An ultrasound and electrical stimulation dual-modal fusion regional anesthesia puncture path planning method and system

Abstract

The application discloses a regional anesthesia puncture path planning method and system based on ultrasonic and electric stimulation bimodal fusion, and belongs to the technical field of medical image processing, which comprises the following steps: constructing a multi-modal space-time synchronous coordinate system, collecting ultrasonic image data and electric stimulation signal data and performing space-time registration to generate multi-modal comprehensive data; constructing a three-dimensional anatomical-function composite model based on the multi-modal comprehensive data; generating a candidate puncture path in the model by using a path planning algorithm; performing multi-dimensional risk assessment on the candidate path, calculating path risk scores and sorting and screening to output a recommended puncture path; and superimposing the recommended path on a real-time ultrasonic image to monitor the needle tip deviation in real time and perform path correction and prompting during the puncture process. The application realizes deep fusion and collaborative analysis of ultrasonic anatomical images and electric stimulation function signals, and improves the accuracy, safety and operation standardization level of regional anesthesia puncture.

Description

Technical Field

[0001] This invention relates to the field of medical image processing technology, and more specifically to a method and system for regional anesthesia puncture path planning that integrates ultrasound and electrical stimulation in a dual-modal manner. Background Technology

[0002] Regional anesthesia involves injecting local anesthetic around a target nerve to achieve sensory and motor blockade in a specific area of ​​the body. It is widely used in surgical anesthesia and postoperative analgesia. The successful implementation of regional anesthesia highly depends on whether the puncture needle can accurately reach the vicinity of the target nerve and accurately inject the local anesthetic. Therefore, the precision and safety of the puncture guidance technique directly determine the anesthetic effect and patient prognosis.

[0003] Traditional blind puncture relies on surface landmarks and tactile feedback, resulting in limited success rates and high risks. Ultrasound-guided techniques offer real-time visualization, significantly improving safety; however, they only provide anatomical information, cannot assess neurological function, have limitations in deep imaging, and rely on subjective experience for path planning. Electrical stimulation can objectively reflect neural excitability, but currently, ultrasound and electrical stimulation are used as independent tools in parallel, lacking spatiotemporal registration and data fusion, thus failing to achieve collaborative analysis of anatomical and functional information. Existing CT / MRI-ultrasound fusion techniques are primarily geared towards tumor puncture and do not address neurophysiological assessment. In summary, existing technologies suffer from shortcomings such as a lack of anatomical-functional information fusion, a lack of quantitative optimization algorithms for path planning, and the absence of systematic risk assessment and automatic navigation correction.

[0004] Therefore, how to provide a regional anesthesia puncture path planning method and system that integrates ultrasound imaging and electrical stimulation signals is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides a regional anesthesia puncture path planning method and system that integrates ultrasound and electrical stimulation in a dual-modal manner. By constructing a multimodal spatiotemporal synchronous coordinate system, the spatiotemporal registration and fusion of ultrasound images and electrical stimulation signals are realized, a three-dimensional anatomical-functional composite model is established, and based on the model, quantitative planning of the puncture path, risk scoring and real-time navigation deviation correction are performed to achieve the accuracy, safety and standardization of operation of regional anesthesia puncture.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: On one hand, the present invention provides a regional anesthesia puncture path planning method integrating ultrasound and electrical stimulation dual-modal approaches, comprising: A multimodal spatiotemporal synchronous coordinate system is constructed, and ultrasound image data and electrical stimulation signal data of the target area are collected. The ultrasound image data and electrical stimulation signal data are spatiotemporally registered to generate multimodal integrated data. Based on multimodal integrated data, a three-dimensional anatomical-functional composite model of the target region is constructed; In the three-dimensional anatomical-functional composite model, at least one candidate puncture path is generated by using a path planning algorithm based on the preset puncture target point and safety constraints. A risk assessment is performed on the candidate puncture paths, a path risk score is calculated, and multiple candidate puncture paths are sorted and filtered according to the path risk score, and a recommended puncture path is output. The recommended puncture path is displayed visually overlaid on a real-time ultrasound image, and the deviation between the needle tip position and the recommended puncture path is monitored in real time during the puncture process. Path correction prompts are given based on the deviation.

[0007] Preferably, the ultrasound image data and electrical stimulation signal data are spatiotemporally registered to generate multimodal integrated data, specifically including: The ultrasound image data and electrical stimulation signal data are preprocessed to obtain standardized ultrasound image sequences, stimulation response characteristic curves, and nerve excitation conduction time series data. Based on the multimodal spatiotemporal synchronous coordinate system, with the spatial position of the ultrasound probe and the acquisition time as the reference, the stimulus response characteristic curve is timestamped with the ultrasound image frame at the corresponding time. Anatomical structural feature points are extracted from the standardized ultrasound image sequence and functional response region feature points are extracted from the electrical stimulation signal data. Spatial feature matching is performed using a registration network based on the iterative nearest point algorithm. Based on the timestamp alignment results and spatial feature matching results, the electrical stimulation signal data is mapped to the pixel coordinate system of the ultrasound image to generate multimodal integrated data.

[0008] Preferably, based on multimodal integrated data, a three-dimensional anatomical-functional composite model of the target region is constructed, including: Voxelization reconstruction is performed on multimodal integrated data to generate three-dimensional anatomical structure data; The functional response intensity of the electrical stimulation signal is mapped onto the surface of the corresponding anatomical structure in a pseudo-color manner to form a functional response heatmap. By integrating the three-dimensional anatomical structure data and functional response heatmap, a three-dimensional anatomical-functional composite model is constructed.

[0009] Preferably, the multimodal integrated data is voxelized and reconstructed to generate three-dimensional anatomical structure data, specifically including: Multimodal integrated data is mapped to a preset three-dimensional voxel grid, and missing voxels are filled by spatial interpolation to construct a complete voxel density field; Based on the voxel density field, the voxels are classified into anatomical structures and labeled with tissue categories using density thresholding and image segmentation algorithms. The tissue boundaries are then optimized through morphological filtering, and the final output is three-dimensional anatomical structure data with anatomical labels.

[0010] Preferably, the functional response intensity of the electrical stimulation signal is mapped onto the surface of the corresponding anatomical structure in a pseudo-color manner to form a functional response thermogram, including: Extract the outer surface mesh nodes of each tissue from the three-dimensional anatomical structure data and establish an anatomical surface spatial index; The functional response intensity values ​​of the electrical stimulation signals in the multimodal integrated data are correlated to the nearest anatomical surface grid node according to spatial coordinates; For mesh nodes that are not directly matched, inverse distance weighted interpolation or radial basis function interpolation is used to calculate their functional response intensity. The interpolated functional response intensity values ​​are converted into color values ​​according to a preset pseudo-color mapping table, and the color values ​​are assigned to the corresponding anatomical surface mesh nodes. The pseudo-color mapping table can be set to a red to blue gradient, where red represents high response / low threshold and blue represents low response / high threshold. After rendering node by node, a functional response heatmap is generated covering the surface of the anatomical structure.

[0011] Preferably, the three-dimensional anatomical structure data and functional response heatmap are integrated to construct a three-dimensional anatomical-functional composite model, including: Spatial registration is performed between the aforementioned functional response heatmap and the three-dimensional anatomical structure data; For voxel regions labeled as nerve trunks / plexuses in three-dimensional anatomical structure data, the corresponding functional response heatmap color information is overlaid to form composite voxel data that simultaneously contains anatomical structure semantic labels and functional response intensity information. For the neural tissue regions in the complex voxel data, functional excitation threshold parameters and stimulus-response conduction velocity parameters are further labeled based on the electrical stimulation signal data; For non-neural tissue anatomical regions, the original anatomical semantic labels and ultrasound echo density information are preserved, but functional response parameters are not labeled. The final output is a three-dimensional anatomical-functional composite model with a hierarchical information structure.

[0012] Preferably, the probabilistic path planning algorithm (PRM) performs path search in the continuous space of the three-dimensional anatomical-functional composite model, specifically including: With the preset puncture target point as the endpoint and the puncture starting point as the starting point, the safety constraints are encoded as penalty terms of the cost function, including: minimum distance penalty between the path and key anatomical structures, average neural excitation threshold penalty of the area traversed by the path, and path length and curvature smoothness penalty. After iterative search, at least one candidate puncture path that satisfies the constraints is generated.

[0013] Preferably, a risk assessment is performed on the candidate puncture path, and a path risk score is calculated, including: calculating the minimum distance between each candidate path and the key anatomical structure, and assigning a structural collision risk weight based on the distance value; The difficulty coefficient of the operation is calculated by combining the path length, needle insertion angle, and needle path curvature. Based on the nerve distribution density and excitation response intensity in the electrical stimulation signal data, assess the probability of nerve damage risk; The path risk score is calculated by weighting the structural collision risk weight, the operational difficulty coefficient, and the probability of nerve damage risk.

[0014] On the other hand, the present invention also provides a regional anesthesia puncture path planning system that integrates ultrasound and electrical stimulation dual modes, comprising: The data acquisition module is used to construct a multimodal spatiotemporal synchronous coordinate system, acquire ultrasound image data and electrical stimulation signal data of the target area, perform spatiotemporal registration on the ultrasound image data and electrical stimulation signal data, and generate multimodal integrated data. The three-dimensional model construction module is communicatively connected to the multimodal data acquisition and spatiotemporal registration module, and is used to construct a three-dimensional anatomical-functional composite model of the target region based on the multimodal integrated data; The path planning module is communicatively connected to the three-dimensional model construction module and is used to generate at least one candidate puncture path in the three-dimensional anatomical-functional composite model based on the preset puncture target point and safety constraints using a path planning algorithm. The path filtering module is communicatively connected to the path planning module. It is used to perform risk assessment on the candidate puncture paths, calculate the path risk score, sort and filter multiple candidate puncture paths according to the path risk score, and output a recommended puncture path. The real-time navigation module, which is connected in communication with the risk assessment and path selection module, is used to display the recommended puncture path in a visual manner overlaid on the real-time ultrasound image, and to monitor the deviation between the needle tip position and the recommended puncture path in real time during the puncture process, and to provide path correction prompts based on the deviation.

[0015] As can be seen from the above technical solutions, compared with the prior art, this invention discloses a regional anesthesia puncture path planning method and system that integrates ultrasound and electrical stimulation in a dual-modal manner. By constructing a multimodal spatiotemporal synchronous coordinate system, it accurately registers ultrasound images and electrical stimulation signals in spatiotemporal space, establishing a three-dimensional anatomical-functional composite model in regional anesthesia. Simultaneously, it provides detailed tissue anatomical structures and nerve excitation thresholds and conduction velocity parameters, overcoming the limitation of ultrasound alone in assessing nerve function. Furthermore, based on the composite model, this invention uses a path planning algorithm to encode safety constraints into a comprehensive cost function, achieving quantitative generation of the puncture path and eliminating reliance on the operator's subjective experience. A multi-dimensional risk assessment mechanism is established, comprehensively weighting the collision risk, operational difficulty, and nerve injury probability to objectively select the optimal path. Simultaneously, this invention overlays the recommended path onto real-time ultrasound images, monitors the deviation between the needle tip position and the preset path in real-time during the procedure, and automatically generates correction prompts, organically combining preoperative quantitative planning with intraoperative dynamic guidance to effectively reduce the safety risks caused by path deviation. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0017] Figure 1 Flowchart provided for this invention; Figure 2 This is a structural schematic diagram provided for the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] This invention discloses a method for regional anesthesia puncture path planning that integrates ultrasound and electrical stimulation in a dual-modal manner, such as... Figure 1 As shown, it includes: A multimodal spatiotemporal synchronous coordinate system is constructed, and ultrasound image data and electrical stimulation signal data of the target area are acquired. The ultrasound image data and electrical stimulation signal data are spatiotemporally registered to generate multimodal integrated data. Among them, ultrasound image data is acquired by an ultrasound probe in B-mode ultrasound or three-dimensional ultrasound scanning mode; electrical stimulation signal data is acquired by a nerve stimulator and a nerve stimulation needle connected to it. The tip of the nerve stimulation needle is equipped with a conductive electrode to release stimulation current and record tissue impedance and nerve excitation response signals.

[0020] Based on multimodal integrated data, a three-dimensional anatomical-functional composite model of the target region is constructed; In the three-dimensional anatomical-functional composite model, at least one candidate puncture path is generated by using a path planning algorithm based on the preset puncture target point and safety constraints. A risk assessment is performed on the candidate puncture paths, a path risk score is calculated, and multiple candidate puncture paths are sorted and filtered according to the path risk score, and a recommended puncture path is output. The recommended puncture path is displayed visually overlaid on a real-time ultrasound image, and the deviation between the needle tip position and the recommended puncture path is monitored in real time during the puncture process. Path correction prompts are given based on the deviation.

[0021] The method for constructing the multimodal spatiotemporal synchronous coordinate system is as follows: a three-dimensional spatial coordinate system is established with the coordinate origin of the spatial positioning device (such as an electromagnetic positioning instrument or an optical positioning instrument) as the origin of the global coordinate system; a time axis is established with the system clock of the data acquisition system as the unified time reference; the spatial positioning device is fixed on the ultrasonic probe and the puncture needle, and the spatial position and attitude information of the ultrasonic probe and the puncture needle are collected in real time.

[0022] Furthermore, the ultrasound image data and electrical stimulation signal data are spatiotemporally registered to generate multimodal integrated data, specifically including: The ultrasound image data and electrical stimulation signal data are preprocessed to obtain standardized ultrasound image sequences, stimulus response characteristic curves, and neural excitation conduction time-series data. Specifically, the ultrasound image data undergoes noise suppression, image enhancement, and edge extraction to obtain standardized ultrasound image sequences. The electrical stimulation signal data is filtered and denoised, and feature extraction is performed to obtain stimulus response characteristic curves and neural excitation conduction time-series data. Noise suppression in the ultrasound image data preprocessing uses median filtering or anisotropic diffusion filtering. Image enhancement uses histogram equalization or contrast-limited adaptive histogram equalization. Edge extraction uses the Canny edge detection operator. Filtering and denoising in the electrical stimulation signal data preprocessing uses a bandpass filter (passband frequency range of 1Hz~10kHz) or wavelet thresholding. Feature extraction includes extracting the amplitude, latency, and waveform morphology features of the stimulus response signal, as well as the time delay and spatial propagation direction features of neural excitation conduction.

[0023] Based on the multimodal spatiotemporal synchronous coordinate system, with the spatial position of the ultrasound probe and the acquisition time as the reference, the stimulus response characteristic curve is timestamped with the ultrasound image frame at the corresponding time. Anatomical structural feature points are extracted from the standardized ultrasound image sequence and functional response region feature points are extracted from the electrical stimulation signal data. Spatial feature matching is performed using a registration network based on the iterative nearest point algorithm.

[0024] Spatial feature matching is performed using a registration network employing an iterative nearest point algorithm. Specifically, this involves treating the anatomical structure feature point set and the functional response region feature point set as two sets of three-dimensional point clouds, and iteratively solving for the optimal rigid body transformation matrix (rotation matrix R and translation vector t) using the iterative nearest point algorithm to minimize the average distance between the two sets of point clouds. The iteration termination condition of the iterative nearest point algorithm is that the mean square error is less than a preset threshold (e.g., 0.5 mm) or the number of iterations reaches a preset maximum value (e.g., 100 times).

[0025] Based on the timestamp alignment results and spatial feature matching results, the electrical stimulation signal data is mapped to the pixel coordinate system of the ultrasound image to generate multimodal integrated data.

[0026] Furthermore, based on multimodal integrated data, a three-dimensional anatomical-functional composite model of the target region is constructed, including: Voxelization reconstruction is performed on multimodal integrated data to generate three-dimensional anatomical structure data; The functional response intensity of the electrical stimulation signal is mapped onto the surface of the corresponding anatomical structure in a pseudo-color manner to form a functional response heatmap. By integrating the three-dimensional anatomical structure data and functional response heatmap, a three-dimensional anatomical-functional composite model is constructed, in which the neural tissue regions are labeled with functional excitation thresholds and stimulus-response conduction velocity parameters.

[0027] In another embodiment, voxelization reconstruction is performed on the multimodal integrated data to generate three-dimensional anatomical structure data, specifically including: Multimodal synthesis data is mapped onto a pre-defined 3D voxel grid, and missing voxels are filled using spatial interpolation to construct a complete voxel density field; specifically: Based on the spatial range of the target area and the preset voxel resolution, a three-dimensional regular voxel grid with a unified coordinate origin and axis direction is established, and the spatial coordinates (x, y, z) of each data point in the multimodal integrated data are mapped to the corresponding cell index of the voxel grid. For multiple data points mapped to the same voxel unit, the average or weighted average of their ultrasonic echo intensity is taken as the initial density value of the voxel. For empty voxels not mapped to any data points, trilinear interpolation or scatter interpolation based on radial basis functions is used to fill them with the density values ​​of neighboring filled voxels until all voxel units have effective density values, forming a complete voxel density field.

[0028] Based on the voxel density field, the voxels are classified into anatomical structures and labeled with tissue categories using density thresholding and image segmentation algorithms. Specifically, the threshold segmentation method is used to distinguish between high-echo and low-echo regions, and the region growing algorithm or a trained 3D U-Net segmentation network is used to automatically classify voxels into different tissue categories such as skin, fat, muscle, blood vessels, bones, and nerve trunks / plexuses, and assign specific anatomical semantic labels to each category. The tissue boundaries are then optimized through morphological filtering, ultimately outputting 3D anatomical structure data with anatomical labels. Specifically, 3D morphological operations are performed on the segmented and labeled voxel data. First, erosion is performed to remove isolated noisy voxel clusters, then dilation is performed to fill small internal voids and cracks, and Gaussian filtering or median filtering is used to smooth the tissue boundaries. The final output is 3D anatomical structure data with clear anatomical boundaries, no voids, and complete tissue labels.

[0029] Furthermore, the functional response intensity of the electrical stimulation signal is mapped onto the surface of the corresponding anatomical structure in a pseudo-color manner to form a functional response thermogram, including: Extract the outer surface mesh nodes of each tissue from the three-dimensional anatomical structure data and establish an anatomical surface spatial index; The functional response intensity values ​​of the electrical stimulation signals in the multimodal integrated data are correlated to the nearest anatomical surface grid node according to spatial coordinates; For mesh nodes that are not directly matched, inverse distance weighted interpolation or radial basis function interpolation is used to calculate their functional response intensity. The interpolated functional response intensity values ​​are converted into color values ​​according to a preset pseudo-color mapping table, and the color values ​​are assigned to the corresponding anatomical surface mesh nodes. The pseudo-color mapping table can be set to a red to blue gradient, where red represents high response / low threshold and blue represents low response / high threshold. After node-by-node rendering, a functional response heatmap is generated over the surface of the anatomical structure, where the color of each node simultaneously represents the anatomical attribution and functional response intensity at that location.

[0030] Furthermore, by integrating the aforementioned three-dimensional anatomical structure data and functional response heatmaps, a three-dimensional anatomical-functional composite model is constructed, including: Spatially register the functional response heatmap with the three-dimensional anatomical structure data to ensure that each voxel or surface node in the functional response heatmap corresponds one-to-one with the voxel or surface node at the corresponding anatomical location in the three-dimensional anatomical structure data. For voxel regions labeled as nerve trunks / plexuses in three-dimensional anatomical structure data, the corresponding functional response heatmap color information is overlaid to form composite voxel data that simultaneously contains anatomical structure semantic labels and functional response intensity information. For the neural tissue region in the complex voxel data, the functional excitation threshold parameter and the stimulus-response conduction velocity parameter are further labeled according to the electrical stimulation signal data. The functional excitation threshold parameter represents the minimum stimulation intensity that causes the neural tissue to produce an observable response, and the stimulus-response conduction velocity parameter represents the conduction rate of the electrical stimulation signal along the neural tissue. The original anatomical semantic labels and ultrasound echo density information are preserved for non-neural tissue anatomical structural regions, but no functional response parameter annotation is performed. The final output is a three-dimensional anatomical-functional composite model with a hierarchical information structure, which includes: an anatomical structure layer, a functional response thermal layer, and a functional parameter annotation layer.

[0031] In another embodiment, the probabilistic path planning algorithm (PRM) performs path search in the continuous space of the three-dimensional anatomical-functional composite model, specifically including: With the preset puncture target point as the endpoint and the puncture starting point as the starting point, the safety constraints are encoded as penalty terms of the cost function, including: minimum distance penalty between the path and key anatomical structures, average neural excitation threshold penalty of the area traversed by the path, and path length and curvature smoothness penalty. After iterative search, at least one candidate puncture path that satisfies the constraints is generated.

[0032] The Probabilistic Path Map (PRM) algorithm performs path search in the continuous space of the 3D anatomical-functional composite model, and its specific steps include: N nodes (N ranges from 500 to 5000) are randomly sampled within the passable space of the three-dimensional anatomical-functional composite model, and invalid nodes located inside key anatomical structures (blood vessels, bones, nerve trunks / plexuses) are removed. For each valid node, search for neighboring nodes within its neighborhood radius r (r ranges from 1mm to 5mm) and attempt to connect them. If the minimum distance between the connecting line segment and the key anatomical structure is greater than the safe distance threshold (e.g., 2mm), then add the connection to the road map. Using the puncture start point as the starting node and the puncture target point as the ending node, the Dijkstra algorithm is used to search for the shortest path in the constructed road map, and the comprehensive cost of each path is calculated. The top K paths (K ranges from 3 to 10) with a total cost less than a preset threshold are output as candidate puncture paths.

[0033] The expression for the comprehensive cost function is as follows:

[0034] Among them, D min (p) is a negative exponential penalty term for the minimum distance between the path and all key anatomical structures. d min (p) represents the minimum distance between the path and the key anatomical structure, and α is the attenuation coefficient (ranging from 0.5 to 2.0); T avg L(p) is the average neural excitation threshold penalty term for the area traversed by the path; L(p) is the path length penalty term; S(p) is the path curvature smoothness penalty term; w1, w2, w3, and w4 are the weight coefficients of each term, satisfying w1+w2+w3+w4=1 and w1≥0.3.

[0035] In another embodiment, a risk assessment is performed on the candidate puncture path, and a path risk score is calculated, including: calculating the minimum distance between each candidate path and the key anatomical structure, and assigning a structural collision risk weight based on the distance value; The difficulty coefficient of the operation is calculated by combining the path length, needle insertion angle, and needle path curvature. Based on the nerve distribution density and excitation response intensity in the electrical stimulation signal data, assess the probability of nerve damage risk; The path risk score is calculated by weighting the structural collision risk weight, the operational difficulty coefficient, and the probability of nerve damage risk.

[0036] Specifically, the formula for calculating the path risk score is as follows:

[0037] Where R1 is the structural collision risk weight, based on the minimum distance d between the path and the key anatomical structure.min Assigning values ​​according to piecewise functions:

[0038] R2 represents the difficulty level of the operation.

[0039] Among them, L norm To normalize the path length, θ norm To normalize the needle angle deviation, κ norm γ1, γ2, and γ3 are the normalized path curvature and the weighting coefficients. R3 represents the probability of nerve damage risk.

[0040] Where ρ is the proportionality coefficient, f(s) is the neural excitation response intensity at point s on the path, and g(s) is the neural distribution density at point s on the path; β1, β2, and β3 are the weighting coefficients of each term, satisfying... .

[0041] Furthermore, the recommended puncture path is displayed visually overlaid on the real-time ultrasound image. Specifically, this includes: projecting and transforming the three-dimensional spatial coordinates of the recommended puncture path onto the pixel coordinate system of the real-time ultrasound image using the spatial positioning parameters of the ultrasound probe and the ultrasound imaging geometric model; drawing the recommended path on the real-time ultrasound image with colored lines (such as solid green lines), and drawing candidate paths with different colors (such as dashed red lines) for the operator's reference.

[0042] Furthermore, the system monitors the deviation between the needle tip position and the recommended puncture path in real time. Specifically, this includes: acquiring the three-dimensional spatial coordinates of the needle tip in real time using a spatial positioning sensor (such as an electromagnetic positioning sensor or an optical positioning marker) installed on the puncture needle; calculating the Euclidean distance from the needle tip coordinates to the nearest point on the recommended puncture path as the lateral deviation; calculating the difference between the projected position of the needle tip along the recommended puncture path and the expected position as the longitudinal deviation; issuing a path deviation warning when the lateral deviation exceeds a first threshold (e.g., 2 mm), and issuing an emergency stop prompt when the lateral deviation exceeds a second threshold (e.g., 5 mm).

[0043] In a specific embodiment of the present invention, the real-time navigation and deviation correction process is as follows: The three-dimensional spatial coordinates of the recommended puncture path are projected and transformed using the spatial positioning parameters of the ultrasound probe and the ultrasound imaging geometric model, and then drawn on the real-time ultrasound image as colored lines (solid green lines, 2 pixels wide); alternative paths are displayed simultaneously as superimposed lines of different colors (dashed red lines). After the puncture begins, the three-dimensional spatial coordinates of the needle tip are obtained in real time by an electromagnetic positioning sensor installed on the puncture needle, with a sampling frequency of 40Hz; Calculate the Euclidean distance from the needle tip coordinates to the nearest point on the recommended puncture path as the lateral deviation d1; calculate the difference between the projected position of the needle tip along the recommended puncture path and the expected position as the longitudinal deviation d2. When d1 < 1mm, a green circle indicates the needle tip is in the correct position on the display interface; when 1mm ≤ d1 < 2mm, a yellow circle indicates a slight deviation and displays an arrow pointing in the direction of the deviation; when 2mm ≤ d1 < 5mm, an orange circle indicates a moderate deviation and an audible alert is issued, while a correction direction guide is displayed on the image; when d1 ≥ 5mm, a red circle indicates a severe deviation and an emergency stop alert is issued. Deviation correction prompts are displayed as arrows and text on the real-time ultrasound image, guiding the operator to adjust the direction and depth of the puncture needle so that the needle tip returns to the recommended path.

[0044] On the other hand, the present invention also provides a regional anesthesia puncture path planning system that integrates ultrasound and electrical stimulation in a dual-modal manner, such as... Figure 2 As shown, it includes: The data acquisition module is used to construct a multimodal spatiotemporal synchronous coordinate system, acquire ultrasound image data and electrical stimulation signal data of the target area, perform spatiotemporal registration on the ultrasound image data and electrical stimulation signal data, and generate multimodal integrated data. The three-dimensional model construction module is communicatively connected to the multimodal data acquisition and spatiotemporal registration module, and is used to construct a three-dimensional anatomical-functional composite model of the target region based on the multimodal integrated data; The path planning module is communicatively connected to the three-dimensional model construction module and is used to generate at least one candidate puncture path in the three-dimensional anatomical-functional composite model based on the preset puncture target point and safety constraints using a path planning algorithm. The path filtering module is communicatively connected to the path planning module. It is used to perform risk assessment on the candidate puncture paths, calculate the path risk score, sort and filter multiple candidate puncture paths according to the path risk score, and output a recommended puncture path. The real-time navigation module, which is connected in communication with the risk assessment and path selection module, is used to display the recommended puncture path in a visual manner overlaid on the real-time ultrasound image, and to monitor the deviation between the needle tip position and the recommended puncture path in real time during the puncture process, and to provide path correction prompts based on the deviation.

[0045] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0046] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for regional anesthesia puncture path planning that integrates ultrasound and electrical stimulation in a dual-modal manner, characterized in that, include: A multimodal spatiotemporal synchronous coordinate system is constructed, and ultrasound image data and electrical stimulation signal data of the target area are collected. The ultrasound image data and electrical stimulation signal data are spatiotemporally registered to generate multimodal integrated data. Based on multimodal integrated data, a three-dimensional anatomical-functional composite model of the target region is constructed; In the three-dimensional anatomical-functional composite model, at least one candidate puncture path is generated by a path planning algorithm based on the preset puncture target point and safety constraints. A risk assessment is performed on the candidate puncture paths, a path risk score is calculated, and multiple candidate puncture paths are sorted and filtered according to the path risk score, and a recommended puncture path is output. The recommended puncture path is displayed visually overlaid on a real-time ultrasound image, and the deviation between the needle tip position and the recommended puncture path is monitored in real time during the puncture process. Path correction prompts are given based on the deviation.

2. The regional anesthesia puncture path planning method integrating ultrasound and electrical stimulation dual-modal approaches according to claim 1, characterized in that, The ultrasound image data and electrical stimulation signal data are spatiotemporally registered to generate multimodal integrated data, specifically including: The ultrasound image data and electrical stimulation signal data are preprocessed to obtain standardized ultrasound image sequences, stimulation response characteristic curves, and nerve excitation conduction time series data. Based on the multimodal spatiotemporal synchronous coordinate system, with the spatial position of the ultrasound probe and the acquisition time as the reference, the stimulus response characteristic curve is timestamped with the ultrasound image frame at the corresponding time. Anatomical structural feature points are extracted from the standardized ultrasound image sequence and functional response region feature points are extracted from the electrical stimulation signal data. Spatial feature matching is performed using a registration network based on the iterative nearest point algorithm. Based on the timestamp alignment results and spatial feature matching results, the electrical stimulation signal data is mapped to the pixel coordinate system of the ultrasound image to generate multimodal integrated data.

3. The regional anesthesia puncture path planning method integrating ultrasound and electrical stimulation dual-modal approaches according to claim 1, characterized in that, Based on multimodal integrated data, a three-dimensional anatomical-functional composite model of the target region is constructed, including: Voxelization reconstruction is performed on multimodal integrated data to generate three-dimensional anatomical structure data; The functional response intensity of the electrical stimulation signal is mapped onto the surface of the corresponding anatomical structure in a pseudo-color manner to form a functional response heatmap. By integrating the three-dimensional anatomical structure data and functional response heatmap, a three-dimensional anatomical-functional composite model is constructed.

4. The regional anesthesia puncture path planning method integrating ultrasound and electrical stimulation dual-modal approaches according to claim 3, characterized in that, Voxelization reconstruction of multimodal integrated data generates three-dimensional anatomical structure data, specifically including: Multimodal integrated data is mapped to a preset three-dimensional voxel grid, and missing voxels are filled by spatial interpolation to construct a complete voxel density field; Based on the voxel density field, the voxels are classified into anatomical structures and labeled with tissue categories using density thresholding and image segmentation algorithms. The tissue boundaries are then optimized through morphological filtering, and the final output is three-dimensional anatomical structure data with anatomical labels.

5. The regional anesthesia puncture path planning method integrating ultrasound and electrical stimulation dual-modal approaches according to claim 3, characterized in that, The functional response intensity of the electrical stimulation signal is mapped onto the surface of the corresponding anatomical structure in a pseudo-color manner to form a functional response heatmap, including: Extract the outer surface mesh nodes of each tissue from the three-dimensional anatomical structure data and establish an anatomical surface spatial index; The functional response intensity values ​​of the electrical stimulation signals in the multimodal integrated data are correlated to the nearest anatomical surface grid node according to spatial coordinates; For mesh nodes that are not directly matched, inverse distance weighted interpolation or radial basis function interpolation is used to calculate their functional response intensity. The interpolated functional response intensity values ​​are converted into color values ​​according to a preset pseudo-color mapping table, and the color values ​​are assigned to the corresponding anatomical surface mesh nodes. The pseudo-color mapping table can be set to a red to blue gradient, where red represents high response / low threshold and blue represents low response / high threshold. After rendering node by node, a functional response heatmap is generated covering the surface of the anatomical structure.

6. The regional anesthesia puncture path planning method integrating ultrasound and electrical stimulation dual-modal approaches according to claim 3, characterized in that, By integrating the aforementioned three-dimensional anatomical structure data and functional response heatmaps, a three-dimensional anatomical-functional composite model is constructed, including: Spatial registration is performed between the aforementioned functional response heatmap and the three-dimensional anatomical structure data; For voxel regions labeled as nerve trunks / plexuses in three-dimensional anatomical structure data, the corresponding functional response heatmap color information is overlaid to form composite voxel data that simultaneously contains anatomical structure semantic labels and functional response intensity information. For the neural tissue regions in the complex voxel data, functional excitation threshold parameters and stimulus-response conduction velocity parameters are further labeled based on the electrical stimulation signal data; For non-neural tissue anatomical regions, the original anatomical semantic labels and ultrasound echo density information are preserved, but functional response parameters are not labeled. The final output is a three-dimensional anatomical-functional composite model with a hierarchical information structure.

7. The regional anesthesia puncture path planning method integrating ultrasound and electrical stimulation dual-modal approaches according to claim 1, characterized in that, The probabilistic path planning algorithm (PRM) performs path search in the continuous space of the three-dimensional anatomical-functional composite model, specifically including: With the preset puncture target point as the endpoint and the puncture starting point as the starting point, the safety constraints are encoded as penalty terms of the cost function, including: minimum distance penalty between the path and key anatomical structures, average neural excitation threshold penalty of the area traversed by the path, and path length and curvature smoothness penalty. After iterative search, at least one candidate puncture path that satisfies the constraints is generated.

8. The regional anesthesia puncture path planning method according to claim 1, characterized in that, A risk assessment is performed on the candidate puncture paths, and a path risk score is calculated, including: Calculate the minimum distance between each candidate path and the key anatomical structure, and assign a collision risk weight to the structure based on the distance value; The difficulty coefficient of the operation is calculated by combining the path length, needle insertion angle, and needle path curvature. Based on the nerve distribution density and excitation response intensity in the electrical stimulation signal data, assess the probability of nerve damage risk; The path risk score is calculated by weighting the structural collision risk weight, the operational difficulty coefficient, and the probability of nerve damage risk.

9. A regional anesthesia puncture path planning system integrating ultrasound and electrical stimulation dual modes, characterized in that, include: The data acquisition module is used to construct a multimodal spatiotemporal synchronous coordinate system, acquire ultrasound image data and electrical stimulation signal data of the target area, perform spatiotemporal registration on the ultrasound image data and electrical stimulation signal data, and generate multimodal integrated data. The three-dimensional model construction module is communicatively connected to the multimodal data acquisition and spatiotemporal registration module, and is used to construct a three-dimensional anatomical-functional composite model of the target region based on the multimodal integrated data; The path planning module is communicatively connected to the three-dimensional model construction module and is used to generate at least one candidate puncture path in the three-dimensional anatomical-functional composite model based on the preset puncture target point and safety constraints using a path planning algorithm. The path filtering module is communicatively connected to the path planning module. It is used to perform risk assessment on the candidate puncture paths, calculate the path risk score, sort and filter multiple candidate puncture paths according to the path risk score, and output a recommended puncture path. The real-time navigation module, which is connected in communication with the risk assessment and path selection module, is used to display the recommended puncture path in a visual manner overlaid on the real-time ultrasound image, and to monitor the deviation between the needle tip position and the recommended puncture path in real time during the puncture process, and to provide path correction prompts based on the deviation.

Images