiple brain puncture paths.

3. The method according to claim 1, characterized in that, The path planning based on the first target region yields multiple brain puncture paths, including: Obtain the coordinate range of the needle insertion point and the coordinate range of the target point; the coordinate range of the needle insertion point and the coordinate range of the target point are determined based on the first target region; For each of the multiple initial paths, multiple perturbation solutions are generated for each path point in the initial path within a second interval; the multiple initial paths are obtained based on the coordinate range of the needle insertion point and the coordinate range of the target point; the second interval is determined based on a safety threshold. Based on the target brain functional area and the location of the blood vessels and nerves closest to the perturbation solution, calculate the cumulative danger distance of the perturbation solution within the puncture needle positioning error; Based on the cumulative danger distance of each of the multiple perturbation solutions, the robustness of the multiple initial paths is evaluated to obtain the robustness target value of the multiple initial paths; The first preset number of initial paths whose robust target value is less than the second threshold is selected as brain puncture paths; the second preset number is greater than the first preset number.

4. The method according to claim 3, characterized in that, The robustness assessment of the multiple initial paths is performed based on the cumulative danger distance of each of the multiple perturbation solutions, and the robustness target value of the multiple initial paths is obtained through the following formula: Among them, f eff (x bs ,x be x is the robust target value of the initial path. bs x is the needle entry point. be For the target point, d(x) bs ,x be q(y) represents the path length from the needle insertion point to the target point, K represents the number of path points in the initial path, N represents the number of perturbation solutions corresponding to each path point, and q(y) represents the number of perturbation solutions corresponding to each path point. k,i ) represents the path point y k The cumulative danger distance of the i-th perturbation solution.

5. The method according to claim 1, characterized in that, After selecting a first preset number of brain puncture paths with robust target values ​​less than a first threshold as target brain puncture paths, the method further includes: Based on the length and cumulative danger distance of the target brain puncture path, a path topography map corresponding to the target brain puncture path is fitted and visualized and output.

6. The method according to claim 5, characterized in that, After fitting and visualizing the path topography map corresponding to the target brain puncture path, the method further includes: In response to the replanning instruction, the first brain MRI image and the second brain MRI image are fused to obtain a third brain MRI image; The third brain MRI image is reconstructed in three dimensions to obtain a third target region. Based on the third target region, the path is replanned to obtain multiple replanned brain puncture paths. The replanned brain puncture paths are then used to replace the original brain puncture paths. The third target region includes: the area where the target brain puncture tissue, target brain functional area, blood vessels, and nerves are located in the third brain MRI image. Based on the third brain MRI image and the first brain MRI image, adjust the registration error and the deformation drift; Based on the adjusted registration error and the adjusted deformation drift, the following steps are repeated: the robustness assessment of the plurality of brain puncture paths is performed to obtain the robustness target value of the plurality of brain puncture paths, and the brain puncture paths with a first preset number of robustness target values ​​less than a first threshold are selected as target brain puncture paths.

7. The method according to claim 5, characterized in that, After fitting and visualizing the path topography map corresponding to the target brain puncture path, the method further includes: In response to a selection command, an execution path is determined from the target brain puncture path; Obtain the real-time brain MRI image corresponding to the execution path, register and fuse the real-time brain MRI image with the first brain MRI image to obtain the fourth brain MRI image and output it. A three-dimensional model was created from the fourth brain MRI image, and the real-time risk of the execution path was assessed. If the real-time danger of the execution path is higher than the danger threshold, a danger warning signal is output and the path is replanned.

8. The method according to claim 1, characterized in that, The robustness assessment of the multiple brain puncture paths is performed based on the puncture needle positioning error, 3D reconstruction error, registration error, and deformation drift to obtain robustness target values ​​for the multiple brain puncture paths, including: A first brain puncture paths are selected from the plurality of brain puncture paths as candidate paths; Based on the puncture needle positioning error, the three-dimensional reconstruction error, the registration error, and the deformation drift, the robustness evaluation of the candidate path is performed to obtain the robustness target value of the candidate path; Record B candidate paths from A candidate paths whose robust target value is less than the third threshold to obtain the robust path; where A and B are both positive integers, and B is less than or equal to A; A second brain puncture path is selected from the plurality of brain puncture paths to replace the candidate path, and the robustness evaluation of the candidate path is re-executed based on the puncture needle positioning error, the three-dimensional reconstruction error, the registration error and the deformation drift to obtain the robustness target value of the candidate path; If the robust target value of the second brain puncture path is less than the robust target value of the robust path, then the second brain puncture path is used to replace the robust path.

9. A brain puncture path planning device, characterized in that, The device, used in NMR-compatible robots, includes: The three-dimensional reconstruction module is used to acquire the first brain MRI image collected before the brain puncture surgery, and to perform three-dimensional reconstruction on the first brain MRI image to obtain the first target region; the first target region includes: the area where the target brain puncture tissue, target brain functional area, blood vessels and nerves are located in the first brain MRI image; The path planning module is used to plan a path based on the first target region to obtain multiple brain puncture paths; The registration module is used to acquire a second brain MRI image acquired during brain puncture surgery, and to register the second brain MRI image with the first brain MRI image to obtain the registration error; The three-dimensional reconstruction module is also used to perform three-dimensional reconstruction on the second brain MRI image to obtain a second target region, and to calculate the deformation drift of the second target region relative to the first target region; the second target region includes: the area where the target brain puncture tissue, target brain functional area, blood vessels and nerves are located in the second brain MRI image; An evaluation module is used to perform robustness evaluation on the multiple brain puncture paths based on puncture needle positioning error, three-dimensional reconstruction error, registration error and deformation drift, and obtain robustness target values ​​for the multiple brain puncture paths. The evaluation module is further configured to select a first preset number of brain puncture paths among the plurality of brain puncture paths whose robust target value is less than a first threshold as target brain puncture paths.

10. A nuclear magnetic resonance compatible robot, characterized in that, The NMR-compatible robot includes a processor coupled to a memory storing at least one computer program instruction, which is loaded and executed by the processor to enable the NMR-compatible robot to implement the method of any one of claims 1-8.

Images