Craniotomy robot path planning method, device, equipment and medium
By using a 3D model of the skull and a fast expanding random tree algorithm to plan the craniotomy path, combined with multi-order Bézier curves and mirror extension filtering algorithms, the problem of inaccurate path planning in traditional craniotomy is solved, achieving a milling effect with high stability and strong perception.
Patent Information
- Application Number
- CN202410899813.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-05
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-07-05
AI Technical Summary
In traditional craniotomy, the milling path is selected based on experience, which leads to high operational difficulty and poor stability. Too many drilling points can easily cause irreversible trauma.
A path planning method based on a 3D skull model is adopted. The fast extended random tree algorithm and attraction point technique are used to plan the skull milling path through a drilling point. The path fitting and smoothing are combined with multi-order Bézier curves and mirror extension filtering algorithm.
It achieves path planning with high precision, strong perception, and stable milling force, reducing the difficulty and trauma risk of craniotomy.
Smart Images

Figure CN118453119B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the fields of neurosurgery and artificial intelligence technology, and in particular to a craniotomy robot path planning method, device, equipment and medium. Background Art
[0002] In the field of neurosurgery, craniotomy is an important way to treat craniocerebral trauma, brain tumors, intracranial hemorrhage, cerebral hematoma, cerebrospinal fluid overflow, traumatic asphyxia, cerebrovascular diseases, etc. In traditional craniotomy, doctors select multiple drilling points according to the location of the tumor and connect the drilling points to determine the milling path. However, this milling path selected based on experience may deviate from the ideal surgical path. At the same time, too many drilling points can easily cause the risk of drilling points breaking through the dura mater and more irreversible trauma, making craniotomy more difficult to operate and less stable. Summary of the invention
[0003] The present invention provides a craniotomy robot path planning method, device, equipment and medium, which are used to solve the defects of high operating difficulty and poor stability of craniotomy in the prior art, and achieve the advantages of high accuracy, strong perception ability, stable milling force, etc., and complete the path planning of skull milling based on only one drilling point.
[0004] The present invention provides a craniotomy robot path planning method, comprising the following steps.
[0005] Determine the skull milling position and drilling point location based on the skull 3D model and tumor location;
[0006] Determine the drilling point position as the root node, determine the tumor position as the obstacle position, and initialize a random tree;
[0007] Determine an attraction point based on the root node and the obstacle position;
[0008] Based on the attraction points and pre-set search weights, a skull milling path is determined using a rapidly expanding random tree algorithm.
[0009] According to a craniotomy robot path planning method provided by the present invention, the method determines the skull milling path based on the attraction points and the preset search weights using a fast expanding random tree algorithm, including:
[0010] For each path node, a random point is selected, and the direction and the target point are determined according to the search weight, the path node and the unit vector of the random point;
[0011] Determine on the random tree the point closest to the target point in the search radius as a neighboring node, and determine, based on the neighboring nodes, the path node closest to the target point in the direction as a new node;
[0012] Based on the search weight, the new node and the attraction point, determine the parent node corresponding to the new node so that the path length from the new node through the parent node to the root node is the shortest;
[0013] The skull milling path is determined using each of the path nodes and a parent node corresponding to the path node.
[0014] According to a craniotomy robot path planning method provided by the present invention, the attraction point includes a first attraction point and a second attraction point, and the attraction point is determined based on the root node and the obstacle position, including:
[0015] Determining the center of gravity of the obstacle position according to the obstacle position;
[0016] The projection point of the root node relative to the center of gravity is determined as the first attraction point, and the root node is determined as the second attraction point.
[0017] According to a craniotomy robot path planning method provided by the present invention, the parent node corresponding to the new node is determined based on the search weight, the new node and the attraction point so that the path length from the new node through the parent node to the root node is the shortest, including:
[0018] Before the new node becomes the first attraction point, the path length between the new node and the root node is determined by using the first attraction point, and the parent node corresponding to the new node is determined so that the path length from the new node through the parent node to the root node is the shortest;
[0019] After the new node becomes the first attraction point, the second attraction point is used to determine the path length between the new node and the root node, and the parent node corresponding to the new node is determined so that the path length from the new node through the parent node to the root node is the shortest.
[0020] According to a craniotomy robot path planning method provided by the present invention, the method further includes:
[0021] Determine the curvature of the skull milling path using the path node, the parent node and the parent node corresponding to the parent node;
[0022] When the curvature of the skull milling path is greater than a preset curvature value, the skull milling path is reselected.
[0023] According to a craniotomy robot path planning method provided by the present invention, the method further includes:
[0024] The search radius is modified according to the number of the random trees, and the parent node corresponding to the new node is modified according to the modified search radius.
[0025] According to a craniotomy robot path planning method provided by the present invention, after determining the skull milling path, the method further includes:
[0026] The skull milling path is fitted by a multi-order Bezier curve, and the skull milling path is smoothed by a mirror extension filtering algorithm.
[0027] The present invention also provides a craniotomy robot path planning device, comprising the following modules:
[0028] The first determination module is used to determine the skull milling position and the drilling point position according to the skull three-dimensional model and the tumor position;
[0029] A processing module, used to determine the drilling point position as a root node, determine the tumor position as an obstacle position, and initialize a random tree;
[0030] A second determination module, configured to determine an attraction point based on the root node and the obstacle position;
[0031] The third determination module is used to determine the skull milling path by using a fast expanding random tree algorithm based on the attraction points and a preset search weight.
[0032] According to a craniotomy robot path planning device provided by the present invention, the third determination module is specifically used for:
[0033] For each path node, a random point is selected, and the direction and the target point are determined according to the search weight, the path node and the unit vector of the random point;
[0034] Determine on the random tree the point closest to the target point in the search radius as a neighboring node, and determine, based on the neighboring nodes, the path node closest to the target point in the direction as a new node;
[0035] Based on the search weight, the new node and the attraction point, determine the parent node corresponding to the new node so that the path length from the new node through the parent node to the root node is the shortest;
[0036] The skull milling path is determined using each of the path nodes and a parent node corresponding to the path node.
[0037] According to a craniotomy robot path planning device provided by the present invention, the attraction point includes a first attraction point and a second attraction point, and the second determination module is specifically used for:
[0038] Determining the center of gravity of the obstacle position according to the obstacle position;
[0039] The projection point of the root node relative to the center of gravity is determined as the first attraction point, and the root node is determined as the second attraction point.
[0040] According to a craniotomy robot path planning device provided by the present invention, the third determination module is specifically used for:
[0041] Before the new node becomes the first attraction point, the path length between the new node and the root node is determined by using the first attraction point, and the parent node corresponding to the new node is determined so that the path length from the new node through the parent node to the root node is the shortest;
[0042] After the new node becomes the first attraction point, the second attraction point is used to determine the path length between the new node and the root node, and the parent node corresponding to the new node is determined so that the path length from the new node through the parent node to the root node is the shortest.
[0043] According to a craniotomy robot path planning device provided by the present invention, the third determination module is also used for:
[0044] Determine the curvature of the skull milling path using the path node, the parent node and the parent node corresponding to the parent node;
[0045] When the curvature of the skull milling path is greater than a preset curvature value, the skull milling path is reselected.
[0046] According to a craniotomy robot path planning device provided by the present invention, the third determination module is also used for:
[0047] The search radius is modified according to the number of the random trees, and the parent node corresponding to the new node is modified according to the modified search radius.
[0048] According to a craniotomy robot path planning device provided by the present invention, the third determination module is also used for:
[0049] The skull milling path is fitted by a multi-order Bezier curve, and the skull milling path is smoothed by a mirror extension filtering algorithm.
[0050] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a craniotomy robot path planning method as described above is implemented.
[0051] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described craniotomy robot path planning methods.
[0052] The craniotomy robot path planning method provided by the present invention determines the skull milling position and the drilling point position through the skull three-dimensional model and the tumor position, then determines the drilling point position as the root node, determines the tumor position as the obstacle position, initializes the random tree, and then determines the attraction point based on the root node and the obstacle position. Finally, based on the attraction point and the pre-set search weight, the skull milling path is determined using a fast expansion random tree algorithm, so as to achieve the path planning of skull milling based on only one drilling point. The method has the advantages of high accuracy, strong perception ability, stable milling force, etc., which greatly reduces the difficulty of skull surgery. BRIEF DESCRIPTION OF THE DRAWINGS
[0053] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0054] Figure 1 It is a flow chart of the craniotomy robot path planning method provided by the present invention.
[0055] Figure 2 It is a schematic diagram of the principle of setting the attraction points provided by the present invention.
[0056] Figure 3 It is a schematic diagram of path planning with different search weights provided by the present invention.
[0057] Figure 4 It is a schematic diagram of the principle of curvature-limited redrawing provided by the present invention.
[0058] Figure 5 It is a schematic diagram of the principle of the edge collapse phenomenon provided by the present invention.
[0059] Figure 6 It is a comparative schematic diagram after mirror filtering provided by the present invention.
[0060] Figure 7 It is a specific flow chart of the craniotomy robot path planning method provided by the present invention.
[0061] Figure 8 It is a structural schematic diagram of the craniotomy robot path planning device provided by the present invention.
[0062] Fig. 9It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION
[0063] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0064] In the field of neurosurgery, craniotomy is an important way to treat craniocerebral trauma, brain tumors, intracranial hemorrhage, cerebral hematoma, cerebrospinal fluid overflow, traumatic asphyxia, cerebrovascular diseases, etc. In traditional craniotomy, doctors select multiple drilling points according to the location of the tumor and connect the drilling points to determine the milling path. However, this milling path selected based on experience may deviate from the ideal surgical path. At the same time, too many drilling points can easily cause the risk of drilling points breaking through the dura mater and more irreversible trauma, making craniotomy more difficult to operate and less stable.
[0065] Therefore, a craniotomy robot path planning method is urgently needed to solve the above problems.
[0066] Combine the following Figure 1-Figure 9 The present invention describes the craniotomy robot path planning method, device, equipment and medium.
[0067] Figure 1 FIG. 1 is one of the flow charts of the craniotomy robot path planning method provided by the present invention, such as Figure 1 As shown, the method includes the following:
[0068] Step 101: Determine the skull milling position and drilling point position according to the skull three-dimensional model and the tumor position.
[0069] During the specific implementation, the computerized tomography (CT) scan images of the skull and tumor are first imported to determine the skull milling position and drilling point position.
[0070] In this step, the feed speed range of the milling cutter and the like may also be determined for calculating the desired curvature in step 104 .
[0071] Step 102: determine the drilling point position as the root node, determine the tumor position as the obstacle position, and initialize the random tree.
[0072] In the specific implementation, the tumor is projected to the skull milling position, the obstacle outline is generated, and the random tree is initialized. In this step, in order to avoid the planned path from contacting the tumor projection, the triangle vertex contacted by the tumor projection is taken as the center of the sphere, and all triangle vertices within the preset radius are removed from the skull triangle. The removed triangle library is used as the state space in the path planning to speed up the convergence of the algorithm.
[0073] Step 103: Determine the attraction point based on the root node and the obstacle position.
[0074] In a specific implementation, the attraction point includes a first attraction point and a second attraction point. When determining the attraction point, the center of gravity of the obstacle position is determined according to the obstacle position, and then the projection point of the root node relative to the center of gravity is determined as the first attraction point, and the root node is determined as the second attraction point. Specifically, Figure 2 As shown, since the starting point and the end point of milling are both drilling points, the end point cannot be used as the attraction point in the initial stage. In one possible implementation method, the centroid point symmetry method is used to confirm the centroid of the tumor projection area, and the projection point of the drilling point relative to the centroid is used as the first attraction point in the initial stage. After reaching the attraction point, the end point is used as the second attraction point again. Of course, the first attraction point can also be driven by other methods, such as using the center of the solid as the projection point, etc., and the embodiments of the present disclosure are not limited to this.
[0075] Step 104: Based on the attraction points and the preset search weights, a skull milling path is determined using a fast expanding random tree algorithm.
[0076] In the specific implementation, for each path node, a random point is selected, and the target point direction and the target point are determined according to the search weight, the path node and the unit vector of the random point. Then, the point closest to the target point direction within the search radius is determined as the neighboring node on the random tree, and the path node closest to the target point direction is determined as the new node based on the neighboring node. Then, based on the search weight, the new node and the attraction point, the parent node corresponding to the new node is determined so that the path length from the new node through the parent node to the root node is the shortest. Finally, each path node and the parent node corresponding to the path node are used to determine the skull milling path.
[0077] Specifically, we first randomly select a random point n in the state space. rand , according to the search weight w c, the unit vectors of the path nodes and random points are used to calculate the direction of the target point, and a point far enough away in the direction of the target point is selected as the target point. Then the point closest to the target point is found in the tree as the neighboring node. According to the set step range, the space between the balls with the neighboring node as the center and the step range as the radius is selected to be the point closest to the target point as the new node. Finally, the search radius r is calculated according to the number of trees, and the potential parent nodes on the tree are searched with the new node as the center and r as the radius. When all potential parent nodes are used as parent nodes, the path length from the new node to the starting point is calculated, and the point with the shortest path length is selected as the parent node. The new node is connected to the parent node, and the connection between the neighboring node and the new node is removed. When calculating the path, the following formula is used for calculation:
[0078]
[0079] in, ; .
[0080] r represents the search step length of the path node, w c It represents the search weight of the direction of the attraction point generated when the target point searches for the node, d new represents a new node, d near represents the neighboring nodes, n rand and n goal It represents the unit vector of the path node and the target point in the path search area, such as Figure 3 As shown in the figure, examples of path planning under different search weights are w c Take examples at 0.1, 0.5, and 0.9.
[0081] In this step, in order to avoid inaccurate routes due to too large a search radius and too slow efficiency due to a small search radius, a large search radius is used in the early stage to explore the map more quickly, and a small radius is used in the later stage to search the most accurate path. The search radius can also be corrected according to the number of random trees, and the parent node corresponding to the new node is corrected according to the corrected search radius. At the same time, the minimum value of the early radius is limited. If the search radius is smaller than the step size in the early stage, an infinite loop will occur. Specifically, the following formula can be used to determine the stage according to the number of trees and dynamically adjust the radius:
[0082]
[0083] Where n is the number of nodes in the tree, d is the dimension of the space, γ is the coefficient of variation, and η is the maximum radius.
[0084] In this step, the curvature of the skull milling path can also be determined by using the path node, the parent node, and the parent node corresponding to the parent node, and then when the curvature of the skull milling path is greater than the preset curvature value, the skull milling path is reselected. Specifically, Figure 4 As shown in the figure, after each exploration of the next target point and path redrawing, the new node, the parent node, and the parent node of the parent node are determined according to the three-dimensional model of the skull. new d parent d ancestry Skull thickness h new 、h parent 、h ancestry , and take the maximum value h max As the input of the neural network, the expected curvature K that meets the torque limit is obtained according to the previously trained neural network. e Calculate d new d parent d ancestry The curvature K of the curve fitted by the cubic polynomial at three points is consistent with the expected curvature K e If the curvature exceeds the limit, repeat the above steps to reselect new nodes. Calculate the points where the length is shortened during the rewiring again and d new d parent Whether the curvature between them meets the constraint, a new node d new Reselect; if it meets the requirements, connect to the new node. It should be noted that the curvature can be calculated through a neural network model or other forms, which is not limited in the embodiments of the present disclosure.
[0085] In addition, after determining the skull milling path, the skull milling path can be fitted by a multi-order Bezier curve, and the skull milling path can be smoothed by a mirror extension filtering algorithm. Specifically, the path is fitted by an n-order Bezier curve, and the path is smoothed by a mirror extension filtering algorithm. The mirror extension method can solve the edge collapse phenomenon caused by the filtering algorithm. First, by adding mirror extensions at both ends of the curve and setting up virtual endpoints, the curve is extended, and the curve after mirror extension is filtered, and the mirror extension part is deleted to restore the endpoint. The parameter equation formula for path fitting by an n-order Bezier curve is as follows:
[0086]
[0087] Where B(t) is the Bernstein basis function, n represents the order, represents the number of ways to select i out of n, Pi is the i-th control point, and t is the parameter used to represent the position of a point on the curve.
[0088] The fitted path curve is not smooth enough because it is fitted by path points. The mirror extension filtering algorithm is used to smooth the path. The mirror extension method can solve the problems caused by the filtering algorithm. Figure 5 To solve the edge collapse phenomenon shown in the figure, we first extend the curve by adding mirror extensions at both ends of the curve and setting up virtual endpoints. Then we filter the curve after mirror extension and finally remove the redundant parts. This allows the filtered trajectory to maintain smoothness while also maintaining the endpoints to avoid edge collapse. Figure 6 As shown, it is a filtering comparison using Bezier curve fitting, where the light color is before filtering and the dark color is after filtering.
[0089] The process of executing the embodiment of the present disclosure is as follows Figure 7 As shown, first initialize the random tree, then execute the determination of state space, random sampling, parent node selection, and then perform curvature detection. If it does not meet the requirements, return to re-execute the determination of state space. If it meets the requirements, rewire until reaching the end point, and finally perform fitting filtering on the skull milling path. The craniotomy robot path planning method provided by the embodiment of the present disclosure is used to introduce the idea of target point pointing expansion, and improve the sampling method of the path node of the Rapidly Exploring Random Tree (RRT) algorithm to achieve the purpose of path pruning, and use Bezier curves to smooth the path so that the final path has coherence. The curvature of the path planned by the traditional algorithm is uncertain, and a curvature-limited redrawing algorithm is used to limit the curvature of the planned path. A dynamic radius algorithm is used to reduce the convergence time, and a mirror extension filtering algorithm is used to solve the edge collapse phenomenon in the path smoothing process.
[0090] The craniotomy robot path planning device provided by the present invention is described below. The craniotomy robot path planning device described below and the craniotomy robot path planning method described above can be referenced to each other.
[0091] like Figure 8 As shown, the craniotomy robot path planning device 800 provided by the present invention includes the following modules:
[0092] The first determination module 801 is used to determine the skull milling position and the drilling point position according to the skull three-dimensional model and the tumor position;
[0093] A processing module 802 is used to determine the drilling point position as a root node, determine the tumor position as an obstacle position, and initialize a random tree;
[0094] A second determination module 803 is used to determine an attraction point based on the root node and the obstacle position;
[0095] The third determination module 804 is used to determine the skull milling path by using a fast expanding random tree algorithm based on the attraction points and the preset search weights.
[0096] According to a craniotomy robot path planning device provided by the present invention, the third determination module 804 is specifically used for:
[0097] For each path node, a random point is selected, and the direction and the target point are determined according to the search weight, the path node and the unit vector of the random point;
[0098] Determine on the random tree the point closest to the target point in the search radius as a neighboring node, and determine, based on the neighboring nodes, the path node closest to the target point in the direction as a new node;
[0099] Based on the search weight, the new node and the attraction point, determine the parent node corresponding to the new node so that the path length from the new node through the parent node to the root node is the shortest;
[0100] The skull milling path is determined using each of the path nodes and a parent node corresponding to the path node.
[0101] According to a craniotomy robot path planning device provided by the present invention, the attraction point includes a first attraction point and a second attraction point, and the second determination module 803 is specifically used for:
[0102] Determining the center of gravity of the obstacle position according to the obstacle position;
[0103] The projection point of the root node relative to the center of gravity is determined as the first attraction point, and the root node is determined as the second attraction point.
[0104] According to a craniotomy robot path planning device provided by the present invention, the third determination module 804 is specifically used for:
[0105] Before the new node becomes the first attraction point, the path length between the new node and the root node is determined by using the first attraction point, and the parent node corresponding to the new node is determined so that the path length from the new node through the parent node to the root node is the shortest;
[0106] After the new node becomes the first attraction point, the second attraction point is used to determine the path length between the new node and the root node, and the parent node corresponding to the new node is determined so that the path length from the new node through the parent node to the root node is the shortest.
[0107] According to a craniotomy robot path planning device provided by the present invention, the third determination module 804 is further used for:
[0108] Determine the curvature of the skull milling path using the path node, the parent node and the parent node corresponding to the parent node;
[0109] When the curvature of the skull milling path is greater than a preset curvature value, the skull milling path is reselected.
[0110] According to a craniotomy robot path planning device provided by the present invention, the third determination module 804 is further used for:
[0111] The search radius is modified according to the number of the random trees, and the parent node corresponding to the new node is modified according to the modified search radius.
[0112] According to a craniotomy robot path planning device provided by the present invention, the third determination module 804 is further used for:
[0113] The skull milling path is fitted by a multi-order Bezier curve, and the skull milling path is smoothed by a mirror extension filtering algorithm.
[0114] Fig. 9 An example of a physical structure diagram of an electronic device is shown in FIG. Fig. 9 As shown, the electronic device may include: a processor 910, a communications interface 920, a memory 930 and a communication bus 940, wherein the processor 910, the communications interface 920 and the memory 930 communicate with each other through the communication bus 940. The processor 910 may call the logic instructions in the memory 930 to execute the craniotomy robot path planning method, which includes:
[0115] Determine the skull milling position and drilling point location based on the skull 3D model and tumor location;
[0116] Determine the drilling point position as the root node, determine the tumor position as the obstacle position, and initialize a random tree;
[0117] Determine an attraction point based on the root node and the obstacle position;
[0118] Based on the attraction points and pre-set search weights, a skull milling path is determined using a rapidly expanding random tree algorithm.
[0119] In addition, the logic instructions in the above-mentioned memory 930 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.
[0120] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to execute the craniotomy robot path planning method provided by the above methods, the method comprising:
[0121] Determine the skull milling position and drilling point location based on the skull 3D model and tumor location;
[0122] Determine the drilling point position as the root node, determine the tumor position as the obstacle position, and initialize a random tree;
[0123] Determine an attraction point based on the root node and the obstacle position;
[0124] Based on the attraction points and pre-set search weights, a skull milling path is determined using a rapidly expanding random tree algorithm.
[0125] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0126] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0127] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A craniotomy robot path planning method, characterized in that: The method comprises: Determine the skull milling position and drilling point location based on the skull 3D model and tumor location; Determine the drilling point position as the root node, determine the tumor position as the obstacle position, and initialize a random tree; Determine an attraction point based on the root node and the obstacle position; Based on the attraction points and pre-set search weights, a skull milling path is determined using a fast expanding random tree algorithm; Wherein, the method of determining the skull milling path by using a fast expanding random tree algorithm based on the attraction points and a preset search weight comprises: For each path node, a random point is selected, and the direction and the target point are determined according to the search weight, the path node and the unit vector of the random point; Determine on the random tree the point closest to the target point in the search radius as a neighboring node, and determine, based on the neighboring nodes, the path node closest to the target point in the direction as a new node; Based on the search weight, the new node and the attraction point, determine the parent node corresponding to the new node so that the path length from the new node through the parent node to the root node is the shortest; The skull milling path is determined using each path node and the parent node corresponding to the path node according to the following formula: ; Wherein, r is the search step length of the path node, is the search weight of the direction of the attraction point generated when the target point searches for the node, is the new node, is the neighboring node, is the unit vector of the path nodes in the path search area, A unit vector pointing in the direction of the target point within the path search area.
2. The craniotomy robot path planning method according to claim 1, characterized in that: The attraction point includes a first attraction point and a second attraction point, and determining the attraction point based on the root node and the obstacle position includes: Determining the center of gravity of the obstacle position according to the obstacle position; The projection point of the root node relative to the center of gravity is determined as the first attraction point, and the root node is determined as the second attraction point.
3. The craniotomy robot path planning method according to claim 2, characterized in that: The determining, based on the search weight, the new node and the attraction point, a parent node corresponding to the new node so that the path length from the new node through the parent node to the root node is the shortest includes: Before the new node becomes the first attraction point, the path length between the new node and the root node is determined by using the first attraction point, and the parent node corresponding to the new node is determined so that the path length from the new node through the parent node to the root node is the shortest; After the new node becomes the first attraction point, the second attraction point is used to determine the path length between the new node and the root node, and the parent node corresponding to the new node is determined so that the path length from the new node through the parent node to the root node is the shortest.
4. The craniotomy robot path planning method according to claim 1, characterized in that: The method further comprises: Determine the curvature of the skull milling path using the path node, the parent node and the parent node corresponding to the parent node; When the curvature of the skull milling path is greater than a preset curvature value, the skull milling path is reselected.
5. The craniotomy robot path planning method according to claim 1, characterized in that: The method further comprises: The search radius is modified according to the number of the random trees, and the parent node corresponding to the new node is modified according to the modified search radius.
6. The craniotomy robot path planning method according to any one of claims 1 to 5, characterized in that: After determining the skull milling path, the method further includes: The skull milling path is fitted by a multi-order Bezier curve, and the skull milling path is smoothed by a mirror extension filtering algorithm.
7. A craniotomy robot path planning device, characterized in that: include: The first determination module is used to determine the skull milling position and the drilling point position according to the skull three-dimensional model and the tumor position; A processing module, used to determine the drilling point position as a root node, determine the tumor position as an obstacle position, and initialize a random tree; A second determination module, configured to determine an attraction point based on the root node and the obstacle position; A third determination module is used to determine the skull milling path using a fast expanding random tree algorithm based on the attraction points and a preset search weight; Wherein, the third determination module is specifically used for: For each path node, a random point is selected, and the direction and the target point are determined according to the search weight, the path node and the unit vector of the random point; Determine on the random tree the point closest to the target point in the search radius as a neighboring node, and determine, based on the neighboring nodes, the path node closest to the target point in the direction as a new node; Based on the search weight, the new node and the attraction point, determine the parent node corresponding to the new node so that the path length from the new node through the parent node to the root node is the shortest; The skull milling path is determined using each path node and the parent node corresponding to the path node according to the following formula: ; Wherein, r is the search step length of the path node, is the search weight of the direction of the attraction point generated when the target point searches for the node, is the new node, is the neighboring node, is the unit vector of the path nodes in the path search area, A unit vector pointing in the direction of the target point within the path search area.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the craniotomy robot path planning method as described in any one of claims 1 to 6 is implemented.
9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the craniotomy robot path planning method as described in any one of claims 1 to 6 is implemented.
Citation Information
Patent Citations
Cooperative interaction control system for craniotomy robot
CN106965175A
Method for determining craniotomy approaching channel through 3D modeling of medical images
CN109893245A
Skull surgery robot system and method
CN112998855A
Cerebral dura mater detection and protection system for skull drilling, electronic equipment and storage medium
CN114869397A
Brain tumor operation planning risk assessment method and device based on deep learning
CN117297771A
