Puncture path planning method and system based on ray projection and collision detection

Abstract

The invention provides a puncture path planning method and system based on ray projection and collision detection, and the method comprises the steps: obtaining a 3D voxel data model of a to-be-punctured object, and carrying out the focus and organ segmentation of the 3D voxel data model according to a pre-trained organ segmentation model; generating uniformly distributed direction vectors from the center of a target focus in the segmented 3D voxel data model to a surrounding space by adopting a Fibonacci spherical sampling algorithm to obtain an initial path set; and performing collision detection on each path in the initial path set, calculating the length of each path passing the collision detection, and determining a candidate puncture path set according to an ascending sequence of the lengths of the paths in combination with a preset candidate threshold value. The technical problems of low reliability and low efficiency of puncture path planning in the prior art are effectively solved.

Description

Technical Field

[0001] This invention relates to the field of medical control, specifically to a puncture path planning method, system, device, and medium based on ray projection and collision detection. Background Technology

[0002] Medical puncture, as a core technology in minimally invasive interventional diagnosis and treatment, has been widely applied in clinical scenarios such as tumor biopsy, lesion ablation, and effusion drainage. It achieves diagnostic sampling or local treatment with minimal trauma, significantly reducing the risks of traditional open surgery and the patient's recovery period. With the deepening of the concept of precision medicine and the development of imaging and robotic technologies, clinicians are placing higher demands on the accuracy, safety, and dynamic adaptability of puncture path planning.

[0003] In clinical practice, the proportion of micronodules ≤10mm in diameter and lesions located adjacent to vital organs is increasing year by year, such as pulmonary ground-glass nodules, hilar tumors, and deep brain hematomas. These necessitate extremely high precision in the puncture path, ensuring the needle accurately reaches the active area of ​​the lesion rather than the necrotic area, while avoiding damage to surrounding blood vessels and nerves. For example, hilar tumors must avoid the portal vein and bile ducts, and deep brain lesions must avoid nerves in functional areas. Traditional pathway planning relies on physician experience combined with two-dimensional imaging judgment, which is prone to misjudgment due to anatomical spatial relationships, leading to target deviation and consequently low biopsy positivity rates or incomplete treatment.

[0004] Furthermore, puncture complications are a core clinical pain point. Complications from traditional manual punctures mainly stem from failure to avoid dangerous structures or intraoperative organ displacement leading to path deviation. Therefore, precise path planning is urgently needed to reduce damage to blood vessels, nerves, and vital organs from the outset, while also lowering the risk of secondary punctures due to inappropriate pathways and alleviating patient suffering.

[0005] The above problems urgently need to be solved. Summary of the Invention

[0006] In view of this, the present invention provides a puncture path planning method, system and storage medium based on ray projection and collision detection to solve the technical problems of low reliability and low efficiency in puncture path planning in the prior art.

[0007] In a first aspect, the present invention provides a puncture path planning method based on ray projection and collision detection, comprising: Obtain a 3D voxel data model of the object to be punctured, and segment the lesion and organ according to the pre-trained organ segmentation model. The Fibonacci spherical sampling algorithm is used to generate uniformly distributed direction vectors from the center of the target lesion in the segmented 3D voxel data model to the surrounding space, thus obtaining an initial path set; Collision detection is performed on each path in the initial path set, and the length of each path that passes the collision detection is calculated. Based on the ascending order of the path lengths and combined with a preset candidate threshold, a candidate puncture path set is determined.

[0008] As an optional implementation, after segmenting the 3D voxel data model into lesions and organs based on the pre-trained organ segmentation model, the method further includes: The arithmetic mean of the coordinates of all vertices in the segmented lesion sub-model is used to obtain the geometric center point of the lesion, which is then used as the center of the target lesion.

[0009] As an optional implementation, it also includes: Assign a corresponding attribute category identifier to each segmented organ sub-model; wherein, the attribute category includes skin surface, venous / arterial vessels, dangerous organs, permissible organs, and bones; Based on the attribute category identifier, construct the corresponding hierarchical bounding body for each forbidden organ sub-model.

[0010] As an optional implementation, after generating uniformly distributed direction vectors and before obtaining the initial path set, the method further includes eliminating invalid direction vectors according to preset direction constraints. Based on the surface puncture entry point where a single direction vector intersects with the skin surface sub-model, an initial path is determined for the internal puncture path from the surface puncture entry point to the center of the target lesion.

[0011] As an optional implementation, the collision detection for each path in the initial path set includes: Determine whether the current path intersects with any of the aforementioned hierarchical bounding volumes; If the current path intersects with any of the aforementioned hierarchical bounding volumes, then the collision detection result for the current path is "fail". If the current path does not intersect with any of the aforementioned hierarchical bounding volumes, then the collision detection result of the current path is passed.

[0012] As an optional implementation, the preset candidate threshold is determined based on the difference between the total length of the puncture needle and the safe length of the extracorporeal extension.

[0013] As an optional implementation, it also includes: In the set of candidate puncture paths, the minimum distance between each path and each layer of enclosing body is calculated, and the path with the largest minimum distance is determined as the target puncture path.

[0014] Secondly, the present invention provides a puncture path planning system based on ray projection and collision detection, comprising: The object segmentation module is used to acquire a 3D voxel data model of the object to be punctured, and to segment the lesion and organ according to the pre-trained organ segmentation model. The ray projection module is used to generate uniformly distributed direction vectors from the center of the target lesion in the segmented 3D voxel data model to the surrounding space using the Fibonacci spherical sampling algorithm, thus obtaining an initial path set. The path filtering module is used to perform collision detection on each path in the initial path set, calculate the length of each path that passes the collision detection, and determine the candidate puncture path set based on the ascending order of the path lengths and a preset candidate threshold.

[0015] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to realize the puncture path planning method based on ray projection and collision detection described in the first aspect.

[0016] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the puncture path planning method based on ray projection and collision detection as described in the first aspect.

[0017] The puncture path planning method, system, and storage medium based on ray projection and collision detection provided by this invention have at least the following beneficial effects: The technical solution provided by this invention, based on a 3D voxel data model and a pre-trained organ segmentation model, can achieve precise segmentation of lesions and surrounding organs. Combined with geometric center point calculation, it locks the core area of ​​the target lesion, effectively adapting to complex scenarios such as small nodules. It avoids target point deviation caused by misjudgment in traditional two-dimensional imaging space, significantly improving the biopsy positivity rate and treatment thoroughness. The Fibonacci spherical sampling algorithm generates multi-directional initial paths, overcoming the limitations of traditional single short paths. By constructing a hierarchical bounding body for prohibited organs, collision detection can accurately avoid dangerous structures such as blood vessels and nerves. Furthermore, the hierarchical bounding body collision detection replaces traditional voxel traversal, significantly reducing computational complexity and improving puncture path planning efficiency. Candidate thresholds, combined with the total length of the puncture needle and the safe extension length outside the body, reserve safety redundancy, prioritizing the optimal path with the greatest distance from dangerous organs. This reduces puncture damage and the risk of secondary punctures from the source, minimizing complications and ensuring both efficiency and safety in the path selection. Attached Figure Description

[0018] To more clearly illustrate and understand the technical solutions in the embodiments of the present invention, the accompanying drawings used in the background technology and embodiment descriptions of the present invention will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the content of the embodiments of the present invention and these drawings without creative effort.

[0019] Figure 1 A schematic diagram of a puncture path planning method based on ray projection and collision detection is shown. Figure 2 A schematic diagram of a puncture path planning system based on ray projection and collision detection is shown. Detailed Implementation

[0020] To make the technical problems solved by the present invention, the technical solutions adopted, and the technical effects achieved clearer, the technical solutions of the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1 See Figure 1 As shown, the present invention provides a puncture path planning method based on ray projection and collision detection, which includes the following steps.

[0022] S1. Obtain the 3D voxel data model of the object to be punctured, and perform lesion and organ segmentation on the 3D voxel data model according to the pre-trained organ segmentation model. S2. Using the Fibonacci spherical sampling algorithm, a uniformly distributed direction vector is generated from the center of the target lesion in the segmented 3D voxel data model to the surrounding space to obtain the initial path set; S3. Perform collision detection on each path in the initial path set and calculate the length of each path that passes the collision detection. Based on the preset candidate threshold and the ascending order of the path lengths, determine the candidate puncture path set.

[0023] In this invention, a 3D voxel data model of the object to be punctured is first obtained, which can be generated by scanning with medical imaging equipment and includes spatial location information of human tissues, organs, and lesions. Then, a pre-trained organ segmentation model, such as a medical image segmentation model based on U-Net, is called to perform pixel-level segmentation on the 3D voxel model, accurately separating the lesion area from the normal organ area, such as the lung, liver, blood vessels, and bones, providing clear spatial boundaries for subsequent path planning.

[0024] The Fibonacci spherical sampling algorithm is an efficient algorithm for generating a uniformly distributed set of points on a sphere. Using the center of the lesion as the sphere's center, it generates uniformly distributed direction vectors radially outwards in three-dimensional space, initially forming a set of directions extending outwards from the lesion center. The number of vectors can be adjusted according to clinical needs, such as 100-500.

[0025] The generated initial path set undergoes two rounds of screening: collision detection eliminates dangerous paths, and a candidate path set is determined by combining path length and a safety threshold. The optimal target path can then be finally determined from this candidate path set.

[0026] This invention, based on a 3D voxel data model and a pre-trained organ segmentation model, enables precise segmentation of lesions and surrounding organs. Combined with geometric center point calculation, it locks onto the core area of ​​the target lesion, effectively adapting to complex scenarios such as small nodules. It avoids target point deviations caused by misjudgments in traditional two-dimensional imaging space, significantly improving biopsy positivity and treatment thoroughness. The Fibonacci spherical sampling algorithm generates multi-directional initial paths, overcoming the limitations of traditional single short paths. By constructing hierarchical bounding bodies for prohibited organs, collision detection accurately avoids dangerous structures such as blood vessels and nerves. Furthermore, hierarchical bounding body collision detection replaces traditional voxel traversal, significantly reducing computational complexity and improving puncture path planning efficiency. Candidate thresholds, combined with the total length of the puncture needle and the safe extension length outside the body, reserve safety redundancy, prioritizing the optimal path with the greatest distance from dangerous organs. This reduces puncture damage and the risk of secondary punctures from the source, minimizing complications and ensuring both efficiency and safety in the path selection.

[0027] In one implementation, after segmenting the 3D voxel data model into lesions and organs according to the pre-trained organ segmentation model, the arithmetic mean of the coordinates of all vertices in the segmented lesion sub-model is calculated to obtain the geometric center point of the lesion, which is used as the target lesion center.

[0028] In this invention, after the lesion and organ are segmented, the coordinates of all vertices of the segmented lesion sub-model are extracted, and the arithmetic mean of the X, Y, and Z three-dimensional coordinates of the vertices is calculated to obtain the geometric center point of the lesion. This point is then used as the target lesion center, i.e., the endpoint reference of the puncture path.

[0029] Furthermore, a density-weighted centroid algorithm can be used instead of the arithmetic mean. Specifically, the spatial coordinates of all voxels within the lesion region are extracted, and the corresponding CT values ​​are obtained. A sigmoid function is applied to map the CT values ​​to weighting coefficients, with higher-density regions, such as enhanced lesions, receiving higher weights. The weighted spatial centroid is then calculated as the final target point coordinates, i.e., the center of the target lesion. Additionally, the principal axis direction of the lesion can be calculated to provide directional guidance for subsequent path planning.

[0030] This invention, by determining the center of the target lesion, accurately reflects the spatial center position of the lesion, ensuring that all subsequent multi-directional paths are based on the center of the target lesion as the endpoint, avoiding the problem of puncture deviation from the lesion due to center positioning deviation, and improving puncture accuracy.

[0031] In one implementation, after generating uniformly distributed direction vectors and before obtaining the initial path set, the method further includes eliminating invalid direction vectors according to preset direction constraints; and determining an initial path for an internal puncture path from the surface puncture entry point to the center of the target lesion based on the surface puncture entry point where a single direction vector intersects with the skin surface sub-model.

[0032] In this invention, after generating a uniform direction vector, known external obstacles are avoided based on preset direction constraints, such as the dorsal direction (the direction the patient points towards the operating table when prone); puncture angles that do not conform to the human anatomical position are avoided, such as avoiding a perpendicular direction to the heart during thoracentesis; and invalid direction vectors that are clearly inconsistent with clinical operation are eliminated, such as directions with too small an angle to the skin surface. This effectively integrates a multi-dimensional safety system encompassing anatomy, physiology, and operation.

[0033] For the remaining valid direction vectors, calculate the intersection point of each vector with the skin surface sub-model, and use this intersection point as the entry point for surface puncture. Then, starting from the entry point for surface puncture and ending at the center of the target lesion, construct a complete internal puncture path. All such paths together form the initial path set.

[0034] By eliminating invalid direction vectors before pre-setting directional constraints, the size of the initial path set is effectively reduced, the load on subsequent collision detection is reduced, and the calculation time is shortened, providing efficiency support for quickly determining the puncture path.

[0035] In one implementation, for each segmented organ sub-model, a unique attribute category identifier is assigned based on its clinical puncture risk level. Specific attribute categories include: Skin surface: puncture entry area, not contraindicated; Veins / arteries: High-risk areas where puncture is prohibited, as it can easily lead to bleeding; Dangerous organs: such as the spinal cord and hilum of the lungs, areas that should not be pierced, as damage to these areas can easily lead to serious complications; Permissible organs: such as muscle, fat, and low-risk puncture-friendly areas; Bone: Hard tissue areas that cannot be punctured, and are prone to needle breakage.

[0036] For organ sub-models with forbidden-penetration attributes, such as blood vessels, dangerous organs, and bones, corresponding hierarchical bounding bodies are constructed, such as axis-aligned bounding boxes (AABB) and directed bounding boxes (OBB), forming simplified collision detection boundaries for forbidden-penetration regions.

[0037] For each path in the initial path set, perform collision detection, traverse the constructed forbidden organ hierarchical bounding volume, and determine whether the current path intersects with any hierarchical bounding volume; If the current path intersects with any level of bounding volume, it means that the path will pass through the forbidden organ. In this case, the collision detection result of the current path is "fail" and it is directly rejected. If the current path does not intersect with any level of bounding volume, it means that the path only passes through the permissible / non-organ region. In this case, the collision detection result of the current path is passed and it is retained.

[0038] This invention, by completely enclosing the space of the restricted organ within a layered enclosure, avoids missed detections due to voxel errors and improves the accuracy of collision detection; it quickly eliminates dangerous paths intersecting with the restricted organ, ensuring that all detected paths are safe paths with no risk of organ damage, thus avoiding puncture complications from the source. This further improves the reliability and efficiency of puncture path planning.

[0039] In one implementation, the preset candidate threshold is determined based on the difference between the total length of the puncture needle and the safe length of extracorporeal extension.

[0040] In this invention, a candidate threshold is set based on the difference between the total length of the puncture needle and the safe length of the external extension, i.e., the candidate threshold = the total length of the puncture needle - the safe length of the external extension. This ensures that the length of the internal segment of the selected candidate path is less than or equal to the threshold, which guarantees that the puncture needle can reach the lesion while maintaining a safe redundant length outside the body surface. This avoids the risk of loss of control due to the lack of an external gripping part of the needle during operation, improves safety, and enhances the reliability of puncture path planning.

[0041] In one implementation, in the set of candidate puncture paths, the minimum distance between each path and each layer of the enclosing body is calculated, and the path with the largest minimum distance is determined as the target puncture path.

[0042] In this invention, by selecting the path with the largest minimum distance from the prohibited organ, the path is further ensured to maintain the farthest safe distance from the dangerous organ while ensuring safety and avoiding collisions. Even if there is an operational deviation during the puncture, it is not easy to touch the prohibited organ, which greatly improves the fault tolerance of the path and reduces the surgical risk.

[0043] Furthermore, before determining the target puncture path, the set of candidate puncture paths can be further narrowed down to a smaller number, such as limiting it to 10. This ensures that the target puncture path can be determined from the top-ranked candidate paths while avoiding excessively deep punctures. Reducing the number of candidate puncture paths decreases computational complexity and further improves the efficiency of determining the target puncture path.

[0044] Furthermore, it should be noted that all anatomical structures and candidate puncture paths can be visualized in a 3D view. Interactive adjustments, such as translation and rotation, are supported for the selected path. After each adjustment, the safety of the path is re-verified, and the final target puncture path is determined. Further details are omitted here.

[0045] Example 2 This embodiment provides a puncture path planning system based on ray projection and collision detection, which is applied to the puncture path planning method based on ray projection and collision detection provided in Embodiment 1 above for illustration.

[0046] Please see Figure 2 As shown, a puncture path planning system based on ray projection and collision detection provided in one embodiment of the present invention may include the following multiple modules.

[0047] The object segmentation module is used to acquire a 3D voxel data model of the object to be punctured, and to segment the lesion and organ according to the pre-trained organ segmentation model. The ray projection module is used to generate uniformly distributed direction vectors from the center of the target lesion in the segmented 3D voxel data model to the surrounding space using the Fibonacci spherical sampling algorithm, thus obtaining an initial path set. The path filtering module is used to perform collision detection on each path in the initial path set, calculate the length of each path that passes the collision detection, and determine the candidate puncture path set based on the ascending order of the path lengths and a preset candidate threshold.

[0048] The puncture path planning system based on ray projection and collision detection provided in this embodiment of the invention can be applied to the puncture path planning method based on ray projection and collision detection provided in Embodiment 1 above. For relevant details, please refer to the above method embodiment. Its implementation principle and technical effect are similar, and will not be repeated here.

[0049] It should be noted that the puncture path planning system based on ray projection and collision detection provided in this embodiment of the invention is only illustrated by the above-described division of functional modules / units. In practical applications, the above functions can be assigned to different functional modules / units as needed, that is, the internal structure of the puncture path planning system based on ray projection and collision detection can be divided into different functional modules / units to complete all or part of the functions described above. Furthermore, the implementation method of the puncture path planning method based on ray projection and collision detection provided in Embodiment 1 above and the implementation system of the puncture path planning system based on ray projection and collision detection provided in Embodiment 2 belong to the same concept. The specific implementation process of the puncture path planning system based on ray projection and collision detection provided in Embodiment 2 is detailed in Embodiment 1 above, and will not be repeated here.

[0050] Example 3 One embodiment of the present invention also provides a computer device, including but not limited to a processor and a memory, wherein the memory stores computer instructions, and the processor executes the computer instructions to implement the method described in the above method embodiment.

[0051] Example 4 One embodiment of the present invention also provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the method described in the above-described method embodiments.

[0052] It should be noted that the aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to general-purpose or special-purpose computers.

[0053] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A puncture path planning method based on ray projection and collision detection, characterized in that, include: Obtain a 3D voxel data model of the object to be punctured, and segment the lesion and organ according to the pre-trained organ segmentation model. The Fibonacci spherical sampling algorithm is used to generate uniformly distributed direction vectors from the center of the target lesion in the segmented 3D voxel data model to the surrounding space, thus obtaining an initial path set; Collision detection is performed on each path in the initial path set, and the length of each path that passes the collision detection is calculated. Based on the ascending order of the path lengths and combined with a preset candidate threshold, a candidate puncture path set is determined.

2. The puncture path planning method based on ray projection and collision detection according to claim 1, characterized in that, After segmenting the 3D voxel data model into lesions and organs based on the pre-trained organ segmentation model, the method further includes: The arithmetic mean of the coordinates of all vertices in the segmented lesion sub-model is used to obtain the geometric center point of the lesion, which is then used as the center of the target lesion.

3. The puncture path planning method based on ray projection and collision detection according to claim 2, characterized in that, Also includes: Assign a corresponding attribute category identifier to each segmented organ sub-model; wherein, the attribute category includes skin surface, venous / arterial vessels, dangerous organs, permissible organs, and bones; Based on the attribute category identifier, construct the corresponding hierarchical bounding body for each forbidden organ sub-model.

4. The puncture path planning method based on ray projection and collision detection according to claim 3, characterized in that: After generating uniformly distributed direction vectors and before obtaining the initial path set, the method further includes eliminating invalid direction vectors according to preset direction constraints. Based on the surface puncture entry point where a single direction vector intersects with the skin surface sub-model, an initial path is determined for the internal puncture path from the surface puncture entry point to the center of the target lesion.

5. The puncture path planning method based on ray projection and collision detection according to claim 4, characterized in that, The collision detection for each path in the initial path set includes: Determine whether the current path intersects with any of the aforementioned hierarchical bounding volumes; If the current path intersects with any of the aforementioned hierarchical bounding volumes, then the collision detection result for the current path is "fail". If the current path does not intersect with any of the aforementioned hierarchical bounding volumes, then the collision detection result of the current path is passed.

6. The puncture path planning method based on ray projection and collision detection according to claim 5, characterized in that, The preset candidate threshold is determined based on the difference between the total length of the puncture needle and the safe length of the external extension.

7. The puncture path planning method based on ray projection and collision detection according to claim 6, characterized in that, Also includes: In the set of candidate puncture paths, calculate the minimum distance between each path and each of the layered enclosing bodies; The path with the largest minimum value is then determined as the target puncture path.

8. A puncture path planning system based on ray projection and collision detection, characterized in that, include: The object segmentation module is used to acquire a 3D voxel data model of the object to be punctured, and to segment the lesion and organ according to the pre-trained organ segmentation model. The ray projection module is used to generate uniformly distributed direction vectors from the center of the target lesion in the segmented 3D voxel data model to the surrounding space using the Fibonacci spherical sampling algorithm, thus obtaining an initial path set. The path filtering module is used to perform collision detection on each path in the initial path set, calculate the length of each path that passes the collision detection, and determine the candidate puncture path set based on the ascending order of the path lengths and a preset candidate threshold.

9. A computer device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes these computer instructions to implement the puncture path planning method based on ray projection and collision detection as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the puncture path planning method based on ray projection and collision detection as described in any one of claims 1-7.

Images