Method for path planning for pulmonary lesion navigation and related devices

Abstract

This application discloses a path planning method and related equipment for lung lesion navigation, which improves the accuracy of path planning for lung lesion navigation, thereby improving the ablation effect and safety. The method includes: inputting lung CT images into an ablation path recommendation model; the ablation path recommendation model performing the following processing: constructing a three-dimensional lung model based on the lung CT images; determining a target path planning scheme in the bronchial tree based on the spatial relationship between the bronchial tree, lesion, blood vessels, and pleura; simulating the ablation device entering the bronchus via the puncture point; puncturing the lesion at the puncture angle; determining the predicted ablation area under the target ablation parameter combination settings; determining candidate path schemes based on the comparison between the predicted ablation area and the preset standard ablation area; and comprehensively scoring and optimizing the candidate path schemes using a multi-objective optimization algorithm to obtain the target path scheme.

Description

Technical Field

[0001] This application relates to the field of path planning for lung lesion navigation, and more specifically, to a path planning method, a path planning system, a path planning device, a computer-readable storage medium, and a computer program product containing instructions for lung lesion navigation. Background Technology

[0002] With the widespread adoption of lung CT screening technology, the detection rate of early pulmonary nodules has significantly improved, leading to an increasing demand for path planning for pulmonary lesion navigation.

[0003] Existing methods for pulmonary lesion navigation mainly rely on technologies such as electromagnetic navigation bronchoscopy to provide lesion localization and bronchial pathway guidance. However, the puncture point location, puncture angle, and ablation parameters still depend on the surgeon's personal experience to determine.

[0004] This method is limited to providing general guidance for bronchial pathways, and the accuracy of path planning for lung lesion navigation is low. It is prone to positioning deviations, incomplete ablation, or damage to normal tissue due to differences in surgeon experience, making it difficult to guarantee the effectiveness and safety of lung lesion ablation. Summary of the Invention

[0005] This application provides a path planning method, a path planning system, a path planning device, a computer-readable storage medium, and a computer program product containing instructions for pulmonary lesion navigation, which are used to improve the accuracy of path planning for pulmonary lesion navigation, thereby improving the lesion ablation effect and safety.

[0006] In a first aspect, embodiments of this application provide a path planning method for navigation of lung lesions, including:

[0007] Acquire lung CT images and input them into a pre-trained ablation path recommendation model. The ablation path recommendation model performs the following processing:

[0008] A three-dimensional lung model is constructed based on the lung CT images. The three-dimensional lung model includes the bronchial tree, lesions, blood vessels, and pleura, and the spatial positional relationship of the lesions relative to the bronchial tree, blood vessels, and pleura is determined.

[0009] Based on the spatial relationship, a target path planning scheme is determined in the bronchial tree that satisfies the requirement that the puncture distance of the ablation device is less than a threshold and is directed toward the lesion, while avoiding the blood vessels and the pleura. The target path planning scheme includes the bronchial entry pathway, the puncture point location, and the puncture angle. The puncture angle represents the entry angle of the ablation device at the end of the bronchial entry pathway.

[0010] The ablation device is simulated to enter the bronchus through the bronchus access pathway corresponding to the target path planning scheme and navigate to the corresponding puncture point. It is then punctured into the lesion at the corresponding puncture angle. Under the target ablation parameter combination settings, the predicted ablation area of ​​the target path planning scheme is determined. Based on the comparison between the predicted ablation area of ​​the target path planning scheme and the preset standard ablation area, a candidate path scheme is determined. The candidate path scheme includes the path planning scheme and the corresponding ablation parameter combination.

[0011] The candidate path schemes are comprehensively scored and optimized based on a multi-objective optimization algorithm from multiple dimensions, including path length, accessibility from the entry angle, safety, and ablation thoroughness, to obtain the target path scheme output by the ablation path scheme recommendation model. The target path scheme is used to ablate the lesion.

[0012] Secondly, embodiments of this application provide a path planning system for navigating lung lesions, the path planning system comprising:

[0013] The acquisition module is used to acquire lung CT images and input the lung CT images into a pre-trained ablation path recommendation model. The ablation path recommendation model performs the following processing:

[0014] A construction module is used to construct a three-dimensional lung model based on the lung CT image. The three-dimensional lung model includes a bronchial tree, lesions, blood vessels, and pleura, and determines the spatial positional relationship of the lesions relative to the bronchial tree, blood vessels, and pleura.

[0015] The determination module is used to determine, based on the spatial positional relationship, a target path planning scheme in the bronchial tree that satisfies the requirement that the puncture distance of the ablation device is less than a threshold and is directed toward the lesion, while avoiding the blood vessels and the pleura. The target path planning scheme includes the bronchial entry pathway, the puncture point location, and the puncture angle. The puncture angle characterizes the entry angle of the ablation device at the end of the bronchial entry pathway.

[0016] The determining module is further configured to simulate the ablation device being guided through the bronchial entry pathway corresponding to the target path planning scheme to the corresponding puncture point, and to puncture the lesion at the corresponding puncture angle. Under the target ablation parameter combination setting, the predicted ablation area of ​​the target path planning scheme is determined, and a candidate path scheme is determined based on the comparison result between the predicted ablation area of ​​the target path planning scheme and the preset standard ablation area. The candidate path scheme includes the path planning scheme and the corresponding ablation parameter combination.

[0017] The output module is used to comprehensively score and optimize the candidate path schemes based on a multi-objective optimization algorithm from multiple dimensions such as path length, accessibility from the entry angle, safety, and ablation thoroughness, so as to obtain the target path scheme output by the ablation path scheme recommendation model.

[0018] Thirdly, embodiments of this application provide a path planning device for navigating lung lesions, comprising:

[0019] Central processing unit, memory, input / output interfaces, wired or wireless network interfaces, and power supply;

[0020] The memory is either a short-term storage memory or a persistent storage memory;

[0021] The central processing unit is configured to communicate with the memory and execute instructions in the memory to perform the aforementioned path planning for lung lesion navigation.

[0022] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a computer, cause the computer to perform the aforementioned path planning for navigation of lung lesions.

[0023] Fifthly, embodiments of this application provide a computer program product containing instructions that, when run on a computer, cause the computer to perform the aforementioned path planning for navigation of lung lesions.

[0024] As can be seen from the above technical solutions, the embodiments of this application have the following advantages: This application intelligently plans the puncture path (bronchial access route) before the operation, accurately locates the puncture point and puncture angle, and simulates the ablation area coverage effect corresponding to the target ablation parameter combination. Based on the multi-objective optimization algorithm, it comprehensively evaluates the path length, accessibility of the entry angle, safety and ablation thoroughness, and determines the target path scheme. The path planning accuracy of lung lesion navigation is high, which can avoid the problems of positioning deviation, incomplete ablation or damage to normal tissue caused by the strong reliance on the operator's experience in the prior art, and significantly improves the effect and safety of lung lesion ablation.

[0025] Accordingly, the path planning system for pulmonary lesion navigation, the path planning device for pulmonary lesion navigation, the computer-readable storage medium, and the computer program product containing instructions provided in this application also have the above-mentioned technical effects. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating a path planning method for navigating lung lesions disclosed in an embodiment of this application.

[0027] Figure 2This is a schematic diagram of the structure of a path planning system for navigating lung lesions, as disclosed in an embodiment of this application.

[0028] Figure 3 This is a schematic diagram of the structure of a path planning device for navigating lung lesions, as disclosed in an embodiment of this application. Detailed Implementation

[0029] This application provides a path planning method, a path planning system, a path planning device, a computer-readable storage medium, and a computer program product containing instructions for pulmonary lesion navigation, which are used to improve the accuracy of path planning for pulmonary lesion navigation, thereby improving the lesion ablation effect and safety.

[0030] Please see Figure 1 , Figure 1 This is a flowchart illustrating a path planning method for pulmonary lesion navigation disclosed in an embodiment of this application. The method includes:

[0031] 101. Obtain lung CT images and input them into a pre-trained ablation path recommendation model. The ablation path recommendation model performs the following processing: construct a three-dimensional lung model based on the lung CT images. The three-dimensional lung model includes the bronchial tree, lesions, blood vessels, and pleura, and determine the spatial positional relationship of the lesions relative to the bronchial tree, blood vessels, and pleura.

[0032] In one alternative implementation, lung CT images include, but are not limited to, input data from different imaging methods such as enhanced CT images (for clearer identification of blood vessels) or bronchoscopic ultrasound images (for assisting in the localization of deep lesions), which are not limited here.

[0033] 102. Based on spatial location relationships, determine the target path planning scheme in the bronchial tree that satisfies the requirement that the puncture distance of the ablation device is less than the threshold and is directed toward the lesion, while avoiding blood vessels and pleura. The target path planning scheme includes the bronchial entry pathway, puncture point location and puncture angle. The puncture angle characterizes the entry angle of the ablation device at the end of the bronchial entry pathway.

[0034] In one alternative implementation, the ablation device may include, but is not limited to, radiofrequency ablation, microwave ablation, and cryoablation devices, with different ablation devices having different effective radii and energy distributions.

[0035] 103. The simulated ablation device is guided through the bronchus access route corresponding to the target path planning scheme to the corresponding puncture point. It is then punctured into the lesion at the corresponding puncture angle. Under the target ablation parameter combination settings, the predicted ablation area of ​​the target path planning scheme is determined. Based on the comparison between the predicted ablation area of ​​the target path planning scheme and the preset standard ablation area, candidate path schemes are determined. The candidate path schemes include the path planning scheme and the corresponding ablation parameter combination.

[0036] In one alternative implementation, the ablation zone extent can be predicted based on a parametric model of the corresponding ablation energy type (such as a radio frequency / microwave thermal conduction model or a frozen ball diffusion model).

[0037] 104. Based on a multi-objective optimization algorithm, candidate path schemes are comprehensively scored and optimized from multiple dimensions such as path length, accessibility from the entry angle, safety, and ablation thoroughness to obtain the target path scheme output by the ablation path scheme recommendation model. The target path scheme is used to ablate the lesion.

[0038] In one alternative implementation, a multi-objective optimization algorithm (such as weighted scoring or Pareto optimization) can be used to select the optimal solution from multiple candidate solutions that takes into account the shortest path, angular accessibility, safe obstacle avoidance (blood vessels / pleura), and thorough ablation (complete coverage of the lesion). The final output is the target path plan, which represents which bronchus to enter from, where to puncture, at what angle to insert the needle, and what combination of ablation parameters (such as how much power to use and how long to burn). The operator can directly perform ablation according to this.

[0039] It is worth mentioning that, through extensive research, the applicant discovered the following unique path planning complexities in lung lesion navigation: First, the bronchial tree of the lung has a branch structure of more than 20 levels, and reaching the lung lesion requires continuous selection of branch paths; second, blood vessels (pulmonary arteries and / or pulmonary veins) run closely alongside the bronchial tree, at a distance of only 1 to 3 millimeters, making it very easy to damage blood vessels and cause massive bleeding during puncture; third, lung lesions are often closely attached to the pleura (with almost no gap), and the pleural cavity is a negative pressure environment, so even slight miscontrol of the puncture angle can easily lead to pneumothorax and lung collapse. In contrast, the path planning for lesions in digestive organs such as the stomach is not as complex: digestive organs such as the stomach are sac-like or tubular structures without tree-like branches; blood vessels are located outside the wall at a certain distance (4-15 mm from the cavity, for example, 4-6 mm from the stomach cavity in the stomach and 5-15 mm from the intestinal wall in the intestine), and do not run alongside or intertwine with the cavity; there is a certain distance (e.g., 4-6 mm) between the peritoneum and digestive organs such as the stomach, and the puncture risk is mainly peritonitis caused by leakage of contents, rather than immediate fatal lung collapse.

[0040] Addressing the difference in path planning complexity between lung lesions and lesions in digestive organs such as the stomach, the applicant found that existing path planning methods for navigation of lesions in digestive organs (considering only extramural vascular obstacle avoidance) are unsuitable for lung scenarios with such high path planning complexity. Therefore, this application proposes a solution. This application constructs a three-dimensional lung model including the bronchial tree, lesion, blood vessels, and pleura. This model accurately determines the spatial relationship of the lesion relative to the bronchial tree, blood vessels, and pleura. Based on this, a target path planning scheme that avoids vascular and pleural constraints is determined, including the bronchial entry route, puncture point location, and puncture angle. Through simulation prediction of the ablation area and comprehensive scoring using a multi-objective optimization algorithm, a refined plan is achieved that simultaneously satisfies the shortest path, vascular obstacle avoidance, pleural angle control, and thorough ablation. This effectively solves the technical challenge of balancing complex path selection, close vascular obstacle avoidance, and close pleural protection in trans-airway lung ablation, significantly improving the ablation effect and safety of lung lesions. Therefore, the path planning method for lung lesion navigation scenarios in this application has outstanding inventiveness.

[0041] In this way, by intelligently planning the puncture path (bronchial access route) before the procedure, accurately locating the puncture point and puncture angle, and simulating the ablation area coverage effect corresponding to the target ablation parameter combination, a multi-objective optimization algorithm is used to comprehensively evaluate the path length, access angle accessibility, safety, and ablation thoroughness to determine the target path plan. The path planning for lung lesion navigation has high accuracy, avoiding the problems of positioning deviation, incomplete ablation, or damage to normal tissue caused by the strong reliance on operator experience in existing technologies, and significantly improving the accuracy and safety of lung lesion ablation. Secondly, this application can incorporate the characteristics of the ablation device (effective radius, energy distribution pattern) into the path planning. For example, for an ablation needle with omnidirectional energy release, it can be planned to be located at the center of the lesion; for directional energy release, the optimal orientation can be calculated to concentrate energy on the lesion and avoid normal tissue, which can significantly improve the ablation effect and safety of the lesion.

[0042] In one optional implementation, a target path planning scheme is determined in the bronchial tree based on spatial location relationships, which satisfies the requirement that the puncture distance of the ablation device is less than a threshold and is directed towards the lesion, while avoiding blood vessels and the pleura. The target path planning scheme includes the bronchial entry route, the puncture point location, and the puncture angle. This includes: using a path search algorithm to determine candidate bronchial entry routes to the lesion in the bronchial tree based on spatial location relationships; determining candidate puncture point locations and candidate puncture angles on the walls of the candidate bronchial entry routes to obtain candidate path planning schemes; and determining the target path planning scheme from the candidate path planning schemes, which satisfies the requirement that the puncture distance of the ablation device is less than a threshold and is directed towards the lesion, while avoiding blood vessels and the pleura. The target path planning scheme includes the bronchial entry route, the puncture point location, and the puncture angle.

[0043] Specifically, graph-based shortest path algorithms, Rapid Random Tree Search (RRT), or AI-based path search algorithms can be used to search for candidate bronchial entry pathways to the lesion within the bronchial tree based on spatial relationships. Candidate puncture point locations and angles are determined on the walls of these pathways, minimizing the puncture distance and aiming towards the lesion (e.g., the lesion center), while avoiding constraints related to blood vessels and the pleura. This yields candidate path planning schemes. The target path planning scheme can be determined from these candidate schemes using a genetic algorithm or particle swarm optimization algorithm for global optimization and selection. It is understood that the specific choices of the aforementioned path search and optimization algorithms are equivalent modifications and can be replaced according to actual needs; no specific limitations are imposed here.

[0044] In this way, by automatically searching for bronchial pathways through path search algorithms and screening candidate locations on the bronchial wall under multiple constraints (while simultaneously satisfying that the puncture distance of the ablation device is less than a threshold and is directed towards the lung lesion, as well as avoiding blood vessels and pleura), a precise and safe path planning scheme can be determined, thereby improving the accuracy of puncture positioning and the safety of the operation.

[0045] In one optional implementation, the candidate path scheme is determined based on the comparison between the predicted ablation area of ​​the target path planning scheme and the preset standard ablation area. This includes: comparing the predicted ablation area of ​​the target path planning scheme with the preset standard ablation area; if the predicted ablation area of ​​the target path planning scheme covers the preset standard ablation area, then a candidate path scheme is determined based on the target path planning scheme and the combination of target ablation parameters; if the predicted ablation area of ​​the target path planning scheme does not cover the preset standard ablation area, then a new target path planning scheme is determined in the bronchial tree based on the spatial location relationship, which satisfies the shortest puncture distance of the ablation device, faces the lesion, and avoids blood vessels and pleura; and a candidate path scheme is determined based on the newly determined target path planning scheme and the adjusted combination of target ablation parameters.

[0046] Specifically, the target ablation parameter combination includes, but is not limited to, ablation power and time. The predicted ablation area can be compared with the preset standard ablation area. If the predicted area completely covers the preset area, the current path scheme and the corresponding ablation parameter combination (such as ablation power and time) are determined as candidate schemes. If it does not completely cover the area, a new path is re-determined based on the spatial relationship to meet the requirements of shortest puncture distance, oriented towards the lesion, and avoiding blood vessels and pleura. Candidate schemes are then determined based on the adjusted ablation parameter combination.

[0047] In this way, when the predicted area does not completely cover the lesion, the path is automatically replanned and the ablation parameters are adjusted to ensure that the ablation range completely covers the lesion, thereby significantly improving the ablation effect and safety.

[0048] In one optional implementation, a target path planning scheme is redefined in the bronchial tree based on spatial location relationships, which satisfies the requirement of minimizing the puncture distance of the ablation device, directing it towards the lesion, and avoiding blood vessels and pleura. Candidate path schemes are then determined based on the redefined target path planning scheme and the adjusted target ablation parameter combination. This includes: redefined multiple sub-path planning schemes in the bronchial tree based on spatial location relationships. Each sub-path planning scheme includes a bronchial entry pathway, a puncture point location, and a puncture angle. These multiple sub-path planning schemes collectively satisfy the requirement of minimizing the puncture distance of the ablation device, directing it towards the lesion, and avoiding blood vessels and pleura. These multiple sub-path planning schemes constitute the redefined target path planning scheme. The ablation device is simulated to navigate to the corresponding puncture point location via the bronchial entry pathway corresponding to each sub-path planning scheme, and punctures into the lesion at the corresponding puncture angle. Under the adjusted target ablation parameter combination settings, the predicted ablation area of ​​each sub-path planning scheme is determined. The predicted comprehensive ablation area is obtained by combining the predicted ablation areas of multiple sub-path planning schemes. Candidate path schemes are determined based on the comparison between the predicted comprehensive ablation area and the preset standard ablation area.

[0049] Specifically, when the predicted ablation area of ​​a single path does not completely cover the lesion, the system iteratively optimizes and adjusts ablation parameters (such as increasing power, extending time, or changing probe type) to expand the single-point ablation range. If complete coverage is still not possible after adjusting ablation parameters, a multi-point ablation strategy is adopted, planning multiple sub-path planning schemes. The predicted ablation areas of multiple sub-path planning schemes are combined to obtain the predicted comprehensive ablation area. Candidate path schemes are determined based on the comparison between the predicted comprehensive ablation area and the preset standard ablation area. The optimal combination of ablation parameters is determined through repeated simulations to ensure that the final ablation area completely covers the lesion, thereby eliminating the risk of tumor residue.

[0050] In this way, by adopting a multi-point ablation strategy, multiple sub-path planning schemes are automatically planned to achieve the best overall effect, which is to minimize the puncture distance of the ablation device, direct it toward the lesion, and avoid blood vessels and pleura. Based on the comparison between the predicted comprehensive ablation area corresponding to the multiple sub-path planning schemes and the preset standard ablation area, candidate path schemes are determined. This solves the problem of insufficient ablation range of single path, ensures complete coverage of larger lesions, and significantly improves the thoroughness of ablation treatment.

[0051] In one optional implementation, a multi-objective optimization algorithm is used to comprehensively score and optimize candidate path schemes from multiple dimensions, including path length, accessibility from the entry angle, safety, and ablation thoroughness, to obtain the target path scheme output by the ablation path scheme recommendation model. This includes: establishing a quantitative scoring index system that includes path length, accessibility from the entry angle, safety, and ablation thoroughness; calculating the comprehensive score of each candidate path scheme using a weighted scoring function based on the quantitative scoring index system; and determining the candidate path scheme whose comprehensive score reaches a preset scoring threshold as the target path scheme.

[0052] Specifically, safety is quantified based on the safe distance between the ablation path and blood vessels and pleura, while ablation thoroughness is quantified based on the coverage of the ablation area on the lesion and the safety boundary. The system can establish a quantitative scoring index system that includes path length (the shorter the puncture distance, the better, to reduce the distance traveled in the lung parenchyma), access angle accessibility (it must be within the bending capability of the bronchoscope and ablation catheter to avoid sharp bends with too small a radius that would prevent the instrument from reaching the target), safety (the path avoids blood vessels and ensures a safe distance from the pleura to reduce the probability of complications), and ablation thoroughness (the higher the coverage of the ablation area, the better, to cover the lesion as much as possible and have a safe margin). Based on this quantitative scoring index system, a multi-objective optimization algorithm is used to calculate the comprehensive score of each candidate path scheme using a quantitative weighted scoring function or Pareto optimization algorithm. The candidate path scheme that reaches the preset scoring threshold is determined as the target path scheme. The target path scheme includes the optimal bronchial entry route, puncture point location, puncture angle, and target ablation parameter combination.

[0053] In this way, by constructing a multi-dimensional quantitative scoring system and using a weighted scoring function for comprehensive optimization, a global balance was achieved in path length, accessibility of entry angle, safety, and ablation thoroughness, avoiding the limitations of single-index optimization and significantly improving the ablation effect and safety of lesions.

[0054] In one optional implementation, after obtaining the target path scheme output by the ablation path scheme recommendation model, the method further includes: obtaining the ablation effect of ablation on the lesion based on the target path scheme, and performing feedback training on the ablation path scheme recommendation model based on the ablation effect.

[0055] Specifically, data on the actual ablation effect after lesion ablation based on the target path plan can be obtained (including intraoperative imaging verification results or postoperative assessment of the ablation range and residual lesion). The actual ablation effect is compared with the expected ablation effect, and the relevant model parameters and algorithms (such as path planning parameters, ablation area prediction algorithms, or scoring weights) of the ablation path plan recommendation model are trained and optimized based on the differences. This can improve the inference accuracy of the ablation path plan recommendation model.

[0056] In one optional implementation, before inputting the lung CT image into the pre-trained ablation path recommendation model, the method further includes: acquiring lung CT image samples labeled with target path schemes; inputting the lung CT image samples into the ablation path recommendation model to obtain the predicted target path scheme output by the ablation path recommendation model; and obtaining the trained ablation path recommendation model when the loss between the predicted target path scheme and the labeled target path scheme reaches a preset convergence condition.

[0057] Specifically, the ablation path recommendation model can be trained using a supervised learning mechanism. First, lung CT image samples labeled with target path schemes are obtained. Then, the lung CT image samples are input into the ablation path recommendation model to obtain the output predicted target path scheme. When the loss between the predicted target path scheme and the labeled target path scheme reaches the preset convergence condition, the trained ablation path recommendation model is obtained.

[0058] In this way, the model is trained based on a large amount of expert experience data, which can learn the complex mapping relationship from CT images to the optimal ablation path, thus possessing automated and high-precision path planning capabilities. This not only reduces the subjective differences and time costs of manual planning, but also significantly improves the ablation effect and safety of lesions.

[0059] For further details, please refer to Figure 2 One embodiment of the path planning system for pulmonary lesion navigation in this application includes:

[0060] The acquisition module is used to acquire lung CT images and input the lung CT images into a pre-trained ablation path recommendation model. The ablation path recommendation model performs the following processing:

[0061] A construction module is used to construct a three-dimensional lung model based on the lung CT image. The three-dimensional lung model includes a bronchial tree, lesions, blood vessels, and pleura, and determines the spatial positional relationship of the lesions relative to the bronchial tree, blood vessels, and pleura.

[0062] The determination module is used to determine, based on the spatial positional relationship, a target path planning scheme in the bronchial tree that satisfies the requirement that the puncture distance of the ablation device is less than a threshold and is directed toward the lesion, while avoiding the blood vessels and the pleura. The target path planning scheme includes the bronchial entry pathway, the puncture point location, and the puncture angle. The puncture angle characterizes the entry angle of the ablation device at the end of the bronchial entry pathway.

[0063] The determining module is further configured to simulate the ablation device being guided through the bronchial entry pathway corresponding to the target path planning scheme to the corresponding puncture point, and to puncture the lesion at the corresponding puncture angle. Under the target ablation parameter combination setting, the predicted ablation area of ​​the target path planning scheme is determined, and a candidate path scheme is determined based on the comparison result between the predicted ablation area of ​​the target path planning scheme and the preset standard ablation area. The candidate path scheme includes the path planning scheme and the corresponding ablation parameter combination.

[0064] The output module is used to comprehensively score and optimize the candidate path schemes based on a multi-objective optimization algorithm from multiple dimensions such as path length, accessibility from the entry angle, safety, and ablation thoroughness, so as to obtain the target path scheme output by the ablation path scheme recommendation model.

[0065] In one alternative implementation, the determining module may be used for:

[0066] The path search algorithm is used to determine the candidate bronchial entry pathways to the lesion in the bronchial tree based on the spatial location relationship;

[0067] The candidate puncture point location and candidate puncture angle are determined on the wall of the candidate bronchial entry pathway to obtain a candidate path planning scheme;

[0068] Among the candidate path planning schemes, a target path planning scheme is determined that satisfies the requirement that the puncture distance of the ablation device is less than a threshold and is directed toward the lesion, while avoiding the blood vessels and the pleura. The target path planning scheme includes the bronchial entry route, the puncture point location, and the puncture angle.

[0069] In one alternative implementation, the determining module may be used for:

[0070] The predicted ablation region of the target path planning scheme is compared with the preset standard ablation region;

[0071] If the predicted ablation region of the target path planning scheme covers the preset standard ablation region, then the candidate path scheme is determined based on the target path planning scheme and the target ablation parameter combination;

[0072] If the predicted ablation area of ​​the target path planning scheme does not cover the preset standard ablation area, a new target path planning scheme is determined in the bronchial tree based on the spatial position relationship, which satisfies the shortest puncture distance of the ablation device, faces the lesion, and avoids the blood vessels and the pleura. The candidate path scheme is determined based on the newly determined target path planning scheme and the adjusted target ablation parameter combination.

[0073] In one alternative implementation, the determining module may be used for:

[0074] Based on the spatial relationship, multiple sub-path planning schemes are redefined in the bronchial tree. Each of the multiple sub-path planning schemes includes the bronchial entry route, puncture point location and puncture angle. The multiple sub-path planning schemes comprehensively satisfy the requirements of the shortest puncture distance of the ablation device and its orientation toward the lesion, as well as avoiding the blood vessels and the pleura. The multiple sub-path planning schemes are the redefined target path planning schemes.

[0075] The simulated ablation device is guided through the bronchial access pathway corresponding to each sub-path planning scheme to the corresponding puncture point location, and punctures into the lesion at the corresponding puncture angle. Under the adjusted target ablation parameter combination settings, the predicted ablation area of ​​each sub-path planning scheme is determined, and the predicted comprehensive ablation area of ​​multiple sub-path planning schemes is obtained by combining the predicted comprehensive ablation area. Based on the comparison results between the predicted comprehensive ablation area and the preset standard ablation area, candidate path schemes are determined.

[0076] In one alternative implementation, the output module can be used for:

[0077] Establish a quantitative scoring index system that includes path length, accessibility of entry angle, safety, and ablation thoroughness;

[0078] Based on the quantitative scoring index system, a weighted scoring function is used to calculate the comprehensive score of each candidate path scheme;

[0079] Candidate path schemes whose comprehensive scores reach a preset score threshold are determined as the target path schemes.

[0080] In one optional implementation, the path planning system for pulmonary lesion navigation further includes:

[0081] The training module is used to obtain the ablation effect of the lesion ablation based on the target path scheme, and to perform feedback training on the ablation path scheme recommendation model based on the ablation effect.

[0082] In one alternative implementation, the training module can also be used for:

[0083] A lung CT image sample is acquired, and the lung CT image sample is labeled with a target path scheme. The lung CT image sample is input into an ablation path scheme recommendation model to obtain a predicted target path scheme output by the ablation path scheme recommendation model. When the loss between the predicted target path scheme and the labeled target path scheme reaches a preset convergence condition, the trained ablation path scheme recommendation model is obtained.

[0084] For further details, please refer to Figure 3 One embodiment of the path planning device for pulmonary lesion navigation in this application includes:

[0085] Central processing unit 301, memory 305, input / output interface 304, wired or wireless network interface 303, and power supply 302;

[0086] Memory 305 is either a short-term storage memory or a persistent storage memory;

[0087] The central processing unit 301 is configured to communicate with the memory 305 and execute instructions stored in the memory 305 to perform the aforementioned operations. Figure 1 The method in the illustrated embodiment.

[0088] Furthermore, embodiments of this application also provide a computer-readable storage medium, which includes instructions that, when executed on a computer, cause the computer to perform the aforementioned... Figure 1 The method in the illustrated embodiment.

[0089] Furthermore, embodiments of this application also provide a computer program product containing instructions, which, when run on a computer, causes the computer to perform the aforementioned... Figure 1 The method in the illustrated embodiment.

[0090] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0091] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0092] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0093] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0094] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0095] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A path planning method for navigation of lung lesions, characterized in that, include: Acquire lung CT images and input them into a pre-trained ablation path recommendation model. The ablation path recommendation model performs the following processing: A three-dimensional lung model is constructed based on the lung CT images. The three-dimensional lung model includes the bronchial tree, lesions, blood vessels, and pleura, and the spatial positional relationship of the lesions relative to the bronchial tree, blood vessels, and pleura is determined. Based on the spatial relationship, a target path planning scheme is determined in the bronchial tree that satisfies the requirement that the puncture distance of the ablation device is less than a threshold and is directed toward the lesion, while avoiding the blood vessels and the pleura. The target path planning scheme includes the bronchial entry pathway, the puncture point location, and the puncture angle. The puncture angle represents the entry angle of the ablation device at the end of the bronchial entry pathway. The ablation device is simulated to enter the bronchus through the bronchus access pathway corresponding to the target path planning scheme and navigate to the corresponding puncture point. It is then punctured into the lesion at the corresponding puncture angle. Under the target ablation parameter combination settings, the predicted ablation area of ​​the target path planning scheme is determined. Based on the comparison between the predicted ablation area of ​​the target path planning scheme and the preset standard ablation area, a candidate path scheme is determined. The candidate path scheme includes the path planning scheme and the corresponding ablation parameter combination. Based on a multi-objective optimization algorithm, the candidate path schemes are comprehensively scored and optimized from multiple dimensions such as path length, accessibility from the entry angle, security, and ablation thoroughness to obtain the target path scheme output by the ablation path scheme recommendation model.

2. The method of claim 1, wherein, The target path planning scheme, determined based on the spatial relationship in the bronchial tree, satisfies the requirement that the ablation device puncture distance is less than a threshold and faces the lesion, while avoiding the blood vessels and the pleura. The target path planning scheme includes the bronchial entry route, puncture point location, and puncture angle, including: The path search algorithm is used to determine the candidate bronchial entry pathways to the lesion in the bronchial tree based on the spatial location relationship; The candidate puncture point location and candidate puncture angle are determined on the wall of the candidate bronchial entry pathway to obtain a candidate path planning scheme; Among the candidate path planning schemes, a target path planning scheme is determined that satisfies the requirement that the puncture distance of the ablation device is less than a threshold and is directed toward the lesion, while avoiding the blood vessels and the pleura. The target path planning scheme includes the bronchial entry route, the puncture point location, and the puncture angle.

3. The method of claim 1, wherein, The process of determining candidate ablation schemes by comparing the predicted ablation region based on the target path planning scheme with the preset standard ablation region includes: The predicted ablation region of the target path planning scheme is compared with the preset standard ablation region; If the predicted ablation region of the target path planning scheme covers the preset standard ablation region, then the candidate path scheme is determined based on the target path planning scheme and the target ablation parameter combination; If the predicted ablation area of ​​the target path planning scheme does not cover the preset standard ablation area, a new target path planning scheme is determined in the bronchial tree based on the spatial position relationship, which satisfies the shortest puncture distance of the ablation device, faces the lesion, and avoids the blood vessels and the pleura. The candidate path scheme is determined based on the newly determined target path planning scheme and the adjusted target ablation parameter combination.

4. The method of claim 3, wherein, The process involves re-determining a target path planning scheme within the bronchial tree based on the spatial location relationship, which minimizes the puncture distance of the ablation device, directs it towards the lesion, and avoids the blood vessels and pleura. The process also includes determining the candidate path scheme based on the re-determined target path planning scheme and the adjusted combination of target ablation parameters, including: Based on the spatial relationship, multiple sub-path planning schemes are redefined in the bronchial tree. Each of the multiple sub-path planning schemes includes the bronchial entry route, puncture point location and puncture angle. The multiple sub-path planning schemes comprehensively satisfy the requirements of the shortest puncture distance of the ablation device and its orientation toward the lesion, as well as avoiding the blood vessels and the pleura. The multiple sub-path planning schemes are the redefined target path planning schemes. The simulated ablation device is guided through the bronchial access pathway corresponding to each sub-path planning scheme to the corresponding puncture point location, and punctures into the lesion at the corresponding puncture angle. Under the adjusted target ablation parameter combination settings, the predicted ablation area of ​​each sub-path planning scheme is determined, and the predicted comprehensive ablation area of ​​multiple sub-path planning schemes is obtained by combining the predicted comprehensive ablation area. Based on the comparison results between the predicted comprehensive ablation area and the preset standard ablation area, candidate path schemes are determined.

5. The method of claim 1, wherein, The multi-objective optimization algorithm comprehensively scores and optimizes the candidate path schemes from multiple dimensions, including path length, accessibility from the entry angle, security, and ablation thoroughness, to obtain the target path schemes output by the ablation path scheme recommendation model, including: Establish a quantitative scoring index system that includes path length, accessibility of entry angle, safety, and ablation thoroughness; Based on the quantitative scoring index system, a weighted scoring function is used to calculate the comprehensive score of each candidate path scheme; Candidate path schemes whose comprehensive scores reach a preset score threshold are determined as the target path schemes.

6. The method of claim 1, wherein, After obtaining the target path scheme output by the ablation path scheme recommendation model, the method further includes: Obtain the ablation effect of the lesion ablation based on the target path scheme; The ablation path recommendation model is trained based on the ablation effect.

7. The method of claim 1, wherein, Before inputting the lung CT image into the pre-trained ablation path recommendation model, the method further includes: Obtain lung CT image samples, wherein the lung CT image samples are labeled with target path schemes; The lung CT image samples are input into the ablation path recommendation model to obtain the predicted target path scheme output by the ablation path recommendation model. When the loss between the predicted target path scheme and the labeled target path scheme reaches the preset convergence condition, the trained ablation path scheme recommendation model is obtained.

8. A path planning system for navigating lung lesions, characterized in that, The path planning system includes: The acquisition module is used to acquire lung CT images and input the lung CT images into a pre-trained ablation path recommendation model. The ablation path recommendation model performs the following processing: A construction module is used to construct a three-dimensional lung model based on the lung CT image. The three-dimensional lung model includes a bronchial tree, lesions, blood vessels, and pleura, and determines the spatial positional relationship of the lesions relative to the bronchial tree, blood vessels, and pleura. The determination module is used to determine, based on the spatial positional relationship, a target path planning scheme in the bronchial tree that satisfies the requirement that the puncture distance of the ablation device is less than a threshold and is directed toward the lesion, while avoiding the blood vessels and the pleura. The target path planning scheme includes the bronchial entry pathway, the puncture point location, and the puncture angle. The puncture angle characterizes the entry angle of the ablation device at the end of the bronchial entry pathway. The determining module is further configured to simulate the ablation device being guided through the bronchial entry pathway corresponding to the target path planning scheme to the corresponding puncture point, and to puncture the lesion at the corresponding puncture angle. Under the target ablation parameter combination setting, the predicted ablation area of ​​the target path planning scheme is determined, and a candidate path scheme is determined based on the comparison result between the predicted ablation area of ​​the target path planning scheme and the preset standard ablation area. The candidate path scheme includes the path planning scheme and the corresponding ablation parameter combination. The output module is used to comprehensively score and optimize the candidate path schemes based on a multi-objective optimization algorithm from multiple dimensions such as path length, accessibility from the entry angle, safety, and ablation thoroughness, so as to obtain the target path scheme output by the ablation path scheme recommendation model.

9. A path planning device for pulmonary lesion navigation, characterized in that, include: Central processing unit and memory; The memory is either a short-term storage memory or a persistent storage memory; The central processing unit is configured to communicate with the memory and execute instructions in the memory to perform the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 7.

Images