A simulation system and method for orthopedic surgery navigation control

Abstract

The application discloses a simulation system and method for orthopedic surgery navigation control, and belongs to the technical field of orthopedic surgery navigation simulation. The method acquires a multimodal monitoring data set of a patient's skeletal part, including preoperative medical image data, intraoperative ultrasonic echo signals and micro-vibration signals collected by a skeletal surface strain sensor array; the preoperative medical image data is subjected to three-dimensional reconstruction to generate a bone tissue anatomical structure model, the intraoperative ultrasonic echo signals are subjected to acoustic impedance feature extraction to generate a soft tissue interface feature sequence, and the micro-vibration signals are subjected to multi-scale entropy value calculation to generate a bone tissue stiffness change fingerprint sequence; the skeletal deformation simulation field is subjected to spatial superposition analysis with a surgical instrument navigation path, and skeletal deformation compensation control parameters are output. The method can realize intraoperative skeletal deformation dynamic simulation and navigation path compensation control, and improves the real-time performance and accuracy of orthopedic surgery navigation.

Description

Technical Field

[0001] This invention relates to the field of orthopedic surgical navigation simulation technology, specifically to a simulation system and method for orthopedic surgical navigation control. Background Technology

[0002] In orthopedic surgical navigation and control simulation, existing technologies typically utilize preoperative medical images to reconstruct a three-dimensional model of bone tissue and obtain the relative positions of surgical instruments and bones through intraoperative ultrasound, optical positioning, or electromagnetic positioning. Preoperative images are static data, making it difficult to reflect the real-time deformation of bone tissue under cutting, drilling, or stress during surgery. While intraoperative ultrasound can provide interface echo signals between soft tissue and bone, conventional processing is only used for display or simple registration, failing to establish an elastic coupling relationship between the bone tissue anatomical structure model and soft tissue interface features. Changes in the soft tissue interface position cannot be correlated with the degrees of freedom of the bone mesh nodes. Micro-vibration signals collected by bone surface strain sensor arrays can reflect changes in the local mechanical state of bone tissue, but existing navigation simulation methods do not perform multi-scale entropy quantization processing on these signals, failing to obtain a fingerprint sequence of bone tissue stiffness changes and unable to use stiffness changes as a dynamic input in deformation simulation. Changes in bone tissue stiffness cause soft tissue interface displacement; due to the lack of elastic coupling mapping and dynamic inversion processes, the simulation results fail to reflect intraoperative bone deformation, and the navigation simulation field remains in the state of a preoperative static geometric model. Surgical instrument navigation paths are typically planned based on static models, without spatial overlay analysis of the real-time deformation simulation field and the navigation path. This lack of data makes it impossible to output bone deformation compensation control parameters, causing surgical instruments to continue moving along their original paths after bone deformation, easily deviating from the target position. The fusion processing of the aforementioned multimodal monitoring data to generate bone deformation compensation control parameters represents a technical problem that has not yet been effectively addressed in orthopedic surgical navigation control simulation. Summary of the Invention

[0003] To address the shortcomings of existing orthopedic surgical navigation and control simulations, such as static preoperative images, lack of elastic coupling mapping between bone and soft tissue, and inability to output deformation compensation control parameters, this invention provides a simulation system and method for orthopedic surgical navigation and control. By performing multimodal processing on preoperative medical image data, intraoperative ultrasound echo signals, and micro-vibration signals collected by a bone surface strain sensor array, a bone tissue anatomical structure model, a soft tissue interface feature sequence, and a bone tissue stiffness change fingerprint sequence are established. An elastic coupling mapping relationship is constructed and dynamically inverted to generate an intraoperative bone deformation simulation field. The deformation simulation field is then spatially superimposed and analyzed with the surgical instrument navigation path to output bone deformation compensation control parameters, which are used to correct the surgical instrument navigation path and attitude.

[0004] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides a simulation method for navigation control in orthopedic surgery, comprising the following steps: A multimodal monitoring data set of the patient's skeletal area is acquired, and the navigation path of the surgical instruments is determined. The multimodal monitoring data set includes preoperative medical imaging data, intraoperative ultrasound echo signals, and micro-vibration signals collected by a bone surface strain sensor array. Through multi-source fusion of preoperative images, intraoperative ultrasound, and surface micro-vibration signals, a data foundation is provided for intraoperative bone deformation simulation and navigation compensation.

[0005] The preoperative medical imaging data is processed into three dimensions to generate a bone tissue anatomical structure model, so that the anatomical morphology of the skeleton can be accurately expressed.

[0006] The intraoperative ultrasound echo signal is processed by extracting acoustic impedance features to generate a soft tissue interface feature sequence, which reflects the dynamic positional changes of the soft tissue-bone interface during the operation.

[0007] The micro-vibration signal is processed by multi-scale entropy calculation to generate a fingerprint sequence of bone tissue stiffness changes. The multi-scale entropy value is used to characterize the evolution of bone tissue stiffness as the surgical process progresses.

[0008] An elastic coupling mapping relationship is constructed between the anatomical model of the bone tissue and the feature sequence of the soft tissue interface, thereby expressing the force transmission and displacement coupling between the bone and soft tissue at the model level.

[0009] Based on the elastic coupling mapping relationship, the fingerprint sequence of bone tissue stiffness change is dynamically inverted to generate an intraoperative bone deformation simulation field. The bone deformation distribution adapted to the intraoperative stiffness change is obtained through dynamic inversion.

[0010] The simulation field of bone deformation is spatially superimposed and analyzed with the navigation path of the surgical instrument to output bone deformation compensation control parameters. The deformation field is used to spatially correct the navigation path and improve the accuracy of surgical navigation.

[0011] Preferably, the preoperative medical image data is subjected to three-dimensional reconstruction processing to generate a bone tissue anatomical structure model, including: performing grayscale threshold segmentation processing on the preoperative medical image data to extract bone contour point cloud data; performing Poisson surface reconstruction processing on the bone contour point cloud data to generate a bone tissue surface mesh model; and performing Laplacian smoothing optimization processing on the bone tissue surface mesh model to generate the bone tissue anatomical structure model. This method can improve the completeness of bone contour extraction and the smoothness of the surface model.

[0012] Preferably, the intraoperative ultrasound echo signal is processed to extract acoustic impedance features to generate a soft tissue interface feature sequence, including: performing Hilbert transform on the intraoperative ultrasound echo signal to obtain an instantaneous amplitude envelope and an instantaneous frequency curve; identifying the acoustic impedance interface location between soft tissue and bone based on the local maxima of the instantaneous amplitude envelope; and arranging the acoustic impedance interface locations in chronological order to generate the soft tissue interface feature sequence. This method utilizes the local extrema of the instantaneous amplitude envelope to identify the interface location, improving the noise resistance of soft tissue interface identification.

[0013] Preferably, the micro-vibration signal is processed by multi-scale entropy calculation to generate a fingerprint sequence of bone tissue stiffness changes. This includes: performing empirical wavelet transform decomposition on the micro-vibration signal to obtain intrinsic modal components at multiple scales; calculating the sample entropy value of each intrinsic modal component to obtain a multi-scale entropy vector; and performing sliding window stitching on the multi-scale entropy vector according to the sampling time sequence to generate the fingerprint sequence of bone tissue stiffness changes. Through multi-scale entropy analysis, minute changes in bone tissue stiffness can be stably captured.

[0014] Preferably, constructing the elastic coupling mapping relationship between the bone tissue anatomical structure model and the soft tissue interface feature sequence includes: performing finite element meshing on the bone tissue anatomical structure model to generate a set of bone mesh nodes and a set of soft tissue mesh nodes; extracting interface contact node pairs between the set of bone mesh nodes and the set of soft tissue mesh nodes; constructing a force transmission coefficient matrix based on the interface contact node pairs; and encapsulating the interface contact node pairs and the force transmission coefficient matrix into the elastic coupling mapping relationship. Describing the elastic coupling through interface contact node pairs and the force transmission coefficient matrix makes the mechanical transmission between bone and soft tissue more consistent with the actual contact state.

[0015] Preferably, the dynamic inversion calculation of the bone tissue stiffness change fingerprint sequence based on the elastic coupling mapping relationship to generate an intraoperative bone deformation simulation field includes: mapping the bone tissue stiffness change fingerprint sequence to the surface node positions of the bone tissue anatomical structure model to generate a node stiffness perturbation vector; transferring the node stiffness perturbation vector to the spatial position corresponding to the soft tissue interface feature sequence through the elastic coupling mapping relationship to generate a soft tissue displacement compensation amount; and spatially superimposing the node stiffness perturbation vector and the soft tissue displacement compensation amount to generate the intraoperative bone deformation simulation field. This dynamic inversion method can convert stiffness changes into spatial displacements of bones and soft tissues, improving the accuracy of deformation simulation.

[0016] Further, the step of transferring the nodal stiffness perturbation vector to the spatial location corresponding to the soft tissue interface feature sequence through the elastic coupling mapping relationship to generate the soft tissue displacement compensation includes: performing a transfer operation on the nodal stiffness perturbation vector between the interface contact node pairs included in the elastic coupling mapping relationship to obtain initial values ​​of soft tissue nodal displacement; performing iterative relaxation correction on the mesh distortion error term in the initial values ​​of soft tissue nodal displacement to obtain corrected initial values ​​of soft tissue nodal displacement; and mapping the corrected initial values ​​of soft tissue nodal displacement to the nodal degrees of freedom at the spatial location corresponding to the soft tissue interface feature sequence to generate the soft tissue displacement compensation. Iterative relaxation correction reduces the impact of mesh distortion on the displacement calculation results.

[0017] Preferably, the method involves spatially overlaying the simulated bone deformation field with the surgical instrument navigation path to output bone deformation compensation control parameters. This includes: extracting isosurfaces from the simulated bone deformation field to obtain bone deformation isosurfaces; discretizing the surgical instrument navigation path into a sequence of path nodes; calculating the spatial distance between each node in the path node sequence and the bone deformation isosurface; and converting the spatial distance into the attitude offset of the corresponding path node to generate the bone deformation compensation control parameters. This method quantifies navigation deviation through the spatial distance between the isosurface and the path nodes, facilitating the formation of executable compensation parameters.

[0018] Furthermore, the step of converting the spatial distance into the attitude offset of the corresponding path node and generating the bone deformation compensation control parameters includes: comparing the spatial distance with a preset safety distance threshold to filter out deviation path nodes that exceed the threshold range; performing inverse distance weighted interpolation on the position coordinates of the deviation path nodes to generate a smooth position correction; calculating the attitude rotation correction matrix by combining the position correction and the forward tangent vector of the current path node; and encapsulating the attitude rotation correction matrix and the position correction into a navigation control command frame to generate the bone deformation compensation control parameters. Through smooth correction and attitude correction of deviation nodes, abrupt changes in the navigation path can be avoided, improving the stability of surgical control.

[0019] This invention also provides a simulation system for orthopedic surgical navigation control, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the aforementioned simulation method. This system integrates multimodal data acquisition, bone tissue anatomical structure modeling, soft tissue interface feature extraction, bone tissue stiffness change fingerprint generation, elastic coupling mapping construction, intraoperative bone deformation simulation, and navigation path compensation, providing real-time and accurate deformation compensation control parameters for orthopedic surgical navigation.

[0020] The technical effects and advantages provided by the present invention in the above technical solution are as follows: By performing empirical wavelet transform decomposition on the micro-vibration signals collected by a bone surface strain sensor array and calculating the sample entropy value for each intrinsic modal component, a multi-scale entropy vector is obtained. This vector is then stitched together in the order of sampling time using a sliding window to form a bone tissue stiffness change fingerprint sequence, enabling a continuous quantitative expression of the stiffness changes in bone tissue caused by cutting and stress changes during surgery. A finite element mesh is performed on the bone tissue anatomical model, extracting the interface contact node pairs between the bone mesh node set and the soft tissue mesh node set, and constructing a force transfer coefficient matrix to form an elastic coupling mapping relationship. The bone tissue stiffness change fingerprint sequence is mapped to the surface node positions of the bone tissue anatomical model to generate a node stiffness perturbation vector. This vector is then transferred to the spatial position corresponding to the soft tissue interface feature sequence through the elastic coupling mapping relationship. After iterative relaxation correction of mesh distortion error, a soft tissue displacement compensation amount is generated, which is then superimposed with the node stiffness perturbation vector to generate an intraoperative bone deformation simulation field. Therefore, this deformation simulation field can be dynamically updated with changes in bone stiffness and soft tissue interface displacement, reducing the deformation simulation bias caused by relying solely on static preoperative images, and making the simulation results closer to the intraoperative bone mechanical state. The simulated skeletal deformation field is subjected to isosurface extraction to obtain skeletal deformation isosurfaces. The surgical instrument navigation path is discretized into a sequence of path nodes, and the spatial distance between each path node and the deformation isosurface is calculated. For deviation path nodes whose spatial distance exceeds a preset safety distance threshold, a smooth position correction is generated through inverse distance weighted interpolation. This correction is then combined with the forward tangent vector of the current path node to calculate the attitude rotation correction matrix. The attitude rotation correction matrix and the position correction are encapsulated into a navigation control command frame, outputting skeletal deformation compensation control parameters. This process quantifies the spatial offset caused by skeletal deformation into the position and attitude compensation of the surgical instrument, allowing the navigation control parameters to update with changes in the simulated deformation field. This prevents deviations caused by the surgical instrument continuing to move along a statically planned path, improving the path-following effect of orthopedic surgical navigation under skeletal deformation conditions. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0022] Figure 1 This is a flowchart of a simulation method for navigation control in orthopedic surgery; Figure 2 This is a flowchart of the generation process for soft tissue interface feature sequences and bone tissue stiffness change fingerprint sequences. Figure 3It is a curve showing the change of sample entropy values ​​of multi-scale intrinsic modal components; Figure 4 This is a schematic diagram of the spatial distance and deviation path nodes between the surgical instrument navigation path and the bone deformation isosurface. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0024] See Figure 1 This invention provides a simulation method for navigation control in orthopedic surgery. It involves acquiring a multimodal monitoring data set of the patient's skeletal region and determining the navigation path for surgical instruments. The multimodal monitoring data set includes preoperative medical imaging data, intraoperative ultrasound echo signals, and micro-vibration signals collected by a bone surface strain sensor array. The preoperative medical imaging data undergoes three-dimensional reconstruction processing to generate a bone tissue anatomical model. The intraoperative ultrasound echo signals are processed to extract acoustic impedance features, generating a soft tissue interface feature sequence. The micro-vibration signals are processed to calculate multi-scale entropy values, generating a bone tissue stiffness change fingerprint sequence. An elastic coupling mapping relationship is constructed between the bone tissue anatomical model and the soft tissue interface feature sequence. Based on the elastic coupling mapping relationship, the bone tissue stiffness change fingerprint sequence is dynamically inverted to generate an intraoperative bone deformation simulation field. The bone deformation simulation field is spatially superimposed with the surgical instrument navigation path to output bone deformation compensation control parameters.

[0025] In practice, grayscale thresholding is performed on preoperative medical imaging data to extract bone contour point cloud data. The preoperative medical imaging data consists of a sequence of CT tomographic images. Each frame of the CT tomographic image contains multiple voxels, each carrying a corresponding grayscale value. In the grayscale thresholding process, the lower limit of the grayscale threshold is set to 300 HU, and the upper limit is set to 1500 HU. The lower limit of 300 HU is based on the fact that the Henlein value of bone tissue in CT images is higher than that of soft tissue but lower than the upper limit of dense cortical bone. The upper limit of 1500 HU is based on the CT value range covering both cancellous and cortical bone. All voxels in the preoperative medical imaging data are traversed, and voxels with grayscale values ​​greater than or equal to the lower limit and less than or equal to the upper limit are identified as bone voxels. The three-dimensional spatial coordinates corresponding to these bone voxels are then recorded in the bone contour point cloud data. In some embodiments, duplicate coordinates in the skeletal contour point cloud data are merged, and coordinate points corresponding to isolated outliers are removed to obtain normalized skeletal contour point cloud data.

[0026] A Poisson surface reconstruction process is performed on the skeletal contour point cloud data to generate a bone tissue surface mesh model. In the Poisson surface reconstruction process, the normal vector of each point in the skeletal contour point cloud data is first estimated to obtain the point normal vector field. An octree spatial partition is constructed based on the spatial bounding box of the skeletal contour point cloud data. The maximum depth of the octree is set to 10 layers, which is based on the balance between the mesh refinement corresponding to the spatial resolution of the skeletal contour point cloud data and the computational cost of solving the Poisson equation. Local basis functions are established on each leaf node of the octree, and the point normal vector field is projected onto the function space spanned by the local basis functions to form a discrete representation of the vector field. The Poisson equation is solved, and the source terms of the Poisson equation are composed of the divergence of the discrete representation of the vector field, yielding the indicator function. Isosurfaces are extracted from the indicator function, with an extraction threshold set to 0.5. This threshold is based on the fact that the function value of the indicator function crosses 0.5 when transitioning from the inside to the outside near the location on the bone tissue surface. The extracted isosurfaces are triangulated to generate the bone tissue surface mesh model.

[0027] A Laplacian smoothing optimization process is applied to the bone tissue surface mesh model to generate a bone tissue anatomical structure model. The Laplacian smoothing optimization is achieved by iteratively updating the 3D coordinates of each vertex in the bone tissue surface mesh model. In each iteration, for the r-th vertex, the 3D coordinates of the r-th vertex are updated according to the following formula: in, This represents the three-dimensional coordinate vector of the r-th vertex of the bone tissue surface mesh model at the t-th iteration; This represents the three-dimensional coordinate vector of the r-th vertex of the bone tissue surface mesh model at the (t+1)-th iteration; This represents the set of a ring-near vertex of the r-th bone tissue surface mesh model vertex; Let represent the three-dimensional coordinate vector of the s-th vertex in the set of vertices in the one-ring neighborhood at the t-th iteration; This represents the weight between the r-th bone tissue surface mesh model vertex and the s-th ring neighbor vertex; This represents the Laplacian smoothing intensity coefficient, ranging from 0 to 1. In this embodiment, it is set to 0.6, based on the principle of preserving the subtle anatomical features of the bone tissue surface while suppressing high-frequency noise in the bone tissue surface mesh model. Weight Cotangent weights are used for calculation, determined by the interior angles of the triangle formed by the r-th vertex of the bone tissue surface mesh model and its adjacent vertices. The iteration count is set to 20. Iteration stops when the sum of the displacements of all vertices in the bone tissue surface mesh model is less than 0.001 mm. The 0.001 mm threshold is set because a vertex displacement below this value has no observable impact on subsequent bone tissue anatomical models. The 20 iterations are set as the upper limit of the number of iterations required for vertex displacement convergence in the bone tissue surface mesh model. After iteration stops, the bone tissue anatomical model is obtained.

[0028] In specific implementation, please refer to Figure 2 The intraoperative ultrasound echo signal was processed using Hilbert transform to obtain the instantaneous amplitude envelope and instantaneous frequency curve. During the Hilbert transform process, an analytic signal of the intraoperative ultrasound echo signal was constructed. The magnitude of the analytic signal was extracted to obtain the instantaneous amplitude envelope. The phase of the analytic signal was differentiated in time to obtain the instantaneous frequency curve. The location of the acoustic impedance interface between soft tissue and bone was identified based on the local maxima of the instantaneous amplitude envelope. In practice, the instantaneous amplitude envelope was scanned, and at each sampling moment, the instantaneous amplitude envelope value at the current sampling moment was compared with the instantaneous amplitude envelope values ​​at five sampling moments to the left and right. When the instantaneous amplitude envelope value at the current sampling moment was greater than the instantaneous amplitude envelope values ​​at five sampling moments to the left and right, the current sampling moment was determined as a local maximum of the instantaneous amplitude envelope. The setting of 5 sampling times to the left and right of the current sampling time is based on the range of sampling points corresponding to the main lobe width of the soft tissue-bone acoustic impedance interface reflected echo in the intraoperative ultrasound echo signal. The spatial location corresponding to the local maximum point of the instantaneous amplitude envelope is taken as the location of the soft tissue-bone acoustic impedance interface. The locations of the soft tissue-bone acoustic impedance interface are arranged in chronological order to generate a soft tissue interface feature sequence. In some embodiments, each item in the soft tissue interface feature sequence includes the time index and spatial coordinates of the soft tissue-bone acoustic impedance interface location.

[0029] The micro-vibration signal is decomposed using empirical wavelet transform to obtain intrinsic mode components at multiple scales. In a specific implementation, a Fourier transform is performed on the micro-vibration signal to obtain its spectrum. The spectrum is then segmented, with the segmentation boundaries determined based on the locations of local minima between the local maxima of the spectral amplitude. These boundary divisions divide the spectrum into multiple continuous frequency bands. For each continuous frequency band, an empirical wavelet function and an empirical scaling function are constructed. The empirical wavelet function acts as a bandpass filter, and the empirical scaling function as a low-pass filter. The micro-vibration signal is filtered using the empirical wavelet function and the empirical scaling function to obtain the intrinsic mode components corresponding to each continuous frequency band. The intrinsic mode components corresponding to multiple continuous frequency bands constitute a multi-scale intrinsic mode component set. In some embodiments, when segmenting the spectrum of the micro-vibration signal, the spectral amplitude is normalized to the interval between 0 and 1, and local maxima with spectral amplitudes greater than 0.1 are selected as effective spectral peaks. The selection of local maxima points with spectral amplitudes greater than 0.1 for micro-vibration signals is based on suppressing over-segmentation of the frequency band caused by low-amplitude noise peaks. The frequency position corresponding to the minimum frequency interval between adjacent effective spectral peaks is used as the spectral segmentation boundary.

[0030] The sample entropy value of each intrinsic mode component is calculated to obtain a multi-scale entropy vector. In specific implementation, phase space reconstruction is performed on each intrinsic mode component to obtain multiple template vectors. For the q-th intrinsic mode component, the sample entropy value of the q-th intrinsic mode component is calculated according to the following formula: in, represents the sample entropy value of the q-th intrinsic mode component; m represents the embedding dimension, which is set to 2. The basis for setting m to 2 is the minimum embedding dimension that can preserve local dynamic features in the phase space reconstruction of the intrinsic mode component. Indicates similarity tolerance. Set to the standard deviation of the q-th intrinsic mode component multiplied by 0.2. The standard deviation of the q-th intrinsic mode component is set to be multiplied by 0.2 based on the similarity tolerance ratio that is robust to amplitude fluctuations of the intrinsic mode component; N represents the number of sampling points for the q-th intrinsic mode component; This indicates that among all template vectors corresponding to the q-th intrinsic mode component with embedding dimension m, the distance is less than the similarity tolerance. The number of matching pairs; This indicates that among all template vectors corresponding to the q-th intrinsic mode component with embedding dimension m+1, the distance is less than the similarity tolerance. The number of matching pairs; Let represent the natural logarithm function. For each intrinsic mode component, calculate the sample entropy value of the q-th intrinsic mode component according to the above formula, and combine the sample entropy values ​​of multiple intrinsic mode components into a multi-scale entropy vector.

[0031] The multi-scale entropy vectors are spliced ​​using a sliding window method according to the sampling time sequence to generate a fingerprint sequence of bone tissue stiffness changes. In the specific implementation, the micro-vibration signal is divided into sliding windows with a length of 256 sampling points and a step size of 128 sampling points. The window length of 256 sampling points is based on the fact that the micro-vibration signal sampling rate is 1024 Hz, and the corresponding time length of the window is 0.25 seconds, which can cover the duration of a single vibration event of bone tissue under the action of surgical instruments. The step size of 128 sampling points is based on the fact that there is 50% overlap between adjacent sliding windows, maintaining the continuity of the bone tissue stiffness change fingerprint sequence on the time axis. Within each sliding window, empirical wavelet transform decomposition and sample entropy calculation are performed to obtain the multi-scale entropy vector corresponding to the current sliding window. The multi-scale entropy vector corresponding to the current sliding window is arranged according to the chronological order of the starting sampling time of the current sliding window to generate the fingerprint sequence of bone tissue stiffness changes. In some embodiments, the bone tissue stiffness change fingerprint sequence is a two-dimensional array, where the row index of the two-dimensional array is the sliding window number, the column index of the two-dimensional array is the intrinsic modal component number, and the elements of the two-dimensional array are the sample entropy values ​​corresponding to the sliding window number and the corresponding intrinsic modal component number.

[0032] See Figure 3 The figure shows the curves of sample entropy values ​​for intrinsic modal components 1 to 5 over time in one of the above embodiments. The horizontal axis represents time in seconds, ranging from 0 to 30 seconds; the vertical axis represents the sample entropy value, indicating the complexity characteristics of the corresponding intrinsic modal component. The five curves are distinguished by different colors and line types. The curves as a whole show a linear upward trend over time, reflecting the dynamic changes in bone tissue stiffness during surgery.

[0033] As can be seen from the curve changes, within the time interval of 0 to 15 seconds, the sample entropy values ​​of each intrinsic modal component increase uniformly and steadily. The sample entropy value of intrinsic modal component 5 is the highest, followed by intrinsic modal components 4, 3, 2 and 1, indicating that the intrinsic modal components of different frequency bands exhibit different levels of complexity in the micro-vibration signal.

[0034] At approximately 15 seconds, the sample entropy values ​​of all intrinsic modal components showed a significant jump, with a large increase in magnitude. Subsequently, they continued to maintain a linear upward trend between 15 and 30 seconds, but the overall level was significantly higher than before 15 seconds. This jump reflects a critical time point in the abrupt change in bone tissue stiffness, possibly corresponding to an event where the action of surgical instruments led to a significant change in bone tissue stiffness.

[0035] In the specific implementation, the bone tissue anatomical structure model is meshed using finite element methods (FEM), employing ten-node tetrahedral elements. The maximum element side length is set to 1.2 mm, a decision based on balancing the surface curvature characteristics of the bone tissue anatomical structure model with the number of FEM nodes. Skeletal mesh nodes are generated within the bone tissue anatomical structure model, while soft tissue mesh nodes are generated in the surrounding soft tissue region, forming a set of both bone and soft tissue mesh nodes. Nodes in the bone mesh node set are labeled with the bone type, and nodes in the soft tissue mesh node set are labeled with the soft tissue type. At the interface between bone and soft tissue, bone and soft tissue mesh nodes coincide, and the bone and soft tissue mesh node sets share the coordinates of the interface nodes.

[0036] Extract the interface contact node pairs between the bone mesh node set and the soft tissue mesh node set. At the interface between bone and soft tissue, for each overlapping node, form an interface contact node pair by combining the bone mesh node at that overlapping node position from the bone mesh node set and the soft tissue mesh node at that overlapping node position from the soft tissue mesh node set. Sort all interface contact node pairs in ascending order according to the node numbers at the interface, generating a list of interface contact node pairs.

[0037] A force transfer coefficient matrix is ​​constructed based on interface contact node pairs. For each interface contact node pair in the list, the normal contact stiffness between the skeleton mesh node and the soft tissue mesh node in the interface contact node pair is obtained. The normal contact stiffness is set to... N / mm, normal contact stiffness set to The N / mm value is based on the typical normal contact stiffness between cortical bone and surrounding soft tissue. The effective contact area corresponding to each interface contact node pair is obtained, calculated as a weighted average of the areas of adjacent finite element faces at the interface. The normal contact stiffness is multiplied by the effective contact area to obtain the diagonal element values ​​corresponding to the interface contact node pairs in the force transfer coefficient matrix. The number of rows in the force transfer coefficient matrix equals the number of interface contact node pairs, and the number of columns equals the number of nodes in the bone mesh node set. For off-diagonal elements in the force transfer coefficient matrix, when there is a shared node relationship between a bone mesh node and a soft tissue mesh node, the off-diagonal element is set to the negative of the product of the normal contact stiffness and the adjacent effective contact area; when there is no shared node relationship between a bone mesh node and a soft tissue mesh node, the off-diagonal element is set to 0. The list of interface contact node pairs and the force transfer coefficient matrix are encapsulated into an elastic coupling mapping relationship.

[0038] The bone tissue stiffness variation fingerprint sequence is mapped to the surface node positions of the bone tissue anatomical structure model, generating a node stiffness perturbation vector. The bone tissue stiffness variation fingerprint sequence is a two-dimensional array, where the row index is the sliding window number, the column index is the intrinsic modal component number, and the elements of the two-dimensional array are the sample entropy values ​​of the corresponding sliding window number and the corresponding intrinsic modal component number. For the surface node positions of the bone tissue anatomical structure model, the spatial coordinates of the surface node positions are obtained. The sliding window number with the closest spatial distance to the surface node position of the bone tissue anatomical structure model is found in the sliding window spatial positions corresponding to the bone tissue stiffness variation fingerprint sequence. The sample entropy value of the first intrinsic modal component number in the multi-scale entropy value vector corresponding to the nearest sliding window number is used as the node stiffness perturbation value. When the spatial distance between the surface node position of the bone tissue anatomical structure model and the sliding window spatial position corresponding to the nearest sliding window number is greater than 2.0 mm, the node stiffness perturbation value is set to 0. The setting of a spatial distance greater than 2.0 mm is based on the radius of the spatial coverage area of ​​a single sliding window in the bone tissue stiffness variation fingerprint sequence. The nodal stiffness perturbation values ​​at the surface node locations of all bone tissue anatomical structure models are combined into a nodal stiffness perturbation vector.

[0039] The nodal stiffness perturbation vector is transferred to the spatial location corresponding to the soft tissue interface feature sequence through an elastic coupling mapping relationship, generating the soft tissue displacement compensation. During the transfer process, the nodal stiffness perturbation vector is transferred between the interface contact node pairs included in the elastic coupling mapping relationship to obtain the initial values ​​of the soft tissue nodal displacements. The transfer operation is performed according to the following formula: in, This is the inverse of the soft tissue stiffness matrix, expressed in millimeters per Newton (mm / N). represents the initial displacement vector of soft tissue nodes. Each component in the initial displacement vector of soft tissue nodes corresponds to the initial displacement of a soft tissue mesh node in the soft tissue mesh node set after the transfer operation; T represents the force transfer coefficient matrix. The number of rows in the force transfer coefficient matrix is ​​equal to the number of interface contact node pairs, and the number of columns in the force transfer coefficient matrix is ​​equal to the number of nodes in the skeleton mesh node set. The elements in the force transfer coefficient matrix are determined by the product of the normal contact stiffness and the effective contact area. This represents the nodal stiffness perturbation vector, where each component corresponds to the nodal stiffness perturbation value at the nodal position on the surface of the bone tissue anatomical structure model.

[0040] Iterative relaxation correction is performed on the mesh distortion error term in the initial value of soft tissue nodal displacement. The mesh distortion error term caused by the irregular mesh shape due to interface contact nodes needs correction. In the iterative relaxation correction, the relaxation factor is set to 0.75, based on the balance between iterative convergence speed and numerical stability. In each iteration of relaxation correction, for each component in the initial value vector of soft tissue nodal displacement, a weighted average of the displacement values ​​of adjacent nodes is calculated to obtain the relaxation correction direction. The current component is shifted along the relaxation correction direction by the relaxation factor multiplied by the difference between the current component and the weighted average. In the weighted average of adjacent nodes, the closer the adjacent node is to the current node, the greater the weight; the distance weight is calculated using the inverse squared distance weight. When the maximum correction value between two consecutive iterations of all components of the initial value vector of soft tissue nodal displacement is less than 0.0001 mm, the iterative relaxation correction is stopped, and the corrected initial value of soft tissue nodal displacement is obtained. The 0.0001 mm setting is based on the fact that the mesh distortion error term in the initial value of soft tissue nodal displacement does not have an observable effect on the subsequent soft tissue displacement compensation.

[0041] The corrected initial values ​​of soft tissue node displacements are mapped to the nodal degrees of freedom at the corresponding spatial locations in the soft tissue interface feature sequence to generate soft tissue displacement compensation. Each item in the soft tissue interface feature sequence contains the time index and spatial coordinates of the acoustic impedance interface location between the soft tissue and bone. For each component of the corrected initial values ​​of soft tissue node displacements, the corresponding spatial coordinates of the soft tissue mesh node are obtained. The acoustic impedance interface location between the soft tissue and bone that is spatially closest to the spatial coordinates of the soft tissue mesh node is found in the soft tissue interface feature sequence. The nodal degrees of freedom at the spatial location corresponding to this closest acoustic impedance interface location are used as the mapping target nodal degrees of freedom. The nodal degrees of freedom include x-direction, y-direction, and z-direction displacement degrees of freedom. The corrected initial values ​​of soft tissue node displacements are then assigned to the x-direction, y-direction, and z-direction displacement degrees of freedom to generate soft tissue displacement compensation.

[0042] The nodal stiffness perturbation vector and the soft tissue displacement compensation amount are spatially superimposed to generate an intraoperative bone deformation simulation field. At the surface node locations of the bone tissue anatomical structure model, the corresponding components of the nodal stiffness perturbation vector and the corresponding components of the soft tissue displacement compensation amount are vector-added to obtain the deformed node position vector. The difference between the deformed node position vector and the undeformed node position vector is taken as the deformation displacement vector. Vector addition is performed on all surface nodes of the bone tissue anatomical structure model to obtain the deformation displacement vectors corresponding to all surface nodes. The deformation displacement vectors corresponding to all surface nodes are combined to form the intraoperative bone deformation simulation field.

[0043] In practice, isosurface extraction is performed on the bone deformation simulation field to obtain bone deformation isosurfaces. The bone deformation simulation field consists of deformation displacement vectors corresponding to all surface nodes of the bone tissue anatomical structure model. The isosurface extraction process uses the moving cube algorithm. The magnitude of the deformation displacement vector corresponding to each surface node in the bone deformation simulation field is taken as the isosurface scalar field. The isosurface extraction threshold is set to 0.05 mm, based on the premise that bone deformation displacements smaller than 0.05 mm do not cause deviations that require compensation for the surgical instrument navigation path. The moving cube algorithm extracts isosurfaces with scalar values ​​equal to 0.05 mm on the 3D mesh formed by the surface nodes, generating bone deformation isosurfaces. The moving cube algorithm takes each triangular facet of the surface mesh of the bone tissue anatomical structure model as input, determines the intersection point between the isosurface and the triangular facet based on the relationship between the scalar values ​​at the three vertices of the triangular facet and the isosurface extraction threshold, and connects the intersection points to form triangular facets on the bone deformation isosurface.

[0044] The surgical instrument navigation path is discretized into a sequence of path nodes. The surgical instrument navigation path is a pre-planned three-dimensional spatial curve, represented by continuous trajectory points. Discrete sampling is performed along the surgical instrument navigation path at 0.5 mm arc length intervals to obtain the path node sequence. The 0.5 mm arc length interval is set based on the resolution requirements of the surgical instrument navigation path compensation control; 0.5 mm corresponds to the minimum spatial step size of the orthopedic surgical navigation control command frame. Each path node in the path node sequence contains the path node's position coordinates and its forward tangent vector. The forward tangent vector is calculated based on the difference in position coordinates between adjacent path nodes on the surgical instrument navigation path and then normalized.

[0045] Calculate the spatial distance between each node in the path node sequence and the skeletal deformation isosurface. For each path node in the path node sequence, calculate the shortest Euclidean distance from the path node's position coordinates to all triangular faces on the skeletal deformation isosurface. For each triangular face on the skeletal deformation isosurface, use the nearest point calculation method between a point and a triangular face to obtain the coordinates and distance of the nearest point from the path node's position coordinates to the triangular face. Take the minimum value among all triangular face distances as the spatial distance between the path node and the skeletal deformation isosurface. Construct a spatial bounding box tree for the triangular faces on the skeletal deformation isosurface. The spatial bounding box tree adopts an axis-aligned bounding box hierarchy, which is constructed by recursively partitioning the triangular face set from top to bottom. The spatial bounding box tree accelerates the traversal of triangular faces on the skeletal deformation isosurface, reducing the distance calculation for missed triangular faces.

[0046] Spatial distance is converted into the attitude offset of the corresponding path node to generate bone deformation compensation control parameters. During this conversion, the spatial distance is compared with a preset safety distance threshold, filtering out deviation path nodes that exceed the threshold range. The preset safety distance threshold is set to 1.0 mm, based on the lower limit of the permissible safe clearance between surgical instrument paths and the bone surface during orthopedic surgery. For each path node in the path node sequence, if the spatial distance between the path node and the bone deformation isosurface is greater than the preset safety distance threshold, the path node is marked as a deviation path node. If the spatial distance between the path node and the bone deformation isosurface is less than or equal to the preset safety distance threshold, the path node is marked as a non-deviation path node.

[0047] The position coordinates of the deviation path nodes are processed using inverse distance weighted interpolation to generate a smooth position correction. For each deviation path node, the two path nodes before and after it in the path node sequence are selected as neighboring path nodes, with a total of 4 neighboring path nodes. The setting of two before and after is based on the principle that inverse distance weighted interpolation maintains the local shape of the surgical instrument navigation path while smoothing the position correction. The position correction of the deviation path node is calculated using the following formula: in, represents the position correction amount of the a-th deviation path node; K represents the number of neighboring path nodes participating in the inverse distance weighted interpolation. K is set to 4. The reason for setting K to 4 is that the two neighboring path nodes before and after the deviation path node provide local proximity information in the inverse distance weighted interpolation without excessively smoothing the surgical instrument navigation path. This indicates that the b-th neighboring path node is related to the b-th neighboring path node. The arc length distance between each deviation path node; This represents the original position coordinates of the b-th neighboring path node; b represents the index of the neighboring path node, and the value of b ranges from 1 to k. This represents a summation operation. When the b-th neighboring path node is summed with the b-th neighboring path node... When the arc length distance between each deviation path node is 0 If the value is 0, then the weight of the b-th neighboring path node is set to 1, and the weights of the remaining neighboring path nodes are calculated according to the above formula.

[0048] After performing inverse distance-weighted interpolation on the deviation path nodes, a smoothed position correction is obtained. The smoothed position correction is a vector representing the distance and direction in which the deviation path node needs to move along the spatial direction.

[0049] The attitude rotation correction matrix is ​​calculated by combining the position correction and the forward tangent vector of the current path node. The current path node is the deviation path node; its forward tangent vector is obtained. The cross product of the deviation path node's forward tangent vector and the position correction is performed to obtain the rotation axis vector. The dot product of the deviation path node's forward tangent vector and the position correction is performed to obtain the rotation angle cosine. The rotation angle is calculated based on the rotation axis vector and the rotation angle cosine. The rotation axis vector is normalized to obtain the unit rotation axis vector. The attitude rotation correction matrix is ​​calculated based on the unit rotation axis vector and the rotation angle. The attitude rotation correction matrix is ​​used to adjust the orientation of the surgical instrument at the deviation path node.

[0050] The attitude rotation correction matrix and position correction amount are encapsulated into navigation control command frames to generate skeletal deformation compensation control parameters. The navigation control command frame includes a frame header, path node index, position correction amount, attitude rotation correction matrix, and a verification field. The frame header identifies the start of the navigation control command frame, the path node index identifies the sequence number of the deviation path node in the path node sequence, the position correction amount is a three-dimensional vector, the attitude rotation correction matrix is ​​a 3x3 rotation matrix, and the verification field verifies the integrity of the navigation control command frame. All navigation control command frames corresponding to deviation path nodes are arranged in ascending order of path node index to generate skeletal deformation compensation control parameters. These skeletal deformation compensation control parameters are used to adjust the surgical instrument navigation path to compensate for skeletal deformation.

[0051] See Figure 4 In the figure, the horizontal axis represents the arc length of the path nodes of the surgical instrument navigation path, in millimeters, ranging from 0 mm to 10 mm; the vertical axis represents the spatial distance between the path node and the bone deformation isosurface, in millimeters. The blue solid line is the spatial distance curve, depicting the trend of the distance between the path node and the corresponding bone deformation isosurface along the arc length direction. The orange "×" marks the deviation path node, that is, the path node whose spatial distance exceeds the safe distance threshold. The red dashed line represents the safe distance threshold of 1.0 mm set in this embodiment.

[0052] The spatial distance curves show that the spatial distances at the path nodes generally fluctuate. Two significant troughs appear at the 6th and 8th millimeters in the 4-5 mm segment, where the spatial distance drops to near or below 0.5 mm. This indicates that the surgical instrument navigation path is relatively close to the bone deformation isosurface at this stage, with minimal deviation. The spatial distances in the remaining segments are all above the 1.0 mm threshold, especially in the 8-10 mm segment, where the spatial distance gradually increases to above 3.0 mm, reaching a maximum of nearly 4.0 mm. This indicates significant bone deformation and a large deviation between the surgical instrument navigation path and the bone surface.

[0053] Orange deviation path nodes are mainly distributed in the interval where the spatial distance curve is 1.0 mm above the safe distance threshold, which conforms to the definition of deviation path nodes in this embodiment. Deviation path nodes are densely distributed and numerous in the path arc length range of 4 mm to 10 mm, indicating that bone deformation in this range has a significant impact on the surgical instrument navigation path and requires compensation control. There are fewer deviation path nodes in the path arc length range of 0 to 4 mm, indicating that bone deformation is smaller in this segment and the surgical instrument path accuracy is relatively high.

[0054] In practical implementation, a simulation system for orthopedic surgical navigation control includes a memory, a processor, and a computer program stored in the memory and running on the processor. The memory comprises a program storage area and a data storage area. The program storage area stores the computer program, while the data storage area stores preoperative medical imaging data, intraoperative ultrasound echo signals, micro-vibration signals acquired by a bone surface strain sensor array, a bone tissue anatomical model, soft tissue interface feature sequences, bone tissue stiffness change fingerprint sequences, elastic coupling mapping relationships, intraoperative bone deformation simulation fields, and bone deformation compensation control parameters. The processor reads the computer program instructions from the memory and executes them.

[0055] During computer program execution, the processor acquires a multimodal monitoring dataset of the patient's bone structure. This dataset includes preoperative medical imaging data, intraoperative ultrasound echo signals, and micro-vibration signals collected by a strain sensor array attached to the bone surface. The preoperative medical imaging data is generated by a preoperative CT scan, the intraoperative ultrasound echo signals are acquired by an intraoperative ultrasound probe, and the micro-vibration signals are collected by a strain sensor array attached to the bone surface. The processor maps the preoperative medical imaging data, intraoperative ultrasound echo signals, and micro-vibration signals to the surgical navigation coordinate system. This unified mapping uses a rigid body registration transformation, which is performed according to the following formula: in, This represents a three-dimensional coordinate vector in the surgical navigation coordinate system. It is a column vector with 3 rows and 1 column; Represents the registration rotation matrix. It is a 3x3 orthogonal matrix. The registration rotation matrix is ​​obtained by rigid body registration of the preoperative markers between the original data coordinate system and the surgical navigation coordinate system. This represents a three-dimensional coordinate vector in the original data coordinate system. It is a column vector with 3 rows and 1 column; This represents the registration translation vector. It is a column vector with 3 rows and 1 column. The registration translation vector is obtained by rigid body registration of the preoperative marker points between the original data coordinate system and the surgical navigation coordinate system.

[0056] The registration rotation matrix and registration translation vector are obtained as follows: At least four marker points are selected from the preoperative medical imaging data. The coordinates of these marker points in the original data coordinate system and their corresponding coordinates in the surgical navigation coordinate system are obtained. The least squares rigid body registration algorithm is used to solve for the registration rotation matrix and registration translation vector. The requirement of at least four marker points is based on the fact that at least three non-collinear points are needed for 3D rigid body registration. Increasing the number of marker points reduces registration errors. The processor determines the surgical instrument navigation path based on the registered multimodal monitoring data set. The surgical instrument navigation path is a pre-planned 3D spatial curve.

[0057] The processor performs 3D reconstruction processing on preoperative medical imaging data to generate a bone tissue anatomical model. The 3D reconstruction process includes grayscale thresholding, Poisson surface reconstruction, and Laplacian smoothing optimization. The processor stores the reconstructed bone tissue anatomical model in the data storage area. The processor also performs acoustic impedance feature extraction processing on intraoperative ultrasound echo signals to generate a soft tissue interface feature sequence. Acoustic impedance feature extraction includes Hilbert transform processing, identification of local maxima in the instantaneous amplitude envelope, and temporal sorting of acoustic impedance interface positions. The processor stores the generated soft tissue interface feature sequence in the data storage area. Finally, the processor performs multi-scale entropy calculation processing on micro-vibration signals to generate a bone tissue stiffness change fingerprint sequence. Multi-scale entropy calculation includes empirical wavelet transform decomposition, entropy calculation of intrinsic modal components, and sliding window stitching. The processor stores the generated bone tissue stiffness change fingerprint sequence in the data storage area.

[0058] The processor constructs an elastic coupling mapping relationship between a bone anatomical structure model and a soft tissue interface feature sequence. This construction involves finite element mesh generation of the bone anatomical structure model, producing sets of bone and soft tissue mesh nodes, extracting interface contact node pairs, constructing a force transfer coefficient matrix based on these pairs, and encapsulating the interface contact node pairs and the force transfer coefficient matrix into an elastic coupling mapping relationship. The processor stores this elastic coupling mapping relationship in a data storage area. Based on this elastic coupling mapping relationship, the processor performs dynamic inversion calculations on the bone stiffness change fingerprint sequence to generate an intraoperative bone deformation simulation field. The dynamic inversion calculation involves mapping the bone stiffness change fingerprint sequence to the surface node positions of the bone anatomical structure model to generate node stiffness perturbation vectors. These vectors are then transferred to the corresponding spatial positions of the soft tissue interface feature sequence through the elastic coupling mapping relationship to generate soft tissue displacement compensation. Finally, the node stiffness perturbation vectors and the soft tissue displacement compensation are spatially superimposed to generate the intraoperative bone deformation simulation field. The processor stores this intraoperative bone deformation simulation field in a data storage area.

[0059] The processor performs spatial overlay analysis on the simulated skeletal deformation field and the surgical instrument navigation path, outputting skeletal deformation compensation control parameters. The spatial overlay analysis includes isosurface extraction of the simulated skeletal deformation field to obtain skeletal deformation isosurfaces, discretizing the surgical instrument navigation path into a sequence of path nodes, calculating the spatial distance between each node in the path node sequence and the skeletal deformation isosurfaces, and converting the spatial distances into attitude offsets for the corresponding path nodes to generate skeletal deformation compensation control parameters. The processor outputs these skeletal deformation compensation control parameters to the surgical instrument drive controller via the navigation control interface.

[0060] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A simulation method for navigation control in orthopedic surgery, characterized in that, include: A multimodal monitoring data set of the patient's bone region is acquired, and a surgical instrument navigation path is determined. The multimodal monitoring data set includes preoperative medical imaging data, intraoperative ultrasound echo signals, and micro-vibration signals collected by a bone surface strain sensor array. The preoperative medical imaging data is subjected to three-dimensional reconstruction processing to generate a bone tissue anatomical structure model. The intraoperative ultrasound echo signals are subjected to acoustic impedance feature extraction processing to generate a soft tissue interface feature sequence. The micro-vibration signal is processed by multi-scale entropy calculation to generate a fingerprint sequence of bone tissue stiffness changes; an elastic coupling mapping relationship is constructed between the bone tissue anatomical structure model and the soft tissue interface feature sequence; based on the elastic coupling mapping relationship, the bone tissue stiffness change fingerprint sequence is dynamically inverted to generate an intraoperative bone deformation simulation field; the bone deformation simulation field is spatially superimposed with the surgical instrument navigation path to output bone deformation compensation control parameters.

2. The simulation method for navigation control in orthopedic surgery according to claim 1, characterized in that, The step of performing three-dimensional reconstruction processing on the preoperative medical image data to generate a bone tissue anatomical structure model includes: performing grayscale threshold segmentation processing on the preoperative medical image data to extract bone contour point cloud data; performing Poisson surface reconstruction processing on the bone contour point cloud data to generate a bone tissue surface mesh model; and performing Laplacian smoothing optimization processing on the bone tissue surface mesh model to generate the bone tissue anatomical structure model.

3. The simulation method for navigation control in orthopedic surgery according to claim 1, characterized in that, The step of extracting acoustic impedance features from the intraoperative ultrasound echo signal to generate a soft tissue interface feature sequence includes: performing Hilbert transform on the intraoperative ultrasound echo signal to obtain an instantaneous amplitude envelope and an instantaneous frequency curve; identifying the acoustic impedance interface location between soft tissue and bone based on the local maxima of the instantaneous amplitude envelope; and arranging the acoustic impedance interface locations in chronological order to generate the soft tissue interface feature sequence.

4. The simulation method for navigation control in orthopedic surgery according to claim 1, characterized in that, The step of performing multi-scale entropy calculation on the micro-vibration signal to generate a fingerprint sequence of bone tissue stiffness changes includes: performing empirical wavelet transform decomposition on the micro-vibration signal to obtain intrinsic modal components at multiple scales; calculating the sample entropy value of each intrinsic modal component to obtain a multi-scale entropy vector; and performing sliding window splicing on the multi-scale entropy vector according to the sampling time sequence to generate the fingerprint sequence of bone tissue stiffness changes.

5. The simulation method for navigation control in orthopedic surgery according to claim 1, characterized in that, The process of constructing the elastic coupling mapping relationship between the bone tissue anatomical structure model and the soft tissue interface feature sequence includes: performing finite element mesh subdivision processing on the bone tissue anatomical structure model to generate a set of bone mesh nodes and a set of soft tissue mesh nodes; extracting the interface contact node pairs between the set of bone mesh nodes and the set of soft tissue mesh nodes; constructing a force transmission coefficient matrix based on the interface contact node pairs; and encapsulating the interface contact node pairs and the force transmission coefficient matrix into the elastic coupling mapping relationship.

6. The simulation method for navigation control in orthopedic surgery according to claim 1, characterized in that, The step of dynamically inverting the bone tissue stiffness change fingerprint sequence based on the elastic coupling mapping relationship to generate an intraoperative bone deformation simulation field includes: mapping the bone tissue stiffness change fingerprint sequence to the surface node positions of the bone tissue anatomical structure model to generate a node stiffness perturbation vector; transferring the node stiffness perturbation vector to the spatial position corresponding to the soft tissue interface feature sequence through the elastic coupling mapping relationship to generate a soft tissue displacement compensation amount; and spatially superimposing the node stiffness perturbation vector and the soft tissue displacement compensation amount to generate the intraoperative bone deformation simulation field.

7. The simulation method for navigation control in orthopedic surgery according to claim 1, characterized in that, The step of spatially superimposing the simulated bone deformation field with the navigation path of the surgical instrument and outputting bone deformation compensation control parameters includes: extracting isosurfaces from the simulated bone deformation field to obtain bone deformation isosurfaces; discretizing the navigation path of the surgical instrument into a sequence of path nodes; calculating the spatial distance between each node in the sequence of path nodes and the bone deformation isosurfaces; and converting the spatial distances into the attitude offsets of the corresponding path nodes to generate the bone deformation compensation control parameters.

8. The simulation method for navigation control in orthopedic surgery according to claim 6, characterized in that, The step of transferring the node stiffness perturbation vector to the spatial location corresponding to the soft tissue interface feature sequence through the elastic coupling mapping relationship to generate the soft tissue displacement compensation amount includes: performing a transfer operation on the node stiffness perturbation vector between the interface contact node pairs included in the elastic coupling mapping relationship to obtain the initial value of the soft tissue node displacement; performing iterative relaxation correction processing on the mesh distortion error term in the initial value of the soft tissue node displacement to obtain the corrected initial value of the soft tissue node displacement; and mapping the corrected initial value of the soft tissue node displacement to the node degrees of freedom at the spatial location corresponding to the soft tissue interface feature sequence to generate the soft tissue displacement compensation amount.

9. A simulation method for navigation control in orthopedic surgery according to claim 7, characterized in that, The step of converting the spatial distance into the attitude offset of the corresponding path node and generating the skeleton deformation compensation control parameters includes: comparing the spatial distance with a preset safety distance threshold and filtering out deviation path nodes that exceed the threshold range; performing inverse distance weighted interpolation on the position coordinates of the deviation path nodes to generate a smooth position correction; calculating the attitude rotation correction matrix by combining the position correction and the forward tangent vector of the current path node; and encapsulating the attitude rotation correction matrix and the position correction into a navigation control command frame to generate the skeleton deformation compensation control parameters.

10. A simulation system for navigation control in orthopedic surgery, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the simulation method for orthopedic surgical navigation control as described in any one of claims 1 to 9.

Images