A needle knife mirror dynamic approach planning method and system

Abstract

The application provides a needle knife mirror dynamic approach planning method and system, relates to the needle knife mirror approach planning technical field, and has the technical scheme as follows: real-time image data of a joint cavity is acquired; the real-time image data is matched and analyzed with preoperative image data, changes of anatomical structures in the joint cavity are identified; according to the identified changes of anatomical structures in the joint cavity and a preset anatomical partition rule, anatomical partition information is dynamically updated, and updated anatomical partition information is obtained; according to the updated anatomical partition information and a preoperatively set approach path, real-time path adjustment is performed, and an optimized needle knife mirror safe approach path is generated; the real-time image data, the updated anatomical partition information and the optimized needle knife mirror safe approach path are superimposed and displayed, so as to guide a doctor to operate. The needle knife mirror dynamic approach planning method and system provided by the application can adapt to the dynamic changes of anatomical structures in the joint cavity during an operation, and has the advantages that the safe operation of surgical instruments is ensured.

Description

Technical Field

[0001] This application relates to the field of needle knife endoscope approach planning technology, and more specifically, to a dynamic approach planning method and system for needle knife endoscopes. Background Technology

[0002] In the field of modern medical technology, minimally invasive surgery is gaining increasing attention due to its advantages such as less trauma and faster recovery. Minimally invasive needle knife endoscopy for arthritis, as an important minimally invasive surgical method, plays a crucial role in the treatment of joint diseases. However, performing precise operations within the confined joint cavity remains a highly challenging task for surgeons. To ensure surgical safety and improve efficiency, preoperative planning of the approach path has become a standard procedure. This procedure aims to pre-determine the path for surgical instruments to enter the joint cavity, avoiding important anatomical tissues such as blood vessels and nerves, thereby reducing the risk of intraoperative injury.

[0003] However, traditional preoperative approach planning methods are based on static preoperative imaging data, ignoring the fact that the joint cavity is not a static environment. During actual surgery, factors such as the manipulation of surgical instruments, tissue traction, and changes in intra-articular pressure can all cause deformation of the soft tissues within the joint cavity, rendering the pre-planned fixed approach inapplicable during the operation, and potentially posing safety risks. This is especially true in complex arthritis surgeries, where the dynamic changes in the intra-articular anatomy are more pronounced, further highlighting the limitations of static planning methods.

[0004] Currently, intraoperative imaging technologies, such as X-ray fluoroscopy and ultrasound, have been applied to some extent in assisting surgery. These technologies are mainly used for the positioning and guidance of surgical instruments, helping surgeons to understand the real-time position of instruments within the joint cavity. However, existing technologies still lack a method to fully utilize real-time intraoperative imaging information to dynamically adjust the approach path to adapt to real-time changes in intraoperative anatomical structures. Therefore, how to effectively utilize real-time intraoperative imaging technology to overcome the challenges posed by dynamic changes in joint cavity anatomy, achieve real-time adjustment of the approach path, and ensure that surgical instruments are always operated within a safe area has become a key technical bottleneck in improving the safety and effectiveness of minimally invasive needle knife endoscopic surgery.

[0005] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0006] The purpose of this application is to provide a dynamic approach planning method and system for needle knife endoscopy, which has the advantages of being able to adapt to the dynamic changes in intra-articular anatomical structures during surgery and ensuring the safe operation of surgical instruments.

[0007] Firstly, this application provides a dynamic approach planning method for a needle knife endoscope, the technical solution of which is as follows:

[0008] The method includes:

[0009] Acquire real-time imaging data of the joint cavity;

[0010] Register and analyze real-time imaging data with preoperative imaging data to identify changes in intra-articular anatomical structures.

[0011] Based on the identified changes in intra-articular anatomical structures and the preset anatomical zoning rules, the anatomical zoning information is dynamically updated to obtain the updated anatomical zoning information.

[0012] Based on the updated anatomical zoning information and the preoperatively set access path, the path is adjusted in real time to generate an optimized safe access path for the needle knife endoscope.

[0013] The system overlays real-time image data, updated anatomical zoning information, and optimized needle knife endoscopy safety access path to guide the doctor's operation.

[0014] Furthermore, in this application, the step of registering and analyzing real-time image data with preoperative image data to identify changes in intra-articular anatomical structures includes:

[0015] Extract anatomical structure contour information from real-time image data;

[0016] The extracted anatomical structure contour information is registered with the anatomical structure contour information in the preoperative image data by nearest point registration, and the anatomical structure change matrix reflecting the changes in the anatomical structure within the joint cavity is calculated.

[0017] The step of dynamically updating the anatomical partition information based on the identified changes in intra-articular anatomical structures and preset anatomical partitioning rules to obtain the updated anatomical partition information includes:

[0018] Based on the anatomical structure change matrix, calculate the anatomical structure change amplitude and determine whether the anatomical structure change amplitude exceeds the preset threshold.

[0019] If the change in anatomical structure exceeds the preset threshold, the sensitivity of anatomical partition update is reduced. Based on the anatomical structure change matrix and the preset anatomical partition rules, the preoperative anatomical partition information is weighted and fused to obtain the updated anatomical partition information. The weight coefficient of the weighted fusion is inversely proportional to the change in anatomical structure.

[0020] If the change in anatomical structure does not exceed the preset threshold, the preoperative anatomical partition information is directly updated according to the anatomical structure change matrix and the preset anatomical partition rules to obtain the updated anatomical partition information.

[0021] Furthermore, in this application, the preset anatomical partitioning rules include:

[0022] Based on the safety level classification rules of anatomical structure type, blood vessels, nerves, ligaments, synovium and cartilage in the joint cavity are assigned different safety levels according to their functional importance;

[0023] Based on the anatomical spatial location regional division rules, the joint cavity space is divided into a safe operation area and a restricted area;

[0024] The buffer setting rules are based on the characteristics of anatomical structures, and different widths of safety buffers are set for different types of anatomical structures.

[0025] Furthermore, in this application, the step of adjusting the path in real time based on the updated anatomical partition information and the preoperatively set approach path to generate an optimized safe approach path for the needle knife endoscope includes:

[0026] Obtain the physical constraint parameters of the needle knife mirror;

[0027] Based on the updated anatomical partition information, a three-dimensional safety distance field within the joint cavity is constructed.

[0028] Based on the preoperatively defined access path and the three-dimensional safety distance field, an initial candidate path set is generated;

[0029] Based on the physical constraint parameters of the needle knife endoscope, the initial candidate path set is screened, and paths that do not meet the physical operation capabilities of the instrument are eliminated, resulting in a candidate path set that meets the physical constraints.

[0030] The candidate path set that meets the physical constraints is comprehensively evaluated for safety and operability, and the path with the best score is selected as the optimized safe access path for the needle knife endoscope.

[0031] Furthermore, in this application, the step of generating an initial candidate path set based on the preoperatively set approach path and the three-dimensional safety distance field includes:

[0032] Multiple path points are generated along the pre-set access path with a preset step size, serving as the starting points of candidate paths.

[0033] With each starting point as the center, random sampling is performed in the three-dimensional safety distance field to generate multiple direction vectors, each direction vector representing the initial direction of a candidate path;

[0034] Along each direction vector, the path is extended with a preset step size. It is determined whether the extended path exceeds the range of the three-dimensional safety distance field. If it does, the extension is stopped; otherwise, the extension continues until the preset maximum path length is reached, thus obtaining an initial candidate path set.

[0035] Furthermore, in this application, the step of filtering the initial candidate path set according to the physical constraint parameters of the needle knife endoscope, eliminating paths that do not meet the physical operation capabilities of the instrument, and obtaining a candidate path set that meets the physical constraints includes:

[0036] Obtain the bending angle threshold, rotation angle threshold, and maximum insertion depth of the needle knife mirror;

[0037] For each candidate path in the initial candidate path set, calculate the bending angle and rotation angle between adjacent path points on the path, and sum them to obtain the total bending angle and total rotation angle of the path.

[0038] Determine whether the total bending angle exceeds the bending angle threshold or the total rotation angle exceeds the rotation angle threshold. If either angle exceeds the threshold, then the candidate path is determined not to meet the physical constraints.

[0039] Calculate the length of the candidate path and determine whether the length exceeds the maximum insertion depth. If it does, the candidate path is determined not to meet the physical constraints.

[0040] Candidate paths that do not meet the physical constraints are removed from the initial candidate path set to obtain a candidate path set that meets the physical constraints.

[0041] Furthermore, in this application, the method also includes:

[0042] Acquire the operation instructions related to the needle knife endoscope during needle knife endoscopy, the operation instructions including at least the injection / aspiration instructions;

[0043] The changes in intra-articular anatomical structures generated based on the type of the operation command and the parameters of the operation command;

[0044] The safe approach path for needle knife endoscopy is updated based on the calculated changes in intra-articular anatomical structures.

[0045] After the operation corresponding to the operation instruction is completed, the deviation data between the actual changes in the intra-articular anatomical structure and the predicted changes in the intra-articular anatomical structure are recorded based on real-time image data.

[0046] The safe approach path of the needle knife mirror is updated a second time based on the deviation data.

[0047] Furthermore, in this application, the step of updating the safety access path of the needle knife mirror a second time based on the deviation data includes:

[0048] Based on the deviation data, a deviation vector field is constructed within the joint cavity, which characterizes the difference between the actual anatomical structure changes and the predicted anatomical structure changes at various locations within the joint cavity.

[0049] Based on the aforementioned deviation vector field, the updated safe approach path of the needle knife scope is adjusted, including:

[0050] Each path point on the path is translated along the deviation vector at the corresponding position in the deviation vector field. The translation step size is adjusted according to the preset learning rate to obtain the adjusted path point.

[0051] The adjusted path points are connected by B-spline curves to form the updated safe access path for the needle knife scope.

[0052] Furthermore, in this application, the step of overlaying and displaying real-time image data, updated anatomical zonal information, and optimized needle knife endoscopy safety access path to guide the doctor's operation includes:

[0053] Obtain the locations of key anatomical structures marked by doctors in real-time image data;

[0054] Based on the updated anatomical zoning information, the minimum distance between key anatomical structures and the optimized needle knife endoscope safe access path is calculated, and based on the preset distance threshold, it is determined whether the optimized needle knife endoscope safe access path poses a risk to key anatomical structures.

[0055] If a risk is constituted, a risk penalty coefficient is calculated based on the safety level of the key anatomical structures and the difference between the minimum distance and the distance threshold. The optimized needle knife endoscope safety approach path is then adjusted based on this risk penalty coefficient to obtain the risk-adjusted needle knife endoscope safety approach path.

[0056] The system overlays real-time imaging data, updated anatomical zoning information, and risk-adjusted safe access routes for needle knife endoscopy to guide the doctor's operation.

[0057] Secondly, this application also proposes a dynamic approach planning system for needle knife endoscopy, used in minimally invasive needle knife endoscopy surgery, the system comprising:

[0058] The acquisition module is used to acquire real-time imaging data of the joint cavity;

[0059] The recognition module is used to register and analyze real-time image data with preoperative image data to identify changes in intra-articular anatomical structures.

[0060] The update module is used to dynamically update the anatomical partition information based on the identified changes in the intra-articular anatomical structure and the preset anatomical partition rules, so as to obtain the updated anatomical partition information.

[0061] The generation module is used to adjust the path in real time based on the updated anatomical partition information and the preoperative set approach path, and generate an optimized needle knife endoscope safe approach path.

[0062] The display module is used to overlay and display real-time image data, updated anatomical zoning information, and optimized needle knife endoscopy safety access path to guide the doctor's operation.

[0063] As can be seen from the above, the needle knife endoscope dynamic approach planning method and system provided in this application utilizes real-time intraoperative image data to dynamically adjust the approach path, solving the problem that traditional preoperative approach path planning methods cannot adapt to the dynamic changes in intraoperative joint cavity anatomical structures. It has the advantages of being able to adapt to the dynamic changes in intraoperative joint cavity anatomical structures and ensuring the safe operation of surgical instruments. Attached Figure Description

[0064] Figure 1 This is a flowchart illustrating a dynamic approach planning method for a needle knife endoscope provided in this application.

[0065] Figure 2 This application provides a schematic diagram of a dynamic approach planning system for a needle knife endoscope.

[0066] In the diagram: 210, Acquisition module; 220, Recognition module; 230, Update module; 240, Generation module; 250, Display module. Detailed Implementation

[0067] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0068] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0069] Reference Figure 1 This application proposes a dynamic approach planning method for needle knife endoscopy, used in minimally invasive needle knife endoscopy surgery. The method includes:

[0070] S110. Acquire real-time imaging data of the joint cavity;

[0071] S120. Register and analyze real-time image data with preoperative image data to identify changes in intra-articular anatomical structures.

[0072] S130. Based on the identified changes in intra-articular anatomical structures and preset anatomical zoning rules, dynamically update the anatomical zoning information to obtain the updated anatomical zoning information.

[0073] S140. Based on the updated anatomical zoning information and the preoperatively set access path, perform real-time path adjustment to generate an optimized needle knife endoscope safe access path.

[0074] S150 overlays and displays real-time image data, updated anatomical zoning information, and optimized needle knife endoscope safety access path to guide the doctor's operation.

[0075] The step of acquiring real-time imaging data of the joint cavity can be achieved by using intraoperative C-arm X-ray machines, ultrasound imaging equipment, or optical tracking systems, or by using an endoscope integrated into the needle knife endoscope to acquire real-time images of the inside of the joint cavity.

[0076] By registering and analyzing real-time imaging data with preoperative imaging data, the steps of changes in intra-articular anatomical structures can be identified. Specifically, registration methods such as iterative nearest point algorithm or mutual information registration algorithm can be used to achieve accurate registration between real-time and preoperative images, thereby analyzing and quantifying the changes in intraoperative joint cavity anatomy relative to preoperative conditions.

[0077] Based on the identified changes in the anatomical structures within the joint cavity and the preset anatomical zoning rules, the steps to dynamically update the anatomical zoning information and obtain the updated anatomical zoning information are as follows. The preset anatomical zoning rules can be pre-defined. For example, based on anatomical atlases or clinical experience, the joint cavity can be divided into areas with different safety levels, and buffer zones can be set for important anatomical structures such as blood vessels and nerves.

[0078] When dynamically updating anatomical partition information, the preoperative anatomical partition information can be adjusted based on the anatomical structure change matrix obtained from registration analysis, thereby achieving dynamic updating of anatomical partitions.

[0079] Based on the updated anatomical partition information and the preoperatively set access path, the path is adjusted in real time to generate an optimized safe access path for the needle knife endoscope. Path planning algorithms such as the fast extended random tree algorithm or the A* search algorithm can be used to search for and generate an optimized safe access path for the needle knife endoscope based on the updated anatomical partition information and the preoperatively set access path.

[0080] The system overlays real-time image data, updated anatomical zoning information, and optimized safe access routes for needle knife endoscopy to guide the surgeon's procedures. For example, the optimized safe access routes can be highlighted or overlaid on the real-time images with different colors, while simultaneously displaying the updated anatomical zoning information, providing the surgeon with intuitive surgical guidance.

[0081] Specifically, this application first acquires real-time imaging data of the joint cavity to provide a data foundation for dynamic path planning. Then, by registering and analyzing the real-time imaging data with preoperative imaging data, it can accurately identify various changes in the anatomical structures within the joint cavity during surgery, such as soft tissue displacement or deformation. For the identified anatomical changes, this application dynamically updates the anatomical partition information according to preset anatomical partitioning rules, ensuring that the anatomical partitioning information reflects the actual situation of the joint cavity in real time. Based on this, this application adjusts the path in real time according to the updated anatomical partitioning information and the preoperatively set access path, generating an optimized needle knife endoscopy safe access path. This ensures that the planned access path always remains within the safe area, effectively avoiding important anatomical structures within the joint cavity. Finally, by overlaying and displaying the real-time imaging data, the updated anatomical partitioning information, and the optimized needle knife endoscopy safe access path, the surgeon can intuitively understand the actual situation of the joint cavity during surgery, the division of safe areas, and the optimized access path, thereby receiving effective guidance during the operation and improving the safety and precision of the surgery. Therefore, the proposed solution enables dynamic planning of the surgical approach for needle knife endoscopy, solving the technical problem of preoperative planned path failure caused by intraoperative changes in anatomical structure.

[0082] This application further proposes steps for registering and analyzing real-time imaging data with preoperative imaging data to identify changes in intra-articular anatomical structures, including:

[0083] Extract anatomical structure contour information from real-time image data;

[0084] The extracted anatomical structure contour information is registered with the anatomical structure contour information in the preoperative image data by nearest point registration, and the anatomical structure change matrix reflecting the changes in the anatomical structure within the joint cavity is calculated.

[0085] The steps for dynamically updating anatomical partition information based on identified changes in intra-articular anatomical structures and pre-defined anatomical partitioning rules to obtain updated anatomical partition information include:

[0086] Based on the anatomical structure change matrix, calculate the anatomical structure change amplitude and determine whether the anatomical structure change amplitude exceeds the preset threshold.

[0087] If the change in anatomical structure exceeds the preset threshold, the sensitivity of anatomical partition update is reduced. Based on the anatomical structure change matrix and the preset anatomical partition rules, the preoperative anatomical partition information is weighted and fused to obtain the updated anatomical partition information. The weight coefficient of the weighted fusion is inversely proportional to the change in anatomical structure.

[0088] If the change in anatomical structure does not exceed the preset threshold, the preoperative anatomical partition information is directly updated according to the anatomical structure change matrix and the preset anatomical partition rules to obtain the updated anatomical partition information.

[0089] In the identification of anatomical structure changes, the first step is to extract the contour information of the anatomical structures from real-time image data. Contour information can be understood as the boundary lines or outlines of the anatomical structures on the image. Extraction methods can employ edge detection algorithms commonly used in image processing, such as Canny edge detection or the Sobel operator. These algorithms can effectively identify regions in the image where pixel grayscale values ​​change abruptly; these regions typically correspond to the boundaries of the anatomical structures.

[0090] Then, the anatomical contour information extracted from the real-time images is registered with the existing anatomical contour information in the preoperative image data using nearest-neighbor registration. Nearest-neighbor registration is a point set registration method that aims to find the optimal transformation relationship between two point sets, minimizing the sum of distances from each point in one set to the nearest point in the other. In this scheme, nearest-neighbor registration establishes the correspondence between the real-time and preoperative anatomical contours, and calculates the anatomical transformation matrix. This matrix describes the displacement, rotation, and deformation of intra-articular anatomical structures during surgery.

[0091] Regarding the dynamic updating of anatomical regions, firstly, the magnitude of anatomical structural changes is calculated based on the anatomical structural change matrix. The magnitude of anatomical structural changes can be understood as the degree or magnitude of the changes. Calculation methods could include calculating the norm of the anatomical structural change matrix or calculating the average displacement of key points on the anatomical structure.

[0092] Then, the calculated magnitude of change is compared with a preset threshold. The preset threshold is a pre-defined value used to determine whether changes in anatomical structures are significant. If the magnitude of change exceeds the preset threshold, it is determined that the anatomical structure has undergone significant changes, and in this case, the sensitivity of anatomical region updates is reduced. The method for reducing update sensitivity is to use a weighted fusion strategy. Weighted fusion means that when updating anatomical region information, both preoperative anatomical region information and change information calculated from real-time images are considered, and different weights are assigned to each.

[0093] The weighting coefficients are inversely proportional to the magnitude of anatomical structural changes, meaning that the greater the magnitude of the change, the higher the weight of preoperative anatomical partition information and the lower the weight of real-time change information. Conversely, if the magnitude of the change does not exceed a preset threshold, it is determined that the anatomical structural changes are small. In this case, the preoperative anatomical partition information is directly updated according to the anatomical structural change matrix and the preset anatomical partitioning rules to obtain the updated anatomical partition information.

[0094] Direct updates mean fully adopting the change information calculated from real-time images to update anatomical partitions, maintaining the sensitivity and timeliness of partition updates.

[0095] Specifically, this approach addresses the potential dynamic changes in the anatomical structure of the joint cavity during minimally invasive needle knife endoscopic surgery by proposing a refined strategy for identifying anatomical changes and an adaptive dynamic update strategy for anatomical partitions.

[0096] In the anatomical structure change recognition stage, contour extraction and nearest-point registration enable a quantitative description of anatomical structure changes, overcoming the limitations of traditional methods in accurately capturing intraoperative anatomical deformation. In the dynamic update of anatomical zones, the concept of the magnitude of anatomical structure change is introduced, and based on this, an adaptive zone update strategy is proposed. When the magnitude of anatomical structure change exceeds a preset threshold, the system determines that a significant change has occurred, reduces the zone update sensitivity, and updates the anatomical zone information using a weighted fusion method. This approach effectively suppresses abrupt changes in anatomical zone information caused by intraoperative noise or transient disturbances, maintaining the stability of the anatomical zones. The design of weighting coefficients that are inversely proportional to the magnitude of anatomical structure change ensures that preoperative anatomical zone information plays a greater role when the magnitude of anatomical structure change is large, guaranteeing the reliability of zone updates.

[0097] When the change in anatomical structure does not exceed a preset threshold, the system determines that the change is minor and directly updates the anatomical zoning information based on real-time images, ensuring the timeliness and sensitivity of zoning updates. Through this hierarchical processing method, this scheme dynamically adjusts the anatomical zoning information update strategy according to the degree of anatomical structure change. When significant changes occur in the anatomical structure, it ensures the stability and reliability of the zoning; when the changes are minor, it ensures the sensitivity and timeliness of zoning updates, thus improving the adaptability of the dynamic approach planning method to intraoperative anatomical structure changes.

[0098] In some specific implementations, during arthroscopic minimally invasive surgery, the surgeon first acquires preoperative CT or MRI image data and plans anatomical partitions and access routes based on this data. During the surgery, real-time image data of the joint cavity is acquired using an arthroscope. The system extracts the contour information of anatomical structures such as cartilage and ligaments from both the real-time and preoperative image data. The contour extraction algorithm employs the Canny edge detection algorithm. Then, the Iterative Closest Point (ICP) algorithm is used to register the real-time anatomical structure contours with the preoperative anatomical structure contours, calculating the anatomical structure change matrix. The anatomical structure change matrix can be decomposed into a rotation matrix and a translation vector. Based on the anatomical structure change matrix, the magnitude of the anatomical structure change is calculated. The magnitude of the change is calculated by calculating the magnitude of the translation vector. A preset threshold is set to 3 mm. When the magnitude of the translation vector exceeds 3 mm, the magnitude of the anatomical structure change is determined to exceed the preset threshold. In this case, the sensitivity of the anatomical partition update is reduced, and a weighted fusion method is used to update the anatomical partition information. The weight coefficient of the weighted fusion is calculated as: weight coefficient = threshold / magnitude of change. The preoperative anatomical partition information is weighted by a weighting coefficient, while the weight of real-time change information is 1 minus the weighting coefficient. When the magnitude of the translation vector does not exceed 3 mm, it is determined that the change in anatomical structure has not exceeded a preset threshold. In this case, the preoperative anatomical partition information is directly updated based on the anatomical structure change matrix and preset anatomical partitioning rules. The updated anatomical partition information is used for subsequent dynamic adjustment of the access path and safety assessment. Through the above steps, the system can adaptively adjust the anatomical partition information update strategy according to the degree of intraoperative anatomical structure change, improving the accuracy and safety of access path planning.

[0099] This application further proposes pre-defined anatomical zoning rules, including:

[0100] Based on the safety level classification rules of anatomical structure type, blood vessels, nerves, ligaments, synovium and cartilage in the joint cavity are assigned different safety levels according to their functional importance;

[0101] Based on the anatomical spatial location regional division rules, the joint cavity space is divided into a safe operation area and a restricted area;

[0102] Buffer setting rules based on anatomical structure characteristics allow for setting different widths of safety buffers for different types of anatomical structures.

[0103] The safety level classification rules based on anatomical structure type are implemented as follows: for example, blood vessels and nerves are assigned the highest safety level because of their highest functional importance and greatest risk of injury; ligaments are assigned a medium safety level; and synovium and cartilage are assigned a lower safety level. The classification of safety levels can be determined based on medical knowledge bases and clinical experience data.

[0104] The anatomical spatial location-based regional division rules are implemented as follows: for example, the joint cavity space can be divided into several regions, which are further divided into safe operating areas and restricted areas based on their anatomical structure distribution and surgical risk level. Safe operating areas are spaces where instruments can be safely operated, while restricted areas are areas that must be strictly avoided, such as areas with dense blood vessels and nerves.

[0105] The buffer zone setting rules based on anatomical structure characteristics are implemented by setting different widths of safety buffer zones for anatomical structures with different safety levels. For example, a wider safety buffer zone is set for high-safety-level blood vessels and nerves to ensure that the access path maintains a sufficient safe distance from these important structures; a narrower safety buffer zone is set for low-safety-level synovium and cartilage. The width of the buffer zone can be adjusted according to factors such as the characteristics of the anatomical structure, surgical risk assessment, and instrument size.

[0106] Specifically, pre-defined anatomical partitioning rules are used to update anatomical partitioning information in dynamic approach planning methods, thereby addressing the lack of specific definitions and structured design for these rules. First, based on safety level classification rules for anatomical structure types, the system can distinguish the importance of different anatomical structures within the joint cavity and prioritize avoiding high-risk structures during path planning. Second, based on regional partitioning rules based on the spatial location of anatomical structures, the system can constrain path planning at a macro level, ensuring that the path generally avoids dangerous areas. Finally, based on buffer zone setting rules based on anatomical structure characteristics, the system can set different widths of safety buffer zones for different types of anatomical structures, thereby controlling safety distances more precisely. Through the synergistic effect of these three rules, pre-defined anatomical partitioning rules provide a comprehensive and detailed anatomical partitioning method, enabling dynamic approach planning methods to more effectively ensure surgical safety.

[0107] This application further proposes the following steps for generating an optimized safe access path for the needle knife endoscope by adjusting the path in real time based on the updated anatomical partition information and the preoperatively set access path:

[0108] Obtain the physical constraint parameters of the needle knife mirror;

[0109] Based on the updated anatomical partition information, a three-dimensional safety distance field within the joint cavity is constructed.

[0110] Based on the preoperatively defined access path and three-dimensional safety distance field, an initial candidate path set is generated;

[0111] Based on the physical constraint parameters of the needle knife endoscope, the initial candidate path set is screened, and paths that do not meet the physical operation capabilities of the instrument are eliminated, resulting in a candidate path set that meets the physical constraints.

[0112] A comprehensive evaluation of safety and operability is conducted on the candidate path set that meets the physical constraints, and the path with the best score is selected as the optimized safe approach path for the needle knife endoscope.

[0113] Specifically, obtaining the physical constraint parameters of the needle knife endoscope can involve reading parameters such as the maximum bending angle, maximum rotation angle, and maximum insertion depth of the endoscope from a preset parameter database. Constructing a three-dimensional safety distance field within the joint cavity can be achieved by first discretizing the joint cavity space into a three-dimensional mesh, and then calculating the shortest distance from the center point of each mesh cell to the restricted area based on the updated anatomical partition information. This distance value is the safety distance value of that mesh cell, and the safety distance values ​​of all mesh cells together constitute the three-dimensional safety distance field.

[0114] To generate an initial candidate path set, a random sampling method can be used. Starting from a point on the pre-defined access path, multiple direction vectors are randomly generated in a safe distance field, and multiple candidate paths are generated by extending along the direction vectors.

[0115] The initial candidate path set is filtered. For each candidate path, its total bending angle, total rotation angle, and path length can be calculated. If any parameter exceeds the physical constraint threshold of the needle knife mirror, the path is eliminated.

[0116] A comprehensive evaluation of safety and operability can be conducted by constructing a comprehensive scoring function. This function takes into account both the safety indicators of the path (such as the minimum distance between the path and the restricted area) and the operability indicators (such as the path length and curvature). The candidate paths are then ranked according to the scoring function, and the path with the highest score is selected as the optimized safe approach path for the needle knife endoscope.

[0117] Specifically, addressing the issue that preoperatively planned pathways in minimally invasive needle knife endoscopy for arthritis are difficult to adapt to dynamic changes in intraoperative anatomical structures, this approach first considers the physical manipulation capabilities of the needle knife endoscope when generating an optimized safe approach. By obtaining the physical constraint parameters of the needle knife endoscope, such as maximum bending angle, maximum rotation angle, and maximum insertion depth, the feasible range for path generation is limited.

[0118] Then, based on the updated anatomical partition information, a three-dimensional safety distance field is constructed to quantify the safety level within the joint cavity.

[0119] Based on this, and guided by the preoperatively set approach path, a series of initial candidate paths are generated, which initially ensure safety.

[0120] Importantly, the candidate paths are then rigorously screened using the physical constraint parameters of the needle knife mirror, eliminating physically infeasible paths to ensure that the final optimized path is one that the needle knife mirror can actually operate on.

[0121] Finally, among the physically feasible paths, the optimal path is selected through a comprehensive evaluation of safety and operability, thereby ensuring that the optimized approach is both safe and practical, and improving the safety and effectiveness of minimally invasive needle knife endoscopic surgery.

[0122] In some specific implementations, the physical constraint parameters of the needle knife endoscope are set as follows: maximum bending angle of 60 degrees, maximum rotation angle of 120 degrees, and maximum insertion depth of 15 cm. When constructing the three-dimensional safety distance field, the joint cavity space is discretized into a cubic mesh with a side length of 1 mm. When generating the initial candidate path set, 10 direction vectors are randomly generated starting from points every 2 mm along the preoperatively defined approach path, and the maximum length of each candidate path is limited to 20 cm. During path selection, the calculation step size for bending and rotation angles is set to 1 mm. For the comprehensive evaluation of safety and operability, the safety score weight is set to 0.7, and the operability score weight is set to 0.3. The safety score indicator uses the minimum distance between the path and the restricted area, and the operability score indicator uses the path length. The scoring function is designed as a weighted sum of the safety score and the operability score.

[0123] By setting the specific parameters mentioned above, an optimized safe approach path for the needle knife endoscope can be generated that satisfies both the physical constraints of the endoscope and ensures safety and operability.

[0124] This application further proposes a step for generating an initial candidate path set based on a preoperatively defined approach path and a three-dimensional safety distance field, including:

[0125] Multiple path points are generated along the pre-set access path with a preset step length, serving as the starting points of candidate paths.

[0126] With each starting point as the center, random sampling is performed in the three-dimensional safe distance field to generate multiple direction vectors, each of which represents the initial direction of a candidate path;

[0127] Along each direction vector, the path is extended with a preset step size. It is determined whether the extended path exceeds the range of the three-dimensional safe distance field. If it does, the extension is stopped; otherwise, the extension continues until the preset maximum path length is reached, thus obtaining the initial candidate path set.

[0128] In the step of generating the initial candidate path set, the starting point is generated by generating multiple path points along the pre-set access path direction according to a preset step size. These path points are used as the starting point of the candidate path. The preset step size can be adjusted according to the actual application scenario. For example, the step size can be set to 1 mm, 2 mm or other values ​​to control the distance and number between the starting points.

[0129] The direction vector generation method involves random sampling within a three-dimensional safe distance field, centered on each starting point. The aim is to obtain multiple direction vectors, each representing the initial extension direction of a candidate path. The range and method of random sampling can be set according to the characteristics of the safe distance field and the desired path diversity. For example, sampling can be performed uniformly within a spherical region around the starting point, or weighted sampling can be performed based on the gradient information of the safe distance field. The path extension and safety judgment method involves extending the path along each direction vector with a preset step size. After each extension, the system determines whether the new path segment exceeds the range of the three-dimensional safe distance field. The judgment method can be to check whether the path point is located within the safe distance field, or to calculate the distance between the path segment and the boundary of the safe distance field. If the path exceeds the safe range, the extension in that direction is stopped; otherwise, the extension continues until the preset maximum path length is reached. The setting of the maximum path length depends on the size of the surgical space and the operating range of the instruments, with the aim of limiting the search space and improving computational efficiency.

[0130] Specifically, generating an initial candidate path set is for a more efficient and comprehensive search of the safe access path space. First, using the pre-defined access path as a reference, a series of starting points are generated along this path with a fixed step size. The advantage of this is that it assumes the pre-planned path has a certain degree of rationality, and expanding the search based on this can improve efficiency. Subsequently, at each starting point, multiple direction vectors are selected in the three-dimensional safe distance field using a random sampling method. These vectors represent multiple potential safe path directions starting from the starting point. The introduction of random sampling increases the diversity of path exploration and avoids searching only in a small area near the pre-operative path. Finally, the path is extended along each selected direction vector, and during the extension process, it is detected in real time whether the path exceeds the boundary of the safe distance field. Once it exceeds the boundary, the extension is stopped immediately to ensure that all generated initial candidate paths are within the safe area. Through the above steps, a set containing multiple safe candidate paths can be obtained. This set covers the path space from the vicinity of the pre-operative path to the boundary of the safe area, providing a high-quality initial path selection range for subsequent path screening and optimization steps.

[0131] In some specific implementations, the preset step size is set to 2 mm. Ten starting points are generated along the pre-defined access path. Uniform random sampling is performed within a spherical region with a radius of 5 mm around each starting point to generate 20 direction vectors. The maximum path length is set to 50 mm, and the step size for path extension is 1 mm. For each direction vector, starting from the starting point, the path extends forward with a step size of 1 mm. After each extension, it is checked whether the new path point is within the three-dimensional safe distance field. If the path point is still within the safe distance field, the path continues to extend until the path length reaches 50 mm or exceeds the boundary of the safe distance field. In this way, up to 20 safe candidate paths with a length not exceeding 50 mm are generated for each starting point, and finally a set containing up to 200 initial candidate paths is obtained.

[0132] This application further proposes the following steps: obtaining the bending angle threshold, rotation angle threshold, and maximum insertion depth of the needle knife mirror; for each candidate path in the initial candidate path set, calculating the bending angle and rotation angle between adjacent path points on the path, and accumulating them to obtain the total bending angle and total rotation angle of the path; determining whether the total bending angle exceeds the bending angle threshold or the total rotation angle exceeds the rotation angle threshold; if either angle exceeds the threshold, the candidate path is determined to not meet the physical constraints; calculating the length of the candidate path and determining whether the length exceeds the maximum insertion depth; if it does, the candidate path is determined to not meet the physical constraints; removing candidate paths that do not meet the physical constraints from the initial candidate path set to obtain a candidate path set that meets the physical constraints.

[0133] Among them, the bending angle threshold, rotation angle threshold, and maximum insertion depth are the physical constraint parameters of the needle knife endoscope, which limit the physical limits of its operation. These parameters can be obtained by consulting the product manual or through experimental testing. The bending angle and rotation angle between adjacent path points are calculated and summed to obtain the total bending angle and total rotation angle, which can be achieved using a vector method. Specifically, for three consecutive path points, the first vector formed by the first two points and the second vector formed by the last two points can be calculated, and the bending angle can be calculated using the formula for the angle between the vectors. The rotation angle can be calculated based on the orientation changes of the path points in space. The path length can be obtained by summing the distances between adjacent path points. Whether a candidate path meets the physical constraints is determined by comparing the calculated total bending angle, total rotation angle, and path length with the corresponding thresholds. If any parameter exceeds the threshold, the path is deemed physically infeasible. Paths that do not meet the physical constraints are eliminated, while paths that do meet them are retained, ensuring that subsequent path optimization and selection are performed within the operable path range.

[0134] Specifically, after generating the initial candidate path set, the system needs to evaluate whether these paths can actually be executed by the needle knife endoscope. The evaluation process first obtains the physical limiting parameters of the needle knife endoscope, such as the maximum allowable bending angle, maximum rotation angle, and maximum insertion depth.

[0135] Subsequently, for each candidate path, the system calculates the bending and rotation amplitudes that the needle knife mirror needs to perform along the path. This is achieved by analyzing the angular changes between consecutive path points and summing these angular changes to obtain the total bending angle and total rotation angle of the entire path.

[0136] Simultaneously, the system also calculates the length of the entire path. After calculation, it compares the total bending angle of the path with the bending angle threshold, the total rotation angle with the rotation angle threshold, and the path length with the maximum insertion depth. If any of these exceed the corresponding threshold, it indicates that the path is physically infeasible, and the needle knife scope cannot operate along this path.

[0137] Therefore, the system will remove these infeasible paths from the candidate path set and only retain those paths that are within the physical operation capabilities of the needle knife endoscope, thereby ensuring that subsequent path optimization and selection are based on actually operable paths, and that the final planned safe approach path can be effectively executed by the doctor.

[0138] In some specific embodiments, the bending angle threshold of the needle knife mirror is set to 60 degrees, the rotation angle threshold is set to 180 degrees, and the maximum insertion depth is set to 15 cm. For one candidate path, the calculated total bending angle is 70 degrees, the total rotation angle is 150 degrees, and the path length is 12 cm. Since the total bending angle of 70 degrees exceeds the bending angle threshold of 60 degrees, this candidate path is determined not to meet the physical constraints and is removed from the candidate path set. Another candidate path has a calculated total bending angle of 50 degrees, a total rotation angle of 170 degrees, and a path length of 18 cm. Since the path length of 18 cm exceeds the maximum insertion depth of 15 cm, this candidate path is also determined not to meet the physical constraints and is removed from the candidate path set. After screening, the remaining candidate paths are all those with a total bending angle less than or equal to 60 degrees, a total rotation angle less than or equal to 180 degrees, and a path length less than or equal to 15 cm. These paths constitute the candidate path set that meets the physical constraints, providing a physically feasible range of path selections for subsequent safety and operability assessments.

[0139] This application further proposes to obtain the operation instructions for the needle knife endoscope during needle knife endoscope surgery, and the operation instructions include at least the injection / aspiration instructions;

[0140] The changes in intra-articular anatomical structures are calculated based on the type and parameters of the operation command.

[0141] The safe approach path for needle knife endoscopy is updated based on the calculated changes in intra-articular anatomical structures.

[0142] After the operation corresponding to the operation command is completed, the deviation data between the actual changes in the intra-articular anatomical structure and the predicted changes in the intra-articular anatomical structure are recorded based on real-time image data.

[0143] The safe approach path of the needle knife scope is updated a second time based on the deviation data.

[0144] In the operation instruction acquisition stage, the system can be configured to listen to the doctor's input of instructions when operating the needle knife endoscope control panel, or to capture the doctor's verbal instructions through voice recognition technology. The operation instruction type can be limited to several preset types, such as injection, aspiration, and irrigation. Each operation instruction may be associated with parameters. For example, the parameter for the injection instruction can be the volume or flow rate of the injected fluid, and the parameter for the aspiration instruction can be the volume of the aspirated fluid. In the anatomical structure change calculation stage, a joint cavity biomechanical model can be pre-established. This model describes the relationship between the pressure, volume, and anatomical structure position and shape within the joint cavity. When an operation instruction is received, the system predicts changes in the anatomical structure within the joint cavity based on the operation instruction type and parameters, combined with the biomechanical model. For example, the injection operation is predicted to cause an increase in the joint cavity volume and the soft tissue to be pushed apart.

[0145] In the safe access route update process, the predicted anatomical changes are used to adjust the original safe access route. The adjustment method can be to directly replan the route in the area where the prediction has changed, or to fine-tune the original route to avoid areas where the prediction may pose a risk.

[0146] In the deviation data recording stage, after the operation command is completed, the system analyzes the real-time image data, identifies the actual changes in anatomical structures, compares the actual changes with the predicted changes, and calculates the deviation. The deviation data can include positional deviation, deformation deviation, etc. In the secondary update stage, the deviation data is used to further optimize the safe access path. For example, if there is a deviation between the actual changes and the predicted changes, the system can adjust the prediction model parameters according to the deviation, or directly correct the path based on the deviation vector field.

[0147] Specifically, regarding fluid injection, when a doctor issues an injection command and sets the injection volume to 10ml of normal saline, the system first records the injection command and parameters. Then, the system calls a preset joint cavity biomechanical model. The model parameters include the initial joint cavity volume and tissue elastic modulus. The model calculates and predicts that after injecting 10ml of fluid, the joint cavity volume will increase, and soft tissue in a specific area will shift. Based on the predicted tissue displacement information, the system updates the safe access path. For example, the original path might be close to a blood vessel, but the prediction shows that the blood vessel will shift away from the original path after injection. In this case, the updated path can be appropriately moved towards the original blood vessel. The system makes fine-tuning of the direction to better utilize the surgical space. After the injection is completed, real-time images show the actual expansion of the joint cavity. The system compares the actual expansion with the expansion predicted by the model to obtain deviation data. For example, the actual blood vessel displacement may deviate from the predicted displacement by 2 mm. Based on this 2 mm deviation vector, the system readjusts the updated path to more accurately avoid blood vessels. Thus, through the operation command prediction and deviation correction mechanism, the safe approach path can respond more quickly and accurately to the dynamic changes in the joint cavity anatomy caused by the surgical operation, improving surgical safety and operational efficiency.

[0148] In some specific implementations, the mechanical model can be a finite element method-based model. This model discretizes the joint cavity and surrounding tissues into a finite number of elements and sets material properties and boundary conditions. The operation command parameters serve as the input conditions of the model. By solving the finite element equations, the displacement and deformation fields of the anatomical structures within the joint cavity are obtained. Deviation data recording can be achieved through image registration. The real-time image after the operation command is completed is registered with the image before the operation command is executed, and the registration transformation parameters are calculated. These transformation parameters characterize the actual changes in the anatomical structures. During the second update, the deviation vector field can be constructed using the radial basis function interpolation method. Using a small number of discrete deviation data points, the deviation vector field of the entire joint cavity space is interpolated. Path adjustment can be achieved using the gradient descent method, with path safety and operability as the objective functions. The deviation information provided by the deviation vector field serves as the gradient information, and the path point positions are iteratively optimized to obtain the safe entry path after the second update.

[0149] This application further proposes a step for secondary updating of the safe access path of the needle knife endoscope based on deviation data, including: constructing a deviation vector field within the joint cavity based on the deviation data, wherein the deviation vector field characterizes the difference between the actual anatomical structure changes and the predicted anatomical structure changes at various locations within the joint cavity; adjusting the updated safe access path of the needle knife endoscope based on the deviation vector field, including: translating each path point along the deviation vector at the corresponding position in the deviation vector field, wherein the translation step size is adjusted according to a preset learning rate to obtain the adjusted path point; and connecting the adjusted path points through a B-spline curve to form the secondary updated safe access path of the needle knife endoscope.

[0150] The construction of the deviation vector field can be implemented as follows:

[0151] First, after the needle knife endoscopy procedure is completed, the actual changes in the intra-articular anatomical structure are obtained through real-time imaging data recording.

[0152] The predicted changes in intra-articular anatomy are calculated based on the type and parameters of the operating command.

[0153] Deviation data is generated by comparing actual changes with predicted changes.

[0154] For each location within the joint cavity, the deviation vector is calculated as the vector difference between the actual anatomical structure location and the predicted anatomical structure location. These deviation vectors are collected in the joint cavity space to construct a deviation vector field.

[0155] During the path adjustment phase, for each path point on the updated safe inbound path, its corresponding position in the deviation vector field is determined. The path point will then be translated along the direction of this deviation vector. The magnitude of the translation is controlled by a preset learning rate.

[0156] The learning rate, a value between 0 and 1, determines the step size for each adjustment. For example, if the learning rate is set to 0.5, the path point will be shifted along the direction of the deviation vector by half the length of the deviation vector.

[0157] By setting the learning rate, excessive oscillations during path adjustment can be avoided, ensuring the stability of the adjustment. After all path points have been adjusted, to ensure the smoothness of the path, the adjusted path points are connected by B-spline curves to generate the final secondary update safe access path. The use of B-spline curves ensures the continuity and differentiability of the path, which better meets the actual needs of needle knife endoscopic surgery.

[0158] Specifically, to address the potential discrepancy between predicted and actual anatomical changes, a method for secondary updating of the safe access path is proposed. During needle knife endoscopic surgery, although the access path is dynamically updated based on real-time imaging data, discrepancies may still exist between the predicted and actual anatomical changes due to the complexity of the anatomical structure and the unpredictability of the surgical procedure. To resolve this issue, after the surgical command is completed, the system records the actual anatomical changes and compares them with the previously predicted changes, calculating the deviation data.

[0159] Deviation data was used to construct a deviation vector field, which visually represents the predicted deviation at various locations within the joint cavity. Subsequently, based on this deviation vector field, the updated safe approach path for the scalpel endoscope was further adjusted.

[0160] Path adjustment is achieved by translating each point on the path along the corresponding deviation vector in the deviation vector field. The translation step size is controlled by a preset learning rate to ensure the stability and convergence of the adjustment process. Finally, the adjusted path points are connected by B-spline curves to generate a safe inbound path that is updated twice.

[0161] Therefore, the safe approach can more accurately adapt to actual anatomical changes, further improving the safety and precision of the surgery.

[0162] In some specific implementations, such as in knee arthroplasty, the surgeon administers an injection command to expand the joint cavity. The system predicts that the injection will result in a 10% increase in joint cavity volume and updates the access path based on this prediction. After the procedure, real-time image analysis shows that the actual joint cavity volume increased by 12%. The deviation is calculated as the 2% difference between the actual increase and the predicted increase. Based on this deviation data, a deviation vector field is constructed. Assuming a point P on the access path has a corresponding deviation vector V, and the learning rate is set to 0.3, the path point P will be translated along the direction of vector V by a distance of 0.3*|V| to obtain a new path point P'. After all path points are adjusted in this way, the new path points are connected by B-spline curves to form a second-updated access path. The second-updated access path more accurately reflects the actual anatomical structure of the joint cavity, providing the surgeon with safer and more precise access guidance.

[0163] This application further proposes to obtain the locations of key anatomical structures marked by doctors in real-time image data;

[0164] Based on the updated anatomical zoning information, the minimum distance between key anatomical structures and the optimized needle knife endoscope safe access path is calculated, and based on the preset distance threshold, it is determined whether the optimized needle knife endoscope safe access path poses a risk to key anatomical structures.

[0165] If a risk is constituted, a risk penalty coefficient is calculated based on the safety level of the key anatomical structures and the difference between the minimum distance and the distance threshold. The optimized needle knife endoscope safety approach path is then adjusted based on this risk penalty coefficient to obtain the risk-adjusted needle knife endoscope safety approach path.

[0166] The system overlays real-time imaging data, updated anatomical zoning information, and risk-adjusted safe access routes for needle knife endoscopy to guide the doctor's operation.

[0167] Among them, obtaining the location of key anatomical structures marked by doctors in real-time image data can be done by doctors directly clicking or circling the anatomical structures in the image on the display interface, and the system records the three-dimensional coordinates of these marked locations.

[0168] By combining the updated anatomical zoning information, the system can query the anatomical zoning of key anatomical structures and their safety levels.

[0169] Calculate the minimum distance between the key anatomical structure and the optimized safe approach path of the needle knife endoscope. Specifically, this can be achieved by calculating the Euclidean distance from the location point of the key anatomical structure to the nearest point on the approach path.

[0170] The preset distance threshold is a standard used to judge the level of risk. Different distance thresholds can be set according to the safety level of different anatomical structures. For example, the distance threshold can be set larger for blood vessels and nerves with high safety levels.

[0171] Determining whether the optimized needle knife endoscope safety approach poses a risk to critical anatomical structures is achieved by comparing the minimum distance with a distance threshold. If the minimum distance is less than the distance threshold, a risk is considered to exist. If a risk exists, a risk penalty coefficient needs to be calculated. The magnitude of the risk penalty coefficient is directly proportional to the safety level of the critical anatomical structure and inversely proportional to the difference between the minimum distance and the distance threshold. The higher the safety level, the smaller the minimum distance, and the larger the risk penalty coefficient.

[0172] The optimized needle knife endoscopy approach path is adjusted based on a risk penalty coefficient. This adjustment can be achieved, for example, by adding a risk penalty term to the objective function of the path optimization algorithm, increasing the risk penalty coefficient to guide the path away from critical anatomical structures. After obtaining the risk-adjusted needle knife endoscopy approach path, it is overlaid with real-time imaging data and updated anatomical partition information for physician reference.

[0173] Specifically, when observing real-time images during surgery, if a surgeon finds that the optimized approach path might be too close to a critical nerve, the surgeon can mark the nerve's location on the display interface. Upon receiving the surgeon's mark, the system first determines the nerve's safety level within the anatomical partition information; for example, the nerve's safety level is set to the highest level. Then, the system calculates the minimum distance between the marked nerve location and the current optimized approach path. Assuming a preset nerve safety distance threshold of 5 mm, a calculated minimum distance of 3 mm is less than the threshold, and the system determines that the current path poses a risk to the nerve. At this point, the system calculates a higher risk penalty coefficient, such as 0.8, based on the nerve's highest safety level and the 3 mm distance difference. In the path optimization algorithm, this risk penalty is added to the original path scoring function. This risk penalty significantly reduces the score of paths close to the nerve, guiding the algorithm to search for a new path. The new path optimization process will tend to select paths further away from the nerve, ultimately generating a risk-adjusted approach path. For example, the minimum distance between the adjusted path and the nerve becomes 6 mm, which is greater than the safety distance threshold, thus eliminating the risk. Finally, the system overlays images containing risk-adjusted access routes, neural markers, and updated anatomical zoning information to the physician, providing safer access guidance.

[0174] Reference Figure 2 This application further proposes a dynamic approach planning system for needle knife endoscopy, used in minimally invasive needle knife endoscopy surgery. The system includes:

[0175] The acquisition module 210 is used to acquire real-time imaging data of the joint cavity;

[0176] The recognition module 220 is used to register and analyze real-time image data with preoperative image data to identify changes in intra-articular anatomical structures.

[0177] The update module 230 is used to dynamically update the anatomical partition information based on the identified changes in anatomical structures and preset anatomical partition rules, so as to obtain the updated anatomical partition information.

[0178] The generation module 240 is used to perform real-time path adjustment based on the updated anatomical partition information and the preoperative set access path, and generate an optimized needle knife endoscope safe access path.

[0179] Display module 250 is used to overlay and display real-time image data, updated anatomical zoning information, and optimized needle knife endoscope safe access path to guide the doctor's operation.

[0180] By utilizing real-time intraoperative imaging data, the approach path can be dynamically adjusted, solving the problem that traditional preoperative approach path planning methods cannot adapt to the dynamic changes in intraoperative anatomical structures within the joint cavity. This approach has the advantage of being able to adapt to the dynamic changes in intraoperative anatomical structures within the joint cavity and ensuring the safe operation of surgical instruments.

[0181] Furthermore, in some preferred embodiments, the dynamic approach planning system for a needle knife endoscope proposed in this application can perform any of the steps in the above methods.

[0182] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A dynamic approach planning system for needle knife endoscopy, used in minimally invasive needle knife endoscopy surgery, characterized in that, The system includes: The acquisition module acquires real-time imaging data of the joint cavity; The recognition module registers and analyzes real-time image data with preoperative image data to identify changes in intra-articular anatomical structures. The update module dynamically updates the anatomical partition information based on the identified changes in the intra-articular anatomical structure and the preset anatomical partitioning rules, thus obtaining the updated anatomical partition information. The generation module adjusts the path in real time based on the updated anatomical partition information and the preoperatively set access path, and generates an optimized needle knife endoscope safe access path. The display module overlays and displays real-time image data, updated anatomical zoning information, and optimized needle knife endoscope safe access path to guide the doctor's operation. The process of adjusting the approach in real time based on the updated anatomical partition information and the preoperatively set approach path to generate an optimized safe approach path for the needle knife endoscope includes: Obtain the physical constraint parameters of the needle knife mirror; Based on the updated anatomical partition information, a three-dimensional safety distance field within the joint cavity is constructed. Based on the preoperatively defined access path and the three-dimensional safety distance field, an initial candidate path set is generated; Based on the physical constraint parameters of the needle knife endoscope, the initial candidate path set is screened, and paths that do not meet the physical operation capabilities of the instrument are eliminated, resulting in a candidate path set that meets the physical constraints. A comprehensive evaluation of safety and operability is performed on the candidate path set that meets the physical constraints, and the path with the best score is selected as the optimized safe access path for the needle knife mirror. The step of filtering the initial candidate path set according to the physical constraint parameters of the needle knife endoscope, eliminating paths that do not meet the physical operation capabilities of the instrument, and obtaining a candidate path set that meets the physical constraints includes: Obtain the bending angle threshold, rotation angle threshold, and maximum insertion depth of the needle knife mirror; For each candidate path in the initial candidate path set, calculate the bending angle and rotation angle between adjacent path points on the path, and sum them to obtain the total bending angle and total rotation angle of the path. Determine whether the total bending angle exceeds the bending angle threshold or the total rotation angle exceeds the rotation angle threshold. If either angle exceeds the threshold, then the candidate path is determined not to meet the physical constraints. Calculate the length of the candidate path and determine whether the length exceeds the maximum insertion depth. If it does, the candidate path is determined not to meet the physical constraints. Candidate paths that do not meet the physical constraints are removed from the initial candidate path set to obtain a candidate path set that meets the physical constraints.

2. The dynamic approach planning system for a needle knife endoscope according to claim 1, characterized in that, The process of registering and analyzing real-time image data with preoperative image data to identify changes in intra-articular anatomical structures includes: Extract anatomical structure contour information from real-time image data; The extracted anatomical structure contour information is registered with the anatomical structure contour information in the preoperative image data by nearest point registration, and the anatomical structure change matrix reflecting the changes in the anatomical structure within the joint cavity is calculated. The step of dynamically updating the anatomical partition information based on the identified changes in intra-articular anatomical structures and preset anatomical partitioning rules to obtain the updated anatomical partition information includes: Based on the anatomical structure change matrix, calculate the anatomical structure change amplitude and determine whether the anatomical structure change amplitude exceeds the preset threshold. If the change in anatomical structure exceeds the preset threshold, the sensitivity of anatomical partition update is reduced. Based on the anatomical structure change matrix and the preset anatomical partition rules, the preoperative anatomical partition information is weighted and fused to obtain the updated anatomical partition information. The weight coefficient of the weighted fusion is inversely proportional to the change in anatomical structure. If the change in anatomical structure does not exceed the preset threshold, the preoperative anatomical partition information is directly updated according to the anatomical structure change matrix and the preset anatomical partition rules to obtain the updated anatomical partition information.

3. The dynamic approach planning system for a needle knife endoscope according to claim 1, characterized in that, The preset anatomical partitioning rules include: Based on the safety level classification rules of anatomical structure type, blood vessels, nerves, ligaments, synovium and cartilage in the joint cavity are assigned different safety levels according to their functional importance; Based on the anatomical spatial location regional division rules, the joint cavity space is divided into a safe operation area and a restricted area; The buffer setting rules are based on the characteristics of anatomical structures, and different widths of safety buffers are set for different types of anatomical structures.

4. The dynamic approach planning system for a needle knife endoscope according to claim 1, characterized in that, The generation of the initial candidate path set based on the preoperatively defined approach path and the three-dimensional safety distance field includes: Multiple path points are generated along the pre-set access path with a preset step size, serving as the starting points of candidate paths. With each starting point as the center, random sampling is performed in the three-dimensional safety distance field to generate multiple direction vectors, each direction vector representing the initial direction of a candidate path; Along each direction vector, the path is extended with a preset step size. It is determined whether the extended path exceeds the range of the three-dimensional safety distance field. If it does, the extension is stopped; otherwise, the extension continues until the preset maximum path length is reached, thus obtaining an initial candidate path set.

5. The dynamic approach planning system for a needle knife endoscope according to claim 1, characterized in that, Also includes: Acquire the operation instructions related to the needle knife endoscope during needle knife endoscopy, the operation instructions including at least the injection / aspiration instructions; The changes in intra-articular anatomical structures generated based on the type of the operation command and the parameters of the operation command; The safe approach path for needle knife endoscopy is updated based on the calculated changes in intra-articular anatomical structures. After the operation corresponding to the operation instruction is completed, the deviation data between the actual changes in the intra-articular anatomical structure and the predicted changes in the intra-articular anatomical structure are recorded based on real-time image data. The safe approach path of the needle knife mirror is updated a second time based on the deviation data.

6. The dynamic approach planning system for a needle knife endoscope according to claim 5, characterized in that, The step of updating the safe access path of the needle knife mirror based on the deviation data includes: Based on the deviation data, a deviation vector field is constructed within the joint cavity, which characterizes the difference between the actual anatomical structure changes and the predicted anatomical structure changes at various locations within the joint cavity. Based on the aforementioned deviation vector field, the updated safe approach path of the needle knife scope is adjusted, including: Each path point on the path is translated along the deviation vector at the corresponding position in the deviation vector field. The translation step size is adjusted according to the preset learning rate to obtain the adjusted path point. The adjusted path points are connected by B-spline curves to form the updated safe access path for the needle knife scope.

7. The dynamic approach planning system for a needle knife endoscope according to claim 1, characterized in that, The method of overlaying and displaying real-time image data, updated anatomical zoning information, and optimized needle knife endoscopy safety access path to guide the doctor's operation includes: Obtain the locations of key anatomical structures marked by doctors in real-time image data; Based on the updated anatomical zoning information, the minimum distance between key anatomical structures and the optimized needle knife endoscope safe access path is calculated, and based on the preset distance threshold, it is determined whether the optimized needle knife endoscope safe access path poses a risk to key anatomical structures. If a risk is constituted, a risk penalty coefficient is calculated based on the safety level of the key anatomical structures and the difference between the minimum distance and the distance threshold. The optimized needle knife endoscope safety approach path is then adjusted based on this risk penalty coefficient to obtain the risk-adjusted needle knife endoscope safety approach path. The system overlays real-time imaging data, updated anatomical zoning information, and risk-adjusted safe access routes for needle knife endoscopy to guide the doctor's operation.

Images