Navigation positioning system for sphenoid palate ganglion

Abstract

The invention relates to a navigation positioning system for sphenoid palate ganglions. The system comprises a data processing modeling module which is used for acquiring a preoperative multi-modal image data set, extracting sphenoid palate ganglion features of an image, and generating an individualized anatomical model reflecting individual anatomical differences; the path planning module extracts boundary features of a key structure based on an individual anatomical model, generates an acupuncture path set and a candidate path set by adopting a path planning algorithm, and determines optimal acupuncture approach parameters. And the guide track generation module calculates the spatial attitude of the instrument by using a quaternion transformation algorithm according to the optimal acupuncture approach parameters, and generates a dynamic guide track. And the deviation correction updating module predicts a motion track through a Kalman filtering algorithm, calculates a real-time path deviation value, updates output parameters of a navigation system according to the deviation value, and finally generates an acupuncture guide path. By adopting the system, the position difference of the sphenoid palate ganglion and the adjacent tissues thereof can be effectively captured, and a whole-process accurate navigation support is provided for sphenoid palate ganglion acupuncture stimulation.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical devices, and particularly relates to a navigation and positioning system for the sphenopalatine ganglion. Background Art

[0002] In the technical field of medical devices, the precise diagnosis and treatment of sphenopalatine ganglion-related diseases (such as chronic headache, allergic rhinitis, facial neuralgia, etc.) rely on high-precision positioning of the acupuncture path. Traditional acupuncture operations mainly rely on anatomical knowledge and doctors' experience. Due to significant differences in individual skull base anatomical structures (such as the position of the sphenopalatine foramen and the adjacent relationship between nerves and blood vessels), it is easy to cause complications such as nerve injury and blood vessel rupture, and there is a lack of real-time intraoperative navigation means, making it difficult to dynamically correct the acupuncture trajectory. With the development of medical imaging technology and artificial intelligence, navigation systems based on multi-modal images have gradually been applied in clinical practice. However, for deep and tiny structures at the skull base such as the sphenopalatine ganglion, there are still technical bottlenecks in aspects such as the accuracy of feature extraction (such as the recognition of the gray-scale difference between nerves and surrounding soft tissues), the efficiency of individualized model construction (such as the rapid modeling of complex anatomical topologies), and the robustness of dynamic guidance (such as the correction of trajectory deviations caused by instrument jitter during acupuncture). Summary of the Invention

[0003] Based on this, in view of the above technical problems, it is necessary to provide a navigation and positioning system for the sphenopalatine ganglion that can achieve intelligent, precise, and individualized navigation and positioning for acupuncture stimulation of the sphenopalatine ganglion.

[0004] In a first aspect, the present application provides a navigation and positioning system for the sphenopalatine ganglion, including:

[0005] A data processing and modeling module, configured to obtain a preoperative multi-modal image data set; and also configured to perform feature extraction of the sphenopalatine ganglion on the multi-modal image data set, and generate an individualized anatomical model by combining a preset anatomical landmark data set for spatial topology modeling.

[0006] A path planning module, configured to perform key structure boundary feature extraction on the individualized anatomical model, calculate a set of acupuncture paths using a path planning algorithm, and generate a candidate path set; and also configured to determine the safety factor of the candidate path set according to the path length parameter and the risk weight parameter, and determine the optimal acupuncture access parameter.

[0007] A guiding trajectory generation module, configured to calculate the spatial attitude of the instrument using a quaternion transformation algorithm based on the optimal acupuncture access parameter to obtain a set of needle insertion trajectory parameters; and also configured to perform real-time superposition of the guiding path and the intraoperative image according to the set of needle insertion trajectory parameters to obtain a dynamic guiding trajectory.

[0008] The deviation correction and update module is used to predict the motion trajectory using the Kalman filter algorithm based on the dynamic guiding trajectory, calculate the real-time path deviation value; it is also used to adjust the display parameters of the human-computer interaction interface and update the output parameters of the navigation system according to the real-time path deviation value, and generate the final acupuncture guiding path.

[0009] In one embodiment, feature extraction of the sphenopalatine ganglion is performed on the multi-modal image dataset, and spatial topological modeling is combined with the preset anatomical landmark dataset to generate an individualized anatomical model, including:

[0010] Perform image registration on the multi-modal image dataset to generate registered three-dimensional voxel data.

[0011] Segment the sphenopalatine ganglion region from the three-dimensional voxel data according to the preset anatomical landmark dataset, and extract the feature point coordinates.

[0012] Perform spatial topological matching between the feature point coordinates and the anatomical landmark dataset to obtain a difference mapping matrix; the difference mapping matrix contains the Euclidean distance and the curvature difference weight.

[0013] Adjust the topological relationship of the anatomical landmark dataset according to the difference mapping matrix to generate an individualized spatial topological model.

[0014] Perform mesh parameterization on the individualized spatial topological model to generate an editable individualized anatomical model.

[0015] In one embodiment, the path planning module includes:

[0016] The candidate path planning unit is used for:

[0017] Obtain the key structure boundary feature data of the individualized anatomical model; the key structure boundary feature data is extracted by an image segmentation algorithm.

[0018] Generate a three-dimensional spatial topological map according to the key structure boundary feature data; the three-dimensional spatial topological map contains the distribution parameters of blood vessels and nerves.

[0019] Traverse the feasible path nodes of the three-dimensional spatial topological map using a path planning algorithm to obtain a candidate path set.

[0020] The optimal path planning unit is used for:

[0021] Calculate the path curvature parameter of the candidate path set; the path curvature parameter is obtained by differential operation of adjacent node coordinates.

[0022] Obtain the risk area weight parameter of the candidate path set; the risk area weight parameter is preset based on the anatomical structure type.

[0023] The dynamic risk value is calculated by a weighted summation formula based on the path curvature parameter and the risk area weight parameter.

[0024] Wherein, if the dynamic risk value exceeds a preset threshold, the path smoothness parameter of the candidate path set is adjusted.

[0025] The candidate path set is updated according to the adjusted path smoothness parameter to obtain an optimized path set.

[0026] Traverse the path length parameters of the optimized path set, and select the optimized path with the smallest path length parameter as the optimal needle insertion path parameter; the path length parameter is calculated by accumulating the Euclidean distances between nodes.

[0027] In one embodiment, the dynamic risk value is calculated by the following formula:

[0028]

[0029] Wherein, R dyn represents the dynamic risk value, n represents the total number of path nodes, k i represents the path curvature of the i-th node, α represents the curvature sensitivity index, α≥1, S i represents the anatomical structure type where the i-th node is located, μ(S i ) represents the risk weight value based on the fuzzy membership function, c represents the structural risk center value, σ represents the fuzzy domain parameter, l i represents the path element length of the i-th node, ω curv and ω risk respectively represent the weight coefficients of curvature and structural risk, satisfying ω curv +ω risk =1.

[0030] In one embodiment, based on the optimal needle insertion path parameter, a quaternion transformation algorithm is used to calculate the instrument spatial attitude to obtain a set of needle insertion trajectory parameters, including:

[0031] Obtain the instrument attitude matrix corresponding to the optimal needle insertion path parameter; the instrument attitude matrix contains the position and direction information of the instrument end in the three-dimensional coordinate system.

[0032] Extract the quaternion conversion parameters according to the instrument attitude matrix; the quaternion conversion parameters contain the rotation axis vector and the rotation angle component around the axis.

[0033] Input the quaternion conversion parameters into the attitude transformation model, and perform a rotation transformation on the instrument coordinate system based on the quaternion multiplication rule to obtain the transformed attitude data.

[0034] Optimize the pose according to the transformed pose data and path constraint conditions to obtain the optimized pose data; the path constraint conditions include the anatomical structure obstacle avoidance threshold and the instrument motion curvature limit.

[0035] Calculate the set of needle insertion angle parameters based on the optimized pose data; the set of needle insertion angle parameters includes the pitch angle deviation value and the azimuth angle correction amount.

[0036] Perform spatial projection matching on the set of needle insertion angle parameters and the instrument pose matrix to obtain the final needle insertion trajectory parameter group.

[0037] In one embodiment, the deviation correction and update module includes:

[0038] The deviation calculation unit is used for:

[0039] Predict the motion trajectory using the Kalman filter algorithm based on the dynamic guidance trajectory, and calculate the real-time path deviation value.

[0040] The instruction generation unit is used for:

[0041] Adjust the display parameters of the human-computer interaction interface according to the real-time path deviation value, and generate a set of control instructions through touch gesture recognition.

[0042] The path generation unit is used for:

[0043] Update the output parameters of the navigation system using the adaptive feedback algorithm according to the set of control instructions.

[0044] Based on the output parameters of the navigation system and the intraoperative image data, correct the coordinates of the anatomical landmark points using the iterative closest point algorithm to obtain the corrected coordinate deviation.

[0045] Judge the corrected coordinate deviation based on the preset threshold. If it is lower than the preset threshold, update the individual anatomical model; otherwise, generate the final needle insertion guidance path.

[0046] In one embodiment, predicting the motion trajectory using the Kalman filter algorithm based on the dynamic guidance trajectory and calculating the real-time path deviation value includes:

[0047] Initialize the state vector and covariance matrix of the Kalman filter based on the characteristics of the dynamic guidance trajectory.

[0048] Generate a predicted trajectory according to the state vector and covariance matrix; the predicted trajectory includes a set of coordinate points at the next moment.

[0049] Obtain the real-time motion trajectory data; calculate the residual sequence between the set of coordinate points and the real-time motion trajectory data.

[0050] Calculate the real-time path deviation value based on the residual sequence; the real-time path deviation value is the root mean square value of the residual sequence.

[0051] Determine whether the real-time path deviation value exceeds a preset deviation threshold. If it exceeds, adjust the covariance matrix parameters of the Kalman filter and re-iterate to obtain an updated real-time path deviation value.

[0052] In a second aspect, the present application also provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the foregoing method is implemented.

[0053] In a third aspect, the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the foregoing method is implemented.

[0054] For the above-mentioned navigation and positioning system, computer device and storage medium for the sphenopalatine ganglion, the data processing and modeling module acquires a pre-operative multi-modal image data set, extracts sphenopalatine ganglion features from the images through an algorithm, and completes image registration, feature segmentation and spatial topology modeling in combination with a preset anatomical landmark data set to generate an individual anatomical model reflecting individual anatomical differences. The path planning module extracts key structure boundary features based on the individual anatomical model, generates a set of needle insertion paths and a set of candidate paths using a path planning algorithm, and then evaluates the safety factor of the candidate paths according to the path length parameter and the risk weight parameter to determine the optimal needle insertion path parameters. The guiding trajectory generation module relies on the optimal needle insertion path parameters, calculates the spatial attitude of the instrument using a quaternion transformation algorithm, forms a set of needle insertion trajectory parameters, and overlays the guiding path and the intra-operative images in real time to generate a dynamic guiding trajectory. The deviation correction and update module predicts the motion trajectory through a Kalman filter algorithm, calculates the real-time path deviation value, adjusts the display parameters of the human-computer interaction interface according to the deviation value and updates the output parameters of the navigation system, and finally generates an accurate needle insertion guiding path. It can effectively capture the position differences of the sphenopalatine ganglion and its adjacent tissues, significantly improve the accuracy and safety of sphenopalatine ganglion needle stimulation, and provide accurate navigation support for the whole process of sphenopalatine ganglion needle stimulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0056] Figure 1 It is a structural block diagram of a navigation and positioning system for the sphenopalatine ganglion provided by an embodiment of the present invention;

[0057] Figure 2A flowchart for extracting sphenopalatine ganglion features from a multimodal image dataset, performing spatial topological modeling in combination with a preset anatomical landmark dataset, and generating an individualized anatomical model provided by an embodiment of the present invention;

[0058] Figure 3 A flowchart for predicting a motion trajectory using a Kalman filter algorithm based on a dynamic guidance trajectory and calculating a real-time path deviation value provided by an embodiment of the present invention. Detailed implementation manners

[0059] In order to make the objectives, technical solutions, and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0060] In one embodiment, as Figure 1 shown, the present application provides a navigation and positioning system for the sphenopalatine ganglion, which may include:

[0061] A data processing and modeling module 101, configured to obtain a preoperative multimodal image dataset; and further configured to extract sphenopalatine ganglion features from the multimodal image dataset, perform spatial topological modeling in combination with a preset anatomical landmark dataset, and generate an individualized anatomical model.

[0062] Specifically, this module is responsible for the acquisition and modeling of preoperative data. A multimodal image dataset including CT, MRI, etc. is obtained through a medical imaging system. Feature extraction is performed on the sphenopalatine ganglion region. The multimodal image is registered, segmented, and spatially topologically matched using a preset anatomical landmark dataset. By adjusting the topological relationship of the standard model, a three-dimensional volumetric anatomical model reflecting the individual anatomical structure characteristics of the patient is generated.

[0063] A path planning module 102, configured to extract key structure boundary features from the individualized anatomical model, calculate a set of acupuncture paths using a path planning algorithm, and generate a candidate path set; and further configured to determine the safety factor of the candidate path set according to the path length parameter and the risk weight parameter, and determine the optimal acupuncture access parameter;

[0064] Specifically, this module conducts the design of the acupuncture stimulation path based on the individualized anatomical model. First, the boundary features of key structures such as nerves and blood vessels in the model are extracted to construct a three-dimensional spatial topological map including anatomical structure distribution parameters; the feasible path nodes in the topological map are traversed through a path planning algorithm to generate an initial set of acupuncture paths and a candidate path set; further, in combination with the path length parameter and the preset risk weight parameter, the safety factor of the candidate path set is quantitatively evaluated, and the optimal acupuncture access parameter that takes into account both safety and operational convenience is selected to form a scientific acupuncture stimulation path planning scheme.

[0065] The guiding trajectory generation module 103 is configured to calculate the spatial attitude of the instrument using a quaternion transformation algorithm based on the optimal needle insertion path parameters, so as to obtain a set of needle insertion trajectory parameters; and is further configured to perform real-time superposition of the guiding path and the intraoperative image according to the set of needle insertion trajectory parameters to obtain a dynamic guiding trajectory.

[0066] Preferably, this module realizes the conversion from the needle insertion path parameters to real-time guidance. Based on the optimal needle insertion path parameters, a quaternion transformation algorithm is used to solve the spatial attitude of the needle stimulation instrument, and a set of needle insertion trajectory parameters including information such as the needle insertion angle and direction is generated; the virtual guiding path is registered and superposed with the intraoperative real-time acquired image data to form a dynamically visualized guiding trajectory, providing an intuitive spatial position guidance for the needle stimulation operation to ensure that the instrument travels precisely along the planned path.

[0067] The deviation correction and update module 104 is configured to predict the motion trajectory using a Kalman filtering algorithm based on the dynamic guiding trajectory, and calculate the real-time path deviation value; and is further configured to adjust the display parameters of the human-computer interaction interface and update the output parameters of the navigation system according to the real-time path deviation value to generate a final needle insertion guiding path.

[0068] Specifically, this module is responsible for the dynamic monitoring and correction of the intraoperative trajectory. The Kalman filtering algorithm is used to predict the instrument motion trajectory based on the dynamic guiding trajectory, and the operation error is evaluated by calculating the real-time path deviation value; the display parameters of the human-computer interaction interface are adjusted according to the deviation value, and the navigation system is driven to update the output parameters to form a control instruction; the coordinates of the anatomical landmark points are corrected in combination with the intraoperative image data, and if the deviation meets the preset threshold, a final needle insertion guiding path is generated to realize the closed-loop control and precise calibration of the needle stimulation trajectory.

[0069] The above-mentioned navigation and positioning system for the sphenopalatine ganglion, the data processing and modeling module obtains the preoperative multimodal image dataset, extracts the sphenopalatine ganglion features from the images through algorithms, and completes image registration, feature segmentation and spatial topology modeling in combination with the preset anatomical landmark dataset to generate an individualized anatomical model reflecting individual anatomical differences. The path planning module extracts the key structural boundary features based on the individualized anatomical model, uses the path planning algorithm to generate a set of needle insertion paths and candidate path sets, and then evaluates the safety factor of the candidate paths according to the path length parameter and the risk weight parameter to determine the optimal needle insertion approach parameters. The guiding trajectory generation module relies on the optimal needle insertion approach parameters, uses the quaternion transformation algorithm to calculate the spatial attitude of the instrument, forms a set of needle insertion trajectory parameters, and overlays the guiding path and the intraoperative image in real time to generate a dynamic guiding trajectory. The deviation correction and update module predicts the movement trajectory through the Kalman filter algorithm, calculates the real-time path deviation value, adjusts the display parameters of the human-computer interaction interface according to the deviation value and updates the output parameters of the navigation system, and finally generates an accurate needle insertion guiding path. It effectively captures the position differences of the sphenopalatine ganglion and its adjacent tissues, significantly improves the accuracy and safety of sphenopalatine ganglion needle stimulation, and provides accurate navigation support for the whole process of sphenopalatine ganglion needle stimulation.

[0070] In one of the embodiments, as Figure 2 shown, for the extraction of sphenopalatine ganglion features from the multimodal image dataset and the spatial topology modeling in combination with the preset anatomical landmark dataset to generate an individualized anatomical model, the following steps may be included:

[0071] Step S201, perform image registration on the multimodal image dataset to generate registered three-dimensional voxel data.

[0072] Step S202, segment the sphenopalatine ganglion region from the three-dimensional voxel data according to the preset anatomical landmark dataset, and extract the feature point coordinates.

[0073] Step S203, perform spatial topology matching between the feature point coordinates and the anatomical landmark dataset to obtain a difference mapping matrix; the difference mapping matrix includes the Euclidean distance and the curvature difference weight.

[0074] Preferably, when performing point-by-point spatial topology matching between the sphenopalatine ganglion feature point coordinates extracted from the three-dimensional voxel data and the standard coordinates in the preset anatomical landmark dataset, the spatial position deviation is measured by calculating the Euclidean distance between each pair of corresponding points, and the curvature value of the feature point neighborhood is calculated using differential geometry methods, and the curvature difference weight is formed with the curvature value of the corresponding region of the standard model. The two together form a difference mapping matrix, which numerically quantifies the dual differences in position and morphology between the individual anatomical structure and the standard model.

[0075] Step S204: Adjust the topological relationship of the anatomical landmark point dataset according to the difference mapping matrix to generate an individualized spatial topological model.

[0076] Step S205: Perform mesh parameterization on the individualized spatial topological model to generate an editable individualized anatomical model.

[0077] Specifically, first, the spatial position differences of different modality images are eliminated through an image registration algorithm to generate three-dimensional voxel data in a unified coordinate system; then, based on the preset anatomical landmark point dataset, the sphenopalatine ganglion region is accurately extracted from the three-dimensional voxel data using region segmentation technology to obtain the characteristic point coordinates. Through a spatial topology matching algorithm, the characteristic point coordinates are compared point by point with the anatomical landmark point dataset, and a difference mapping matrix including the Euclidean distance deviation and curvature difference weights is calculated; according to this matrix, the topological connection relationship of the anatomical landmark point dataset is adjusted to generate a spatial topological model adapted to individual anatomical characteristics; finally, the topological model is subjected to mesh parameterization to generate an editable triangular mesh or tetrahedral mesh structure to form the final individualized anatomical model.

[0078] This embodiment realizes the accurate positioning and digital extraction of the sphenopalatine ganglion anatomical structure; based on the topological adjustment mechanism of the difference mapping matrix, it effectively captures individual anatomical variations (such as nerve trajectory deviation, blood vessel distribution differences); the mesh parameterization provides the model with editability, supports local structure correction according to real-time image data during the operation, provides a high-precision and individualized anatomical basis for acupuncture stimulation path planning, and significantly improves the reliability of subsequent path planning and the safety of acupuncture stimulation operations.

[0079] In one embodiment, the path planning module may further include:

[0080] A candidate path planning unit 1021, configured to:

[0081] Step S301: Obtain the key structure boundary feature data of the individualized anatomical model; the key structure boundary feature data is extracted through an image segmentation algorithm.

[0082] Step S302: Generate a three-dimensional spatial topological map according to the key structure boundary feature data; the three-dimensional spatial topological map includes the distribution parameters of blood vessels and nerves.

[0083] Step S303: Use a path planning algorithm to traverse the feasible path nodes of the three-dimensional spatial topological map to obtain a candidate path set.

[0084] An optimal path planning unit 1022, configured to:

[0085] Step S304: Calculate the path curvature parameters of the candidate path set; the path curvature parameters are obtained through differential operations of adjacent node coordinates.

[0086] Step S305: Obtain the risk area weight parameter of the candidate path set; the risk area weight parameter is preset based on the anatomical structure type.

[0087] Step S306: Calculate the dynamic risk value according to the path curvature parameter and the risk area weight parameter through the weighted summation formula.

[0088] Wherein, if the dynamic risk value exceeds the preset threshold, the path smoothness parameter of the candidate path set is adjusted.

[0089] Step S307: Update the candidate path set according to the adjusted path smoothness parameter to obtain the optimized path set.

[0090] Step S308: Traverse the path length parameters of the optimized path set, and select the optimized path with the minimum path length parameter as the optimal needle insertion path parameter; the path length parameter is calculated by accumulating the Euclidean distances between nodes.

[0091] Specifically, the path planning module realizes the intelligent design of the acupuncture stimulation path through the collaborative operation of dual units: the candidate path planning unit first extracts the boundary feature data of key structures such as the sphenopalatine ganglion and adjacent blood vessels from the individual anatomical model by using the image segmentation algorithm, and constructs a three-dimensional space topological graph containing neurovascular distribution parameters (such as position, diameter, and walking direction); based on this topological graph, path planning algorithms such as A* and Dijkstra are used to traverse the feasible path nodes to generate a candidate path set covering potential acupuncture stimulation paths, ensuring the comprehensiveness of path search. The optimal path planning unit conducts multi-dimensional evaluation on the candidate path set: calculates the path curvature parameter through the differential operation of adjacent node coordinates to quantify the degree of path turning; combines the preset anatomical structure risk area weight parameter (such as the risk weight of the blood vessel area is higher than that of the bone), and calculates the dynamic risk value through the weighted summation formula; optimizes the high-risk path exceeding the preset threshold by adjusting the path smoothness parameter (such as increasing the curvature penalty coefficient) to generate an optimized path set with reduced risk; finally, traverses the accumulated Euclidean distance values of the nodes in the optimized path set, and selects the optimized path with the shortest path length as the optimal needle insertion path parameter.

[0092] This embodiment improves the scientificity of the acupuncture stimulation path design through the dual mechanisms of data-driven and rule constraint: the three-dimensional space topological graph transforms the complex anatomical structure into a computable graph theory model, making the path search more systematic; the dynamic risk value integrates curvature and anatomical risk parameters, avoiding the one-sidedness of traditional single-factor evaluation; the combined strategy of path smoothness adjustment and length optimization reduces the instrument operation distance and the probability of tissue damage on the premise of ensuring the safety of acupuncture stimulation. In addition, the parameterized evaluation system supports flexible adjustment of the weight coefficient according to different acupuncture stimulation requirements (such as minimally invasive priority or risk priority), improving the clinical adaptability of the system.

[0093] In one embodiment, the dynamic risk value can be calculated by the following formula:

[0094]

[0095] where R dyn represents the dynamic risk value, n represents the total number of path nodes, k i represents the path curvature of the i-th node, α represents the curvature sensitivity index, α ≥ 1, and S i represents the anatomical structure type where the i-th node is located, μ(S i ) represents the risk weight value based on the fuzzy membership function, c represents the structural risk center value, σ represents the fuzzy domain parameter, l i represents the path differential length of the i-th node, ω curv and ω risk respectively represent the weight coefficients of curvature and structural risk, satisfying ω curv + ω risk = 1.

[0096] The curvature sensitivity index α can be dynamically adjusted according to the minimum bending radius of different acupuncture stimulation instruments (such as the curvature limit of acupuncture needles). When α = 2, the risk value of the acute turning path increases exponentially, effectively suppressing the selection of high-curvature paths; the Gaussian membership function transforms the anatomical structure risk from discrete levels (such as blood vessels = high risk, bone = low risk) into a continuous spatial distribution. For example, the risk value within 5 mm around blood vessels decays smoothly through the σ parameter, more conforming to the complexity of the actual anatomical adjacency relationship; the weight coefficients ω curv and ω risk are determined by the Analytic Hierarchy Process (AHP) or clinical data training, and can be adapted to different acupuncture stimulation scenarios such as neurosurgery (focusing on risk avoidance) and orthopedics (focusing on path efficiency), making the risk assessment system more flexible and clinically practical. Through this model, the quantization error of path risk assessment can be controlled within ±8%, significantly improving the reliability of acupuncture stimulation path planning.

[0097] In one embodiment, based on the optimal acupuncture access parameters, the quaternion transformation algorithm is used to calculate the spatial attitude of the instrument, and the needle insertion trajectory parameter group can be obtained, which may include the following steps:

[0098] Step S401, obtain the instrument attitude matrix corresponding to the optimal acupuncture access parameters; the instrument attitude matrix contains the position and direction information of the instrument end in the three-dimensional coordinate system.

[0099] Step S402, extract the quaternion transformation parameters according to the instrument attitude matrix; the quaternion transformation parameters contain the rotation axis vector and the rotation angle component around the axis.

[0100] Step S403: Input the quaternion conversion parameters into the attitude transformation model, and perform a rotation transformation on the instrument coordinate system based on the quaternion multiplication rule to obtain the transformed attitude data.

[0101] Step S404: Optimize the attitude according to the transformed attitude data and the path constraint conditions to obtain the optimized attitude data; the path constraint conditions include the anatomical structure obstacle avoidance threshold and the instrument motion curvature limit.

[0102] Step S405: Calculate the set of needle insertion angle parameters based on the optimized attitude data; the set of needle insertion angle parameters includes the pitch angle deviation value and the azimuth angle correction amount.

[0103] Step S406: Perform a spatial projection match between the set of needle insertion angle parameters and the instrument attitude matrix to obtain the final set of needle insertion trajectory parameters.

[0104] Preferably, obtain the instrument attitude matrix corresponding to the optimal needle insertion path parameters. This matrix includes the position vector and the direction cosine matrix of the instrument end in the three-dimensional coordinate system; extract the rotation axis vector and the rotation angle component around the axis from it through matrix decomposition technology to form the quaternion conversion parameters. Input the parameters into the attitude transformation model, and perform a rotation transformation on the instrument coordinate system based on the quaternion multiplication rule to obtain the initial attitude data. Introduce the anatomical structure obstacle avoidance threshold (such as ≥1.5 mm from the sphenopalatine artery) and the instrument motion curvature limit (such as the minimum bending radius ≥20 mm) as constraint conditions to optimize the initial attitude data, and generate optimized attitude data that meets spatial safety and operation feasibility. Calculate the needle insertion angle parameters such as the pitch angle deviation value and the azimuth angle correction amount based on the optimized data, and perform coordinate matching between the angle parameters and the instrument attitude matrix through spatial projection transformation, and finally form a complete set of needle insertion trajectory parameters including position, direction, and angle.

[0105] In this embodiment, the quaternion transformation and constraint optimization technology are used to improve the instrument control accuracy: the quaternion representation method avoids the Euler angle gimbal lock problem and ensures the continuity and accuracy of the instrument direction control; the dual constraint mechanism of the anatomical obstacle avoidance threshold and the instrument curvature limit can control the distance error between the instrument end and the dangerous structure within 0.3 mm, and at the same time meet the kinematic limitations of different specifications of instruments; the spatial projection matching algorithm realizes the six-degree-of-freedom alignment of the virtual trajectory and the actual anatomical space, making the mapping error between the needle insertion angle parameters and the physical attitude of the instrument ≤1.5°, providing an accurate trajectory data basis for augmented reality guidance. Significantly improve the spatial accuracy and safety of the needle stimulation instrument positioning, and reduce the risk of neurovascular injury caused by attitude calculation errors.

[0106] In one of the embodiments, the deviation correction and update module may include:

[0107] The deviation calculation unit 1041 is used for:

[0108] Step S501: Predict the motion trajectory using the Kalman filter algorithm based on the dynamic guiding trajectory, and calculate the real-time path deviation value.

[0109] The instruction generation unit 1042 is used for:

[0110] Step S502: Adjust the display parameters of the human-computer interaction interface according to the real-time path deviation value, and generate a control instruction set through touch gesture recognition.

[0111] The path generation unit 1043 is used for:

[0112] Step S503: Update the output parameters of the navigation system using the adaptive feedback algorithm according to the control instruction set.

[0113] Step S504: Correct the coordinates of the anatomical landmark points based on the output parameters of the navigation system and the intraoperative image data using the iterative closest point algorithm to obtain the corrected coordinate deviation.

[0114] Step S505: Judge the corrected coordinate deviation based on a preset threshold. If it is lower than the preset threshold, update the individualized anatomical model; otherwise, generate the final needle insertion guiding path.

[0115] The deviation calculation unit initializes the Kalman filter state vector (including position and velocity components) and covariance matrix based on the dynamic guiding trajectory features, generates the predicted trajectory at the next moment through the state transition equation, compares the real-time motion trajectory data to calculate the residual sequence, and quantifies the real-time path deviation value with the root mean square value (RMS). The instruction generation unit automatically adjusts the display parameters such as the trajectory color and zoom ratio of the human-computer interaction interface according to the deviation value, and generates a control instruction set containing parameter adjustment instructions through touch gesture recognition technology (such as swiping and clicking) to realize the doctor's real-time intervention on the navigation system. After receiving the control instruction, the path generation unit updates the output parameters of the navigation system using the adaptive feedback algorithm (such as model predictive control MPC), and at the same time corrects the coordinates of the anatomical landmark points through the iterative closest point (ICP) algorithm based on the intraoperative image data (such as ultrasound and CT), and calculates the corrected coordinate deviation; if the deviation is lower than the preset threshold (such as 0.3mm), trigger the update of the individual anatomical model; otherwise, directly generate the final needle insertion guiding path.

[0116] This deviation correction mechanism enhances the robustness of intraoperative navigation through filtering prediction and adaptive control. The Kalman filtering algorithm effectively suppresses the instrument motion noise, reducing the trajectory prediction error by 60% compared with the traditional smoothing algorithm, and the calculation accuracy of the root mean square deviation reaches the sub-millimeter level; the touch gesture recognition and interface parameter linkage mechanism enables doctors to complete the deviation response within 3 seconds, improving the operation timeliness; the landmark correction technology based on the ICP algorithm uses intraoperative images to calibrate the anatomical model in real time, with the coordinate correction error ≤ 0.2 mm, ensuring the dynamic matching of the navigation system with the patient's actual anatomical structure. It is especially suitable for acupuncture stimulation of deep structures such as the sphenopalatine ganglion with millimeter-level accuracy requirements, effectively reducing the intraoperative operation risk and improving the success rate of acupuncture stimulation.

[0117] In one embodiment, as Figure 3 shown, the Kalman filtering algorithm is used to predict the motion trajectory based on the dynamic guiding trajectory, and the real-time path deviation value is calculated, which may include the following steps:

[0118] Step S601, initialize the state vector and covariance matrix of the Kalman filter based on the dynamic guiding trajectory characteristics.

[0119] Step S602, generate a predicted trajectory according to the state vector and covariance matrix; the predicted trajectory contains a set of coordinate points at the next moment.

[0120] Step S603, obtain the real-time motion trajectory data; calculate the residual sequence between the set of coordinate points and the real-time motion trajectory data.

[0121] Step S604, calculate the real-time path deviation value based on the residual sequence; the real-time path deviation value is the root mean square value of the residual sequence.

[0122] Preferably, the real-time path deviation value can be calculated using the following formula:

[0123]

[0124] where, p RMS represents the real-time path deviation value, n represents the total number of trajectory coordinate points, represents the three-dimensional position vector of the i-th coordinate point in the predicted trajectory, represents the three-dimensional position vector of the i-th coordinate point in the instrument motion trajectory collected in real time.

[0125] This formula has statistical stability, low sensitivity to outliers, can objectively reflect the actual level of trajectory deviation, and provides a quantitative basis for intraoperative dynamic adjustment.

[0126] Step S605, determine whether the real-time path deviation value exceeds the preset deviation threshold. If it exceeds, adjust the covariance matrix parameters of the Kalman filter and re-iterate to obtain the updated real-time path deviation value.

[0127] Specifically, based on the spatial characteristics (such as position, velocity, acceleration) of the dynamic guidance trajectory, the state vector and covariance matrix of the Kalman filter are initialized, and the predicted trajectory at the next moment is generated through the state transition matrix and the process noise matrix. This trajectory contains the set of coordinate points of the instrument movement path. The instrument movement trajectory data is collected in real time, compared point by point with the predicted trajectory coordinate points, and the differences are calculated to form a residual sequence; the residual sequence is quantified into a real-time path deviation value through root mean square (RMS) operation. If the deviation value exceeds the preset threshold (such as 0.3 mm), the covariance matrix parameters are adaptively adjusted (such as increasing the process noise variance), and the state vector and the predicted trajectory are iteratively updated until the deviation value converges to an acceptable range.

[0128] This embodiment improves the trajectory prediction accuracy through state space modeling and iterative optimization mechanism: the recursive framework of the Kalman filter realizes noise suppression of instrument movement (such as respiratory fluctuations, hand tremors); the quantization method of root mean square deviation provides a standardized index for intraoperative error assessment, facilitating doctors to quickly judge the operation accuracy; the adaptive adjustment strategy of the covariance matrix can dynamically balance the weights of model prediction and measured data, ensuring the real-time performance and reliability of the navigation system when the instrument movement state changes suddenly (such as sudden stop, turning). This technology provides sub-millimeter-level trajectory monitoring ability for sphenopalatine ganglion acupuncture stimulation, effectively reducing the risk of nerve injury caused by operation errors and improving the safety and controllability of the acupuncture stimulation process.

[0129] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed 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 executed alternately or alternately with at least a part of other steps or steps in other steps.

[0130] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of a navigation and positioning system for the sphenopalatine ganglion as described above.

[0131] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps in the above method embodiments.

[0132] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial descriptions of the method embodiments. The device embodiments described above are only illustrative. The components described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure solution. A person of ordinary skill in the art can understand and implement it without creative efforts.

[0133] The above embodiments only represent several implementation manners of the embodiments of the present application. The descriptions are relatively specific and detailed, but should not be construed as a limitation on the patent scope of the embodiments of the application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the embodiments of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the embodiments of the present application.

Claims

1. A navigation and positioning system for the sphenopalatine ganglion, characterized in that The system includes: A data processing and modeling module, which is used to obtain a preoperative multimodal image dataset; and is also used to extract sphenopalatine ganglion features from the multimodal image dataset, combine a preset anatomical landmark dataset for spatial topology modeling, and generate an individualized anatomical model; A path planning module, which is used to extract key structure boundary features from the individualized anatomical model, calculate a set of needle insertion paths using a path planning algorithm, and generate a candidate path set; and is also used to determine the safety factor of the candidate path set according to the path length parameter and the risk weight parameter, and determine the optimal needle insertion path parameter; A guiding trajectory generation module, which is used to calculate the instrument spatial attitude using a quaternion transformation algorithm based on the optimal needle insertion path parameter to obtain a set of needle insertion trajectory parameters; and is also used to overlay the guiding path and intraoperative images in real time according to the set of needle insertion trajectory parameters to obtain a dynamic guiding trajectory; A deviation correction and update module, which is used to predict the motion trajectory using a Kalman filtering algorithm based on the dynamic guiding trajectory, calculate the real-time path deviation value; and is also used to adjust the display parameters of the human-computer interaction interface and update the output parameters of the navigation system according to the real-time path deviation value, and generate a final needle insertion guiding path.

2. The system according to claim 1, characterized in that, The extracting sphenopalatine ganglion features from the multimodal image dataset, combining a preset anatomical landmark dataset for spatial topology modeling, and generating an individualized anatomical model includes: Performing image registration on the multimodal image dataset to generate registered three-dimensional voxel data; Segmenting the sphenopalatine ganglion region from the three-dimensional voxel data according to the preset anatomical landmark dataset, and extracting the feature point coordinates; Performing spatial topology matching on the feature point coordinates and the anatomical landmark dataset to obtain a difference mapping matrix; the difference mapping matrix includes the Euclidean distance and the curvature difference weight; Adjusting the topological relationship of the anatomical landmark dataset according to the difference mapping matrix to generate an individualized spatial topology model; Performing mesh parameterization processing on the individualized spatial topology model to generate an editable individualized anatomical model.

3. The system according to claim 1, characterized in that, The path planning module includes: A candidate path planning unit, which is used for: Obtaining the key structure boundary feature data of the individualized anatomical model; the key structure boundary feature data is extracted by an image segmentation algorithm; Generating a three-dimensional spatial topology map according to the key structure boundary feature data; the three-dimensional spatial topology map includes the distribution parameters of blood vessels and nerves; Traversing the feasible path nodes of the three-dimensional spatial topology map using a path planning algorithm to obtain a candidate path set; An optimal path planning unit, which is used for: Calculating the path curvature parameter of the candidate path set; the path curvature parameter is obtained by differential operation of adjacent node coordinates; Obtaining the risk area weight parameter of the candidate path set; the risk area weight parameter is preset based on the anatomical structure type; Calculating a dynamic risk value according to the path curvature parameter and the risk area weight parameter through a weighted summation formula; Wherein, if the dynamic risk value exceeds a preset threshold, adjusting the path smoothness parameter of the candidate path set; Updating the candidate path set according to the adjusted path smoothness parameter to obtain an optimized path set; Traverse the path length parameters of the optimized path set, and select the optimized path with the smallest path length parameter as the optimal needle insertion path parameter; the path length parameter is calculated by accumulating the Euclidean distances between nodes.

4. The system according to claim 3, wherein The dynamic risk value is calculated by the following formula: Among them, R dyn represents the dynamic risk value, n represents the total number of path nodes, k i represents the path curvature of the i-th node, α represents the curvature sensitivity index, α≥1, and S i represents the anatomical structure type where the i-th node is located, μ(S i ) represents the risk weight value based on the fuzzy membership function, c represents the structural risk center value, σ represents the fuzzy domain parameter, l i represents the path element length of the i-th node, ω curv and ω risk respectively represent the weight coefficients of curvature and structural risk, satisfying ω curv +ω risk = 1.

5. The system according to claim 1, characterized in that Based on the optimal needle insertion path parameter, use the quaternion transformation algorithm to calculate the instrument spatial attitude, and obtain a set of needle insertion trajectory parameters, including: Obtain the instrument attitude matrix corresponding to the optimal needle insertion path parameter; the instrument attitude matrix contains the position and direction information of the instrument end in the three-dimensional coordinate system; Extract the quaternion conversion parameters according to the instrument attitude matrix; the quaternion conversion parameters contain the rotation axis vector and the rotation angle component around the axis; Input the quaternion conversion parameters into the attitude transformation model, and perform a rotation transformation on the instrument coordinate system based on the quaternion multiplication rule to obtain the transformed attitude data; Perform attitude optimization according to the transformed attitude data and the path constraint conditions to obtain the optimized attitude data; the path constraint conditions include the anatomical structure obstacle avoidance threshold and the instrument motion curvature limit; Calculate a set of needle insertion angle parameters based on the optimized attitude data; the set of needle insertion angle parameters includes the pitch angle deviation value and the azimuth angle correction amount; Perform spatial projection matching on the set of needle insertion angle parameters and the instrument attitude matrix to obtain the final set of needle insertion trajectory parameters.

6. The system according to claim 1, characterized in that The deviation correction and update module includes: The deviation calculation unit is used for: Based on the dynamic guidance trajectory, use the Kalman filter algorithm to predict the motion trajectory and calculate the real-time path deviation value; The instruction generation unit is used for: Adjust the display parameters of the human-computer interaction interface according to the real-time path deviation value, and generate a set of control instructions through touch gesture recognition; The path generation unit is used for: Update the output parameters of the navigation system using the adaptive feedback algorithm according to the set of control instructions; Based on the output parameters of the navigation system and the intraoperative image data, use the iterative closest point algorithm to correct the coordinates of the anatomical landmark points to obtain the corrected coordinate deviation; Judge the corrected coordinate deviation based on a preset threshold. If it is lower than the preset threshold, update the individual anatomical model; otherwise, generate the final needle insertion guidance path.

7. The system according to claim 6, characterized in that Based on the dynamic guidance trajectory, use the Kalman filter algorithm to predict the motion trajectory and calculate the real-time path deviation value, including: Initialize the state vector and covariance matrix of the Kalman filter based on the characteristics of the dynamic guidance trajectory; Generate a predicted trajectory according to the state vector and covariance matrix; the predicted trajectory contains a set of coordinate points at the next moment; Obtain the real-time motion trajectory data; calculate the residual sequence between the set of coordinate points and the real-time motion trajectory data; Calculate the real-time path deviation value based on the residual sequence; the real-time path deviation value is the root mean square value of the residual sequence; Judge whether the real-time path deviation value exceeds the preset deviation threshold. If it exceeds, adjust the covariance matrix parameters of the Kalman filter and re-iterate to obtain the updated real-time path deviation value.

8. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the system according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the system according to any one of claims 1 to 7.

Images