Multi-mode image navigation system for gout minimally invasive treatment

Abstract

The invention relates to the technical field of medical image navigation, in particular to a gout minimally invasive treatment multi-modal image navigation system which comprises a data acquisition module, a biomechanical fusion modeling module, a dynamic path planning module and a multi-physical field execution control module. The data acquisition module realizes synchronous acquisition of crystal-structure stress gradient, surface energy density and viscoelasticity parameters of the synovial fluid; the fusion modeling module generates a three-dimensional crystallization adhesion risk coefficient and fluid resistance prediction matrix based on a finite element and fluid-solid coupling algorithm; the path planning module dynamically calculates motion compensation and suction parameters in combination with real-time instrument poses, and outputs a six-degree-of-freedom navigation instruction set; and the execution control module drives the magnetic navigation puncture needle, the self-adaptive suction device and the laser ablation unit to realize cooperative execution of pose regulation and control, flow regulation and focus form switching. According to the method, gout crystallization accurate identification, path risk-avoiding navigation and targeted removal are realized, and the safety and the removal rate of minimally invasive intervention are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of medical image navigation technology, and in particular to a multimodal image navigation system for minimally invasive treatment of gout. Background Technology

[0002] With the widespread application of minimally invasive interventional techniques in gout treatment, clinical practice has placed higher demands on the precise identification, pathway avoidance, and local removal of uric acid crystals. While imaging technologies such as ultrasound, CT, and MRI provide some support for intraoperative localization, their limitations in information dimension and resolution across single modalities make it difficult to fully represent the coupling relationship between crystals and surrounding tissues in terms of mechanical, energy, and fluid properties. In recent years, techniques such as magnetically guided puncture, laser ablation, and aspiration removal have been gradually introduced into gout intervention procedures. However, how to systematically couple image recognition, pathway planning, and the execution control system remains a key challenge in achieving "intelligent and precise intervention."

[0003] The existing technology has obvious bottlenecks in the following aspects, which directly limit the safety and treatment integrity of intraoperative navigation: (1) The acquisition process relies on a single image modality and lacks comprehensive perception of crystal stress field, surface energy and viscoelastic properties of synovial fluid, resulting in one-sided and unstable intraoperative model construction; (2) The path planning process lacks a biomechanical driving mechanism and cannot dynamically respond to local adhesion risk and fluid resistance changes, which can easily cause the instrument to enter a high-risk area; (3) The execution control process is scattered and lacks a multi-physics field collaborative execution strategy based on multimodal fusion data, which cannot achieve the spatiotemporal unity of puncture, aspiration and ablation actions. Summary of the Invention

[0004] This invention provides a multimodal image navigation system for minimally invasive gout treatment. It proposes a three-dimensional biomechanical fusion model driven by multimodal in situ imaging, and uses this model as the core to connect navigation planning and control execution into a holistic solution, which systematically improves the perception intelligence, path safety and treatment closed-loop capability of minimally invasive gout intervention.

[0005] A multimodal image-guided system for minimally invasive gout treatment includes a data acquisition module, a biomechanical fusion modeling module, a dynamic path planning module, and a multiphysics execution control module, among which: The data acquisition module integrates a dual-frequency ultrasound probe, an energy-dispersive CT scanner, and a microscopic optical coherence tomography device to simultaneously acquire stress gradient data at the crystallization-tissue interface, surface energy distribution data of crystallization, and viscoelastic parameters of synovial fluid. The biomechanical fusion modeling module receives the stress gradient data of the crystallization-tissue interface, the crystallization surface energy distribution data, and the viscoelastic parameters of the synovial fluid output by the data acquisition module, and generates a three-dimensional biomechanical fusion model through the crystallization surface energy compensation algorithm. The three-dimensional biomechanical fusion model includes the crystallization adhesion risk coefficient and the fluid resistance prediction matrix. The dynamic path planning module receives the biomechanical fusion model generated by the biomechanical fusion modeling module, and calculates the following parameters by combining it with real-time device pose data: Device motion compensation vector based on adhesion risk coefficient; Optimization parameters for the suction path based on the fluid resistance matrix; Output a dynamic navigation instruction set, which includes instrument pose correction amount and suction flow control curve; The multiphysics execution control module receives and outputs a set of dynamic navigation instructions, synchronously driving the magnetic navigation puncture needle, the adaptive negative pressure aspiration device, and the laser ablation unit, wherein: Magnetic navigation puncture needle: Six degrees of freedom motion compensation based on instrument position correction; Adaptive negative pressure suction device: dynamically adjusts suction power according to the suction flow control curve; Laser ablation unit: Adjusts the shape of the ablation focus (ellipsoid / columnar / multifocal) based on the surface energy distribution data of the crystal.

[0006] Optionally, the data acquisition module includes: Initial configuration: The dual-frequency ultrasound probe is set to alternate between high-frequency mode (20MHz) and low-frequency mode (5MHz). The high-frequency mode is used for elastic imaging to acquire stress gradient data of the crystal-tissue interface, and the low-frequency mode is used to penetrate deep tissues. The energy dispersive CT scanner is configured with a tube voltage of 80-140kVp and a multi-energy spectral separation mode to generate energy dissipation curves of energy distribution data on the crystal surface. The micro-optical coherence tomography device was set to a center wavelength of 1300 nm and an axial resolution of 5 μm for dynamic tomographic scanning of the viscoelastic parameters of synovial fluid. Synchronous data acquisition: The dual-frequency ultrasound probe, energy spectrum CT scanner, and micro-optical coherence tomography device are synchronously activated by hardware trigger signals to ensure that the three acquire data under the same time reference. A radio frequency signal sequence output by a dual-frequency ultrasound probe to provide stress gradient data at the crystallization-tissue interface; The projected energy spectrum matrix of the energy distribution data on the crystal surface output by the energy spectroscopy CT scanner; The microscopic optical coherence tomography device outputs a three-dimensional interference signal of the viscoelastic parameters of the synovial fluid; Extraction and fusion of stress gradient data at the crystallization-tissue interface: The radio frequency signal sequence output by the dual-frequency ultrasound probe is input into the elastic inversion algorithm to generate a two-dimensional elastic map of the stress gradient data at the crystallization-tissue interface; The two-dimensional elastic map is converted into a three-dimensional stress gradient distribution tensor aligned with the coordinate system of the energy spectrum CT scanner using a spatial interpolation algorithm.

[0007] Reconstruction and calibration of surface energy distribution data of crystals: The projected energy spectrum matrix output by the energy spectrum CT scanner is input into the base material decomposition algorithm to separate the energy spectrum characteristics of uric acid crystals from those of the surrounding tissue; Based on the decomposed energy spectrum characteristics, an energy density cloud map of the crystal surface energy distribution data is generated through a surface energy calculation model. The refractive index of the energy density cloud map was calibrated using the viscoelastic parameters of the synovial fluid output by the micro-optical coherence tomography device, and the refractive compensation data of the crystal surface energy distribution were obtained. Multimodal registration of synovial fluid viscoelastic parameters: The three-dimensional interference signal output from the microscopic optical coherence tomography device is input into the viscoelastic deconvolution algorithm to generate the frequency-strain rate response curve of the synovial fluid viscoelastic parameters; The frequency-strain rate response curve is spatiotemporally registered with the three-dimensional stress gradient distribution tensor and the refractive compensated crystal surface energy distribution data to form a three-dimensional viscoelastic field of synovial fluid viscoelastic parameters. Data fusion and output: The three-dimensional stress gradient distribution tensor, the refractive compensated crystal surface energy distribution data, and the three-dimensional viscoelastic field are fused in multiple modes to generate a synchronously acquired data package.

[0008] Optionally, the data acquisition module further includes: Quality verification and anomaly marking of synchronous acquisition data packets: The generated synchronous acquisition data packets are input into the data quality assessment algorithm, which calculates the quality score of the synchronous acquisition data packets based on the signal-to-noise ratio of the three-dimensional stress gradient distribution tensor, the standard deviation of the energy density of the crystal surface energy distribution data after refraction compensation, and the spatial continuity parameters of the three-dimensional viscoelastic field. If the quality score is lower than the preset threshold, it is marked as an abnormal data packet and an abnormal type label is output. The abnormal type label includes stress gradient distortion, surface energy refraction shift or viscoelastic field fracture. Dynamic parameter reconfiguration based on anomaly type labels: Matching parameter adjustment rules for dual-frequency ultrasound probes, energy dispersive CT scanners, and microscopic optical coherence tomography devices according to anomaly type labels. If the label indicates stress gradient distortion, adjust the high-frequency mode transmission power of the dual-frequency ultrasound probe to 150% of the reference value and shorten the low-frequency mode pulse interval to 10μs. If the label is surface energy refraction offset, then reload the uric acid crystal energy spectrum feature library in the base material decomposition algorithm of the energy spectrum CT scanner and start the refractive index dynamic compensation mode of the micro-optical coherence tomography device. If the label indicates viscoelastic field fracture, increase the scanning frame rate of the micro-optical coherence tomography device to 200 Hz and expand the sampling bandwidth of the frequency-strain rate response curve of the synovial fluid viscoelastic parameters. Local reacquisition and replacement of abnormal data packets: Based on the adjusted parameters of the dual-frequency ultrasound probe, energy dispersive CT scanner, and micro-optical coherence tomography (EDT), local reacquisition is initiated only for the anatomical region corresponding to the abnormality type label. The dual-frequency ultrasonic probe outputs a local radio frequency signal sequence to replace the abnormal data block in the original three-dimensional stress gradient distribution tensor. The energy spectrum CT scanner outputs a local projection energy spectrum matrix, which replaces the abnormal energy spectrum region in the original refractive compensated crystal surface energy distribution data. The micro-optical coherence tomography device outputs a local three-dimensional interference signal, which replaces the fracture viscoelastic parameters in the original three-dimensional viscoelastic field. Fusion verification of the corrected synchronous acquisition data packet: The three-dimensional stress gradient distribution tensor after local reacquisition, the crystal surface energy distribution data after refraction compensation, and the three-dimensional viscoelastic field are fused in a second multi-modal manner to generate the corrected synchronous acquisition data packet. A data quality assessment algorithm is executed on the corrected synchronously acquired data packets. If the quality score reaches the preset threshold, a verification pass signal is output; otherwise, an iterative correction loop of dynamic parameter reconfiguration → local reacquisition → fusion verification is triggered until the verification passes. Dynamic data loading and biomechanical model update: The validated and corrected data packets are synchronously collected and sent to the biomechanical fusion modeling module.

[0009] Optionally, the biomechanical fusion modeling module includes: Preprocessing and feature extraction of synchronous acquisition data packets: Receive synchronous acquisition data packets output by the data acquisition module, which include three-dimensional stress gradient distribution tensor, refractive compensated crystal surface energy distribution data and three-dimensional viscoelastic field; Anisotropic noise reduction filtering is performed on the three-dimensional stress gradient distribution tensor to generate a noise-reduced three-dimensional stress gradient distribution tensor. Energy density thresholding was performed on the refractive-compensated crystal surface energy distribution data to extract energy densities greater than 0.8 J / m². 2 High-density regions of crystalline surface energy; Fluid-structure interaction registration was performed on the three-dimensional viscoelastic field to generate the displacement coupling matrix between the synovial fluid viscoelastic parameters and the tissue interface; Finite element modeling of crystal adhesion risk coefficient: The denoised three-dimensional stress gradient distribution tensor is input into the crystal-tissue contact mechanics model, and the contact stress distribution at the crystal-soft tissue interface is calculated based on Hertz contact theory. By superimposing the high-density region of crystalline surface energy with the contact stress distribution, a three-dimensional probability distribution map of the initial crystalline adhesion risk coefficient is generated using the energy correction term formula in the crystalline surface energy compensation algorithm.

[0010] Optionally, the biomechanical fusion modeling module further includes: Fluid-structure interaction calculation of fluid resistance prediction matrix: Input the synovial fluid viscoelastic parameters and the displacement coupling matrix of the tissue interface into the Navier-Stokes equation, and combine the noise-reduced three-dimensional stress gradient distribution tensor to calculate the flow velocity field and shear stress field of synovial fluid on the crystallized surface. The initial weight distribution of the fluid resistance prediction matrix is ​​constructed based on the spatial gradient of the velocity field and the shear stress field. A three-dimensional dynamic response model of the fluid resistance prediction matrix is ​​generated by dynamically correcting the attenuation factor of the fluid resistance prediction matrix using viscoelastic parameters. Iterative calibration of the three-dimensional biomechanical fusion model: The three-dimensional probability distribution map of the initial crystallization adhesion risk coefficient is coupled and iterated with the three-dimensional dynamic response model of the fluid resistance prediction matrix. Within the high-density region of crystalline surface energy, the energy density difference of the crystalline surface energy distribution data after refraction compensation is used as a benchmark to calculate the crystalline adhesion risk coefficient. If the energy density difference is greater than 0.5 J / m 2 Then adjust the surface energy compensation coefficient in the crystallization surface energy compensation algorithm until the energy density difference is less than or equal to 0.5 J / m³. 2 ; The calibrated crystal adhesion risk coefficient and fluid resistance prediction matrix are spatially weighted and fused to generate an initial version of the three-dimensional biomechanical fusion model; Model output and dynamic path planning interface: The initial version of the 3D biomechanical fusion model is sent to the dynamic path planning module as the input source for the device motion compensation vector and suction path optimization parameters; The three-dimensional biomechanical fusion model includes a three-dimensional probability distribution map of the final crystallization adhesion risk coefficient and a three-dimensional dynamic response model of the fluid resistance prediction matrix.

[0011] Optionally, the dynamic path planning module includes: Registration of the three-dimensional biomechanical fusion model with real-time instrument pose data: receiving the three-dimensional biomechanical fusion model output by the biomechanical fusion modeling module, wherein the three-dimensional biomechanical fusion model includes a three-dimensional probability distribution map of the crystal adhesion risk coefficient and a three-dimensional dynamic response model of the fluid resistance prediction matrix; The three-dimensional probability distribution map of the crystal adhesion risk coefficient is spatially registered with the real-time instrument pose data (including the coordinates of the puncture needle tip, the posture angle, and the movement speed) to generate an instrument tip safety margin index. The safety margin index is defined as the inverse product of the crystal adhesion risk coefficient and the distance to the instrument tip. Dynamic calculation of instrument motion compensation vector: Based on the three-dimensional dynamic response model of the instrument tip safety margin index and fluid resistance prediction matrix, the following parallel calculations are performed: Adhesion risk avoidance path planning: In the region where the crystal adhesion risk coefficient is >0.7, the gradient descent method is used to calculate the motion compensation vector of the device to avoid the high-risk region and output the six-degree-of-freedom compensation amount; Fluid resistance optimization path generation: Extract fluid blockage areas with resistance values ​​> 1.5 kPa from the fluid resistance prediction matrix, and generate suction path optimization parameters through the minimum resistance path algorithm; Synthesis and verification of dynamic navigation instruction sets: Inputting six-degree-of-freedom compensation quantities and suction path optimization parameters into the instruction synthesis engine: The device pose correction amount is generated based on the six-degree-of-freedom compensation amount, which includes a linear displacement vector and rotational Euler angles; Based on the optimal suction angle and critical suction flow rate, combined with the real-time updated data of the fluid resistance prediction matrix, a cubic spline function of the suction flow rate control curve is generated. By integrating the instrument pose correction amount and the suction flow control curve, an initial version of the dynamic navigation instruction set is formed.

[0012] Optionally, the dynamic path planning module further includes: Real-time prediction of instrument-crystallization collision risk: Input the initial version of the dynamic navigation instruction set into the collision prediction model: Based on the three-dimensional probability distribution diagram of the crystal adhesion risk coefficient, the motion trajectory of the puncture needle after the instrument pose correction is simulated. If the trajectory of the motion intersects with a region where the crystal adhesion risk coefficient is greater than 0.9, it is marked as a high-risk collision zone, and the collision risk level is output. The collision risk level is divided into low, medium and high. Iterative correction of collision avoidance instructions: If the collision risk level is high, path replanning is triggered, including: In the dynamic calculation of the motion compensation vector of the device, a collision constraint condition is added to the adhesion risk avoidance path planning: the threshold of the crystal adhesion risk coefficient is increased from 0.7 to 0.6; Recalculate the motion compensation vector of the device to generate updated six-degree-of-freedom compensation quantities; The updated six-degree-of-freedom compensation values ​​are input into the synthesis and verification of the dynamic navigation instruction set and the real-time prediction of the instrument-crystallization collision risk for secondary verification until the collision risk level is reduced to medium or low. Output and execution interface of dynamic navigation instruction set: The verified dynamic navigation instruction set is sent to the multiphysics execution control module. The dynamic navigation instruction set includes the final instrument pose correction amount and the final suction flow control curve.

[0013] Optionally, the multiphysics execution control module includes: Parsing and execution allocation of dynamic navigation instruction set: Receives the dynamic navigation instruction set output by the dynamic path planning module, the dynamic navigation instruction set including the final instrument pose correction amount and the final suction flow control curve; The final instrument pose correction amount is allocated to the magnetic navigation puncture needle control channel to generate a six-degree-of-freedom motion compensation command for the magnetic navigation puncture needle. The final suction flow control curve is assigned to the control channel of the adaptive negative pressure suction device to generate a suction power modulation waveform. Simultaneously, the surface energy distribution data of the crystallized surface (from the data acquisition module) is extracted and allocated to the control channel of the laser ablation unit; Six-degree-of-freedom motion compensation execution of the magnetic navigation puncture needle: The magnetic navigation puncture needle is driven according to six-degree-of-freedom motion compensation commands, including: The linear displacement vector is analyzed, and a displacement magnetic field is generated by the triaxial gradient coil of the magnetic navigation system to drive the tip of the puncture needle to move with an accuracy of 0.1 mm / ms. The Euler angles of rotation are analyzed, and the puncture needle attitude angle is adjusted by a rotating torque magnetic field generator, with an angular resolution of 0.5°. Real-time feedback of the actual position data of the puncture needle to the dynamic path planning module forms a closed-loop position control; Dynamic flow control of the adaptive negative pressure suction device: The adaptive negative pressure suction device is driven according to the suction power modulation waveform, including: The final suction flow control curve is converted into a pulse width modulation duty cycle sequence to control the motor speed of the negative pressure pump; Based on real-time updated data from the fluid resistance prediction matrix (from the biomechanical fusion modeling module), when the resistance value suddenly changes by more than 10%, the coefficients of the suction flow control curve are dynamically readjusted.

[0014] Optionally, the multiphysics execution control module further includes: Dynamic switching of the focal morphology of the laser ablation unit: The laser ablation unit is driven based on the surface energy distribution data of the crystal, including: When the energy density in the surface energy distribution data of the crystal is >1.2 J / m 2 At that time, the ellipsoid focal shape is activated; When the energy density is between 0.8 and 1.2 J / m³ 2 At that time, the columnar focal point pattern is activated; When energy density < 0.8 J / m 2 Furthermore, when the crystal distribution dispersion is greater than 40%, the multifocal morphology is activated; Spatiotemporal synchronization control executed by multiphysics: Establishing the action timing protocol for the magnetic navigation puncture needle, adaptive negative pressure aspiration device, and laser ablation unit, including: Phase 1: After the magnetically guided puncture needle moves to the target position, the adaptive negative pressure aspiration device is activated after a 50ms delay; Phase 2: When the suction flow rate reaches 90% of the critical suction flow rate, the laser ablation unit is triggered to emit. Phase 3: During the laser ablation period, the flow rate of the adaptive negative pressure suction device is as follows: To lift and remove ablation products; Real-time verification and iterative triggering of execution results: Real-time acquisition of residual crystallization data after ablation using a microscopic optical coherence tomography device in the data acquisition module. If the residual crystal surface energy distribution data is greater than 20% of the initial value, it is marked as an area of ​​insufficient ablation; The coordinates of the insufficiently ablated area are sent to the dynamic path planning module, triggering a local replanning of the dynamic navigation instruction set until the residual crystal surface energy distribution data is ≤ 5% of the initial value.

[0015] The beneficial effects of this invention are: This invention achieves the simultaneous acquisition of crystal-tissue interface stress, surface energy density, and synovial fluid viscoelastic parameters through a multimodal image data acquisition mechanism that integrates dual-frequency ultrasound, energy spectrum CT, and microscopic optical coherence tomography. By combining algorithms such as signal-to-noise ratio analysis, refraction compensation, and viscoelastic registration, a three-dimensional biomechanical model fused from multiple sources is constructed. This model effectively characterizes the risk of crystal adhesion and the distribution of local fluid resistance, ensuring the authenticity of the anatomical and physical fields upon which the navigation path is based, and improving the accuracy of lesion identification and intervention target identification.

[0016] This invention combines the distribution of crystal adhesion risk coefficients with the fluid resistance prediction matrix, and uses a six-degree-of-freedom path compensation and minimum resistance path algorithm to dynamically generate motion compensation vectors and suction optimization parameters for surgical instruments. By introducing safety margin indicators and collision prediction mechanisms, high-risk areas are dynamically avoided. Combined with real-time adjustment of the aspiration flow control curve, the risks of crystallization disturbance, aspiration blockage, and intraoperative puncture deviation are significantly reduced, achieving a closed-loop safety optimization of path planning, control commands, and instrument execution.

[0017] This invention, based on a dynamic navigation instruction set, drives a magnetic navigation puncture needle, an adaptive negative pressure aspiration device, and a laser ablation unit to perform spatial pose compensation, dynamic aspiration adjustment, and focus energy mode switching, respectively. By setting phased control logic and synchronous triggering conditions, the ablation products are efficiently removed during the ablation process. Combined with real-time monitoring and feedback from a microscopic optical coherence tomography device, the system can automatically identify and replan areas with insufficient ablation, effectively improving the thoroughness of uric acid crystal removal and minimizing tissue trauma and postoperative residual risks. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this 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 for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the system flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the biomechanical fusion modeling module in an embodiment of the present invention. Detailed Implementation

[0020] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0021] It should be noted that the use of terms such as "an embodiment," "an embodiment," "an exemplary embodiment," and "some embodiments" in the specification indicates that the described embodiment may include a specific feature, structure, or characteristic, but not every embodiment necessarily includes that specific feature, structure, or characteristic. Furthermore, when a specific feature, structure, or characteristic is described in connection with an embodiment, implementing such a feature, structure, or characteristic in conjunction with other embodiments (whether explicitly described or not) should be within the knowledge of those skilled in the art.

[0022] Generally, terms can be understood at least partly from their use in context. For example, depending at least partly on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in a singular sense, or a combination of features, structures, or characteristics in a plural sense. Additionally, the term "based on" can be understood not necessarily to convey an exclusive set of factors, but rather, alternatively, depending at least partly on the context, to allow for the presence of other factors that are not necessarily explicitly described.

[0023] like Figure 1-2 As shown, the multimodal image navigation system for minimally invasive gout treatment includes a data acquisition module, a biomechanical fusion modeling module, a dynamic path planning module, and a multiphysics execution control module, wherein: The data acquisition module integrates a dual-frequency ultrasound probe, an energy-dispersive CT scanner, and a microscopic optical coherence tomography device to simultaneously acquire stress gradient data at the crystallization-tissue interface, surface energy distribution data of crystallization, and viscoelastic parameters of synovial fluid. The biomechanical fusion modeling module receives the stress gradient data of the crystallization-tissue interface, the crystallization surface energy distribution data, and the viscoelastic parameters of the synovial fluid output by the data acquisition module. It generates a three-dimensional biomechanical fusion model through the crystallization surface energy compensation algorithm. The three-dimensional biomechanical fusion model includes the crystallization adhesion risk coefficient and the fluid resistance prediction matrix. The dynamic path planning module receives the biomechanical fusion model generated by the biomechanical fusion modeling module, and calculates the following parameters by combining it with real-time device pose data: Device motion compensation vector based on adhesion risk coefficient; Optimization parameters for the suction path based on the fluid resistance matrix; Output dynamic navigation command set, which includes instrument pose correction amount and suction flow control curve; The multiphysics execution control module receives and outputs a set of dynamic navigation instructions, synchronously driving the magnetic navigation puncture needle, the adaptive negative pressure aspiration device, and the laser ablation unit, wherein: Magnetic navigation puncture needle: Six degrees of freedom motion compensation based on instrument position correction; Adaptive negative pressure suction device: dynamically adjusts suction power according to the suction flow control curve; Laser ablation unit: Adjusts the shape of the ablation focus (ellipsoid / columnar / multifocal) based on the surface energy distribution data of the crystal.

[0024] The data acquisition module includes: Initial configuration: The dual-frequency ultrasound probe is set to alternate between high-frequency mode (20MHz) and low-frequency mode (5MHz). The high-frequency mode is used for elastic imaging to acquire stress gradient data of the crystal-tissue interface, and the low-frequency mode is used to penetrate deep tissues. The energy dispersive CT scanner is configured with a tube voltage of 80–140 kVp and a multi-energy spectral separation mode to generate energy dissipation curves of energy distribution data on the crystal surface. The micro-optical coherence tomography device was set to a center wavelength of 1300 nm and an axial resolution of 5 μm for dynamic tomographic scanning of the viscoelastic parameters of synovial fluid. Synchronous data acquisition: The dual-frequency ultrasound probe, energy spectrum CT scanner, and micro-optical coherence tomography device are synchronously activated by hardware trigger signals to ensure that the three acquire data under the same time reference. A radio frequency signal sequence output by a dual-frequency ultrasound probe to provide stress gradient data at the crystallization-tissue interface; The projected energy spectrum matrix of the energy distribution data on the crystal surface output by the energy spectroscopy CT scanner; The microscopic optical coherence tomography device outputs a three-dimensional interference signal of the viscoelastic parameters of the synovial fluid; Extraction and fusion of stress gradient data at the crystallization-tissue interface: The radio frequency signal sequence output from the dual-frequency ultrasound probe is input into the elastic inversion algorithm to generate a two-dimensional elastic map of the stress gradient data at the crystallization-tissue interface. The calculation formula for the elastic inversion algorithm is as follows: ,in, Position on a two-dimensional elasticity graph Stress gradient at the location, The shear modulus field is obtained by frequency response inversion. The strain tensor at the corresponding position is obtained from the radio frequency signal through sliding window correlation processing.

[0025] The two-dimensional elastic map is converted into a three-dimensional stress gradient distribution tensor aligned with the coordinate system of the energy dispersive CT scanner using a spatial interpolation algorithm (such as cubic spline interpolation). ; Reconstruction and calibration of surface energy distribution data of uric acid crystals: The projected energy spectrum matrix output from the energy spectrum CT scanner is input into the matrix material decomposition algorithm to separate the energy spectrum characteristics of uric acid crystals from the surrounding tissue, and the linear attenuation coefficient of the uric acid crystal region is calculated. ; The energy density cloud map of the crystal surface energy distribution data is generated based on the surface energy calculation model. The model calculation formula is as follows: ; in, For position Surface energy density per unit area at that location, For energy-spectral CT at different energy levels The linear decay coefficient under the condition, The intensity function of the X-ray energy spectrum. This is a material-specific weighting factor (obtained through experimental calibration).

[0026] The refractive index of the energy density contour map was calibrated using the viscoelastic parameters of the synovial fluid output from a microscopic optical coherence tomography (EMT) device. A correction factor was used during the calibration process. The calculation formula is as follows: ,in, The surface energy density after refraction compensation. For frequency The related complex refractive index of the synovial fluid.

[0027] Multimodal registration of synovial fluid viscoelastic parameters: The three-dimensional interference signal output from the microscopic optical coherence tomography (OCT) device is input into the viscoelastic deconvolution algorithm to generate frequency-strain rate response curves. The viscoelastic deconvolution algorithm is a generalized Maxwell model, and its viscoelastic modulus expression is: ; in, For the complex modulus, The energy storage modulus reflects elastic behavior. Loss modulus, reflecting viscous behavior. ω is the angular frequency.

[0028] Pair the frequency-strain rate response curve with the three-dimensional stress gradient distribution tensor Data on the distribution of crystal surface energy after refraction compensation Spatiotemporal registration was performed using a multimodal image registration algorithm (based on maximizing mutual information) to generate a three-dimensional viscoelastic field of synovial fluid viscoelastic parameters. Data fusion and output: The three-dimensional stress gradient distribution tensor, the refractive compensated crystal surface energy distribution data and the three-dimensional viscoelastic field are fused in a multi-modal manner. The dimensionality reduction strategy of joint feature tensor splicing and principal component analysis (PCA) is adopted to generate a synchronous acquisition data package of the data acquisition module and send it to the biomechanical fusion modeling module.

[0029] The data acquisition module also includes: Quality verification and anomaly marking of synchronously acquired data packets: The generated synchronously acquired data packets are input into the data quality assessment algorithm, which calculates the data quality score based on the following three indicators. (1) Signal-to-noise ratio of the three-dimensional stress gradient distribution tensor , is represented as: ,in, For three-dimensional tensors The mean, For three-dimensional tensors The variance.

[0030] (2) Energy density standard deviation of crystal surface energy distribution data after refraction compensation , is represented as: ,in, For the first Surface energy density of individual primitive points The average value for all voxel points. The total number of voxels; (3) Spatial continuity parameters of three-dimensional viscoelastic field , defined as the inverse ratio of the mean to the complex modulus gradient between neighboring voxels: ; in, For complex modulus field , The number of sampling points for continuous detection. This represents the gradient of the complex modulus.

[0031] The comprehensive scoring function is as follows: ; in The weighting coefficients (system preset empirical values, satisfying...) (If rating) Below the preset threshold If it is an exception, it is marked as an exception data packet and an exception type label is output. Exception type labels include: Stress gradient distortion: ; Surface energy refraction shift: ; Viscoelastic field fracture: ; in , , These are the individual lower limit thresholds for each quality indicator.

[0032] Dynamic parameter reconfiguration based on exception type labels: Based on the exception type labels, the following rules are invoked to reconfigure device parameters: If the label indicates stress gradient distortion, then: adjust the transmission power of the high-frequency mode of the dual-frequency ultrasonic probe to 150% of the reference value. The pulse interval of the low-frequency mode is shortened to 10μs; If the label is surface energy refraction offset, then: reload the uric acid crystal energy spectrum feature library of energy spectrum CT and start the refractive index dynamic compensation mode of the micro-optical coherence tomography device. If the label indicates a viscoelastic fracture, increase the scanning frame rate of the optical coherence tomography (OCT) device to 200 Hz and expand the sampling bandwidth of the frequency-strain rate response curve to [missing information]. ; Local re-acquisition and replacement of abnormal data packets: Based on the dissection area corresponding to the selected tag, only that area is re-acquisitioned after reconfiguring the device parameters. Dual-frequency ultrasound probe outputs local radio frequency signal sequence Replace the original stress tensor area; The output local projection energy spectrum matrix of the energy spectrum CT scanner ,replace area; Microscopic optical coherence tomography device outputs local interference signal ,replace area.

[0033] Fusion verification of the corrected synchronously acquired data packets: Perform secondary multimodal fusion on the above three types of re-acquired data, update the synchronously acquired data packets, and re-input them into the quality assessment algorithm to calculate the score. ; like > Output a verification pass signal; Otherwise, based on the abnormal label, the iterative correction loop of dynamic parameter reconfiguration → local re-acquisition → fusion verification is triggered again until the score meets the conditions.

[0034] Dynamic data loading and biomechanical model update: The final verified and corrected synchronously collected data packets are sent to the biomechanical fusion modeling module to build a highly reliable three-dimensional biomechanical fusion model.

[0035] The biomechanical fusion modeling module includes: Preprocessing and feature extraction of synchronous acquisition data packets: Receive synchronous acquisition data packets output by the data acquisition module. The synchronous acquisition data packets include: Three-dimensional stress gradient distribution tensor ; Refraction-compensated crystal surface energy distribution data ; Three-dimensional viscoelastic field ; right Anisotropic filtering is performed using a covariance adaptive filter guided by the structure tensor to generate a denoised three-dimensional stress gradient distribution tensor. ; right Perform energy density threshold segmentation and extract samples that meet the conditions. J / In the region, a high-density region of crystalline surface energy is obtained. ; right and Perform fluid-structure interaction registration to generate the interface displacement coupling matrix between the viscoelastic field and the stress field. ; Finite element modeling of crystal adhesion risk coefficient: using the denoised stress tensor Input a crystallization-tissue contact mechanics model based on Hertz contact theory and calculate the local stress field in the contact region. ,in: ; in, For local contact force, by Projection integration is obtained, The contact radius is determined based on the local stiffness and morphology model. The radius is the relative contact position radius. For the contact stress tensor, High-density surface energy regions Superimpose, perform energy correction, and calculate the correction energy term in the crystal surface energy compensation algorithm: ; in, The total energy compensation per unit volume. This is the surface energy compensation coefficient (set during system initialization or in subsequent iterative calibration). The surface energy density after refraction compensation is used to generate a crystal adhesion risk coefficient based on the integral result. The three-dimensional probability distribution diagram.

[0036] Fluid-structure interaction calculation of fluid resistance prediction matrix: and Input the Navier-Stokes equations for fluid-space coupling solution, combined with stress tensor Calculate the velocity field of synovial fluid on the crystallization surface With shear stress field Its basic governing equations are: ; in, The density of the synovial fluid. Frequency-dependent viscosity, by Real part derivation, This is a volume force term; Based on the spatial variation rate of the velocity field and the shear stress field: An initial weight map of the fluid resistance prediction matrix is ​​constructed, and its local response attenuation factor is dynamically corrected using viscoelastic parameters. ; ; in, To be based on the complex modulus The viscous impedance factor is set based on the frequency response.

[0037] Iterative calibration of the three-dimensional biomechanical fusion model: plotting the crystal adhesion risk coefficient. With fluid resistance prediction matrix Perform the following iterative coupling: In the region Inside, with Using this as a reference, calculate: ; If there exists any Then adjust and re-execute Calculate until all regions satisfy: ; The revised and Perform weighted spatial fusion to generate a three-dimensional biomechanical fusion model: ; in Let be the fusion coefficient, satisfying .

[0038] Model output and dynamic path planning interface: This will integrate the model. The output is sent to the dynamic path planning module and used as the basis for the following path decisions: based on The provided 3D probability map outputs the motion compensation vector of the device; based on Output the suction path optimization parameter set.

[0039] The final generated three-dimensional biomechanical fusion model includes: Three-dimensional probability distribution of crystal adhesion risk coefficient Three-dimensional dynamic response model of fluid resistance prediction matrix .

[0040] The dynamic path planning module includes: Registration of the 3D biomechanical fusion model with real-time device pose data: Receive the 3D biomechanical fusion model output from the biomechanical fusion modeling module. The fusion model includes: Three-dimensional probability distribution of crystal adhesion risk coefficient ; Three-dimensional dynamic response model of fluid resistance prediction matrix ; Will With real-time instrument pose data Spatial registration is performed, among which... The coordinates of the puncture needle tip. These are attitude angles (Euler angles). Instantaneous velocity; Define the safety margin index of the instrument tip for: ; in, From the current puncture tip to the nearest high-risk area The Euclidean distance, if If it is, then it is marked as a path segment with potential adhesion risk.

[0041] Dynamic calculation of motion compensation vector of the device: based on safety margin index and fluid resistance prediction matrix The following two path planning processes are executed in parallel: Adhesion risk avoidance path planning: while satisfying In the region, an avoidance path is iteratively generated using the three-dimensional gradient descent method: ; Generate a six-degree-of-freedom compensation vector along the negative gradient direction: ; Attitude Adaptive Rotation Matching ; in This is the step size coefficient.

[0042] Fluid resistance optimization path generation: under the condition of satisfying Within a range of kPa, fluid blockage channels are identified, and the minimum resistance path algorithm is used to solve for the optimal suction angle and critical suction flow rate. Optimal suction angle This minimizes the angle between the suction vector and the path of least resistance. Critical suction flow rate Satisfying shear stress The minimum flow rate.

[0043] Synthesis and verification of dynamic navigation instruction sets: Input the output of the two path generation modules mentioned above into the instruction synthesis engine: (1) Generate the device pose correction amount based on the compensation vector: displacement vector Rotation angle ; (2) Optimize parameters based on suction path Real-time updated gradient in the resistance prediction matrix The suction flow control curve is constructed using a cubic spline function: ; coefficient Calculated according to the following rules: ; That is, the response of the suction flow rate to the dynamic changes in resistance is negatively correlated.

[0044] (3) Merge the above two types of instructions to generate the initial version of the dynamic navigation instruction set.

[0045] Real-time prediction of instrument-crystallization collision risk: Input the initial version of the dynamic navigation instruction set into the collision prediction model: Based on the instrument pose correction, the puncture needle trajectory is simulated. ; If any If so, it is marked as a high-risk collision zone; Output collision risk level: ; Iterative correction of collision avoidance instructions: If the collision risk level is "high", the following correction process will be initiated: (1) Reduce the adhesion risk coefficient threshold in the original path planning from 0.7 to 0.6; (2) Re-execute gradient descent calculation based on the new threshold and update the six-degree-of-freedom compensation vector; (3) Repeat the synthesis and verification of dynamic navigation instruction set - instrument - real-time prediction of crystallization collision risk, repeat the synthesis of instructions and collision verification until the collision risk level is lower than "high"; The dynamic navigation instruction set output and execution interface: The verified navigation instruction set is sent as the final output to the multiphysics execution control module. Final instrument pose correction: used to drive the magnetically guided puncture needle to complete six-degree-of-freedom spatial motion compensation; Final suction flow control curve: used to drive the adaptive negative pressure suction device to dynamically adjust the suction power and achieve low-resistance and precise suction.

[0046] The multiphysics execution control module includes: Parsing and Execution Allocation of Dynamic Navigation Instruction Sets: Receives the dynamic navigation instruction set output by the dynamic path planning module. The instruction set includes: Final instrument pose correction ; Final suction flow control curve Will: The pose correction amount is allocated to the magnetic navigation puncture needle control channel to generate a six-degree-of-freedom compensation control command. The suction flow control curve is assigned to the control channel of the adaptive negative pressure suction device to generate a suction power modulation waveform. Simultaneously extract crystal surface energy distribution data It is allocated to the laser ablation unit control channel.

[0047] Six-degree-of-freedom motion compensation execution of the magnetic navigation puncture needle: based on the six-degree-of-freedom compensation control command: (1) Analytical displacement vector A triaxial magnetic field is generated by a magnetic field gradient generator. The tip of the puncture needle is moved with an accuracy of 0.1 mm / ms. (2) Analyze the Euler angles of rotation Through a three-dimensional rotating magnetic field Generate directional torque, adjust attitude angle, angular resolution And provide real-time pose feedback To the dynamic path planning module, closed-loop correction of control error: ; Dynamic flow control of the adaptive negative pressure suction device: The suction flow control curve is as follows: ; Converted to PWM duty cycle control signal The mapping relationship for controlling the speed of the negative pressure pump motor is as follows: ; in, To control the gain factor (set by the pump response characteristics); Based on real-time resistance prediction data provided by the biomechanical fusion modeling module: If the rate of change of resistance If the initial suction rate is increased to kPa / s, then the suction rate will be dynamically increased. ; like If the pressure is kPa / s, then reduce the basic suction flow rate: ; in, The system adaptively adjusts the increment.

[0048] Dynamic switching of the focal morphology of the laser ablation unit: based on the surface energy density of the crystal. Select the following laser focus mode: (1) Ellipsoid focus (high energy density), when J / ; ; Columnar focal point (medium density), when J / ; ; Column height = 3mm; Multi-focus mode (low density + discrete), when And crystallization dispersion : ; Spatiotemporal synchronization control for multiphysics execution: Define the following execution phases: Phase 1 (Navigation → Aspiration): When the magnetically guided puncture needle reaches the target position... Then, the suction system is started after a 50ms delay; Phase 2 (Aspiration-Ablation): When the real-time aspiration flow rate... At the critical suction flow rate, the laser ablation unit is activated; Phase 3 (Enhanced Ablation with Simultaneous Aspiration): During laser ablation, the aspiration flow rate is adjusted to: ; This is to remove local ablation products and maintain regional transparency and safety.

[0049] Real-time verification and iterative triggering of execution results: Data on residual crystallization after ablation is acquired using a microscopic optical coherence tomography device. If the residual surface energy density If so, it is determined to be an area of ​​insufficient ablation; Extract the coordinates of this area and send them to the dynamic path planning module to trigger local replanning and navigation command regeneration until the termination condition is met: .

[0050] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0051] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A multimodal image-guided system for minimally invasive treatment of gout, characterized in that, It includes a data acquisition module, a biomechanical fusion modeling module, a dynamic path planning module, and a multiphysics execution control module, among which: The data acquisition module integrates a dual-frequency ultrasound probe, an energy-dispersive CT scanner, and a microscopic optical coherence tomography device to simultaneously acquire stress gradient data at the crystallization-tissue interface, surface energy distribution data of crystallization, and viscoelastic parameters of synovial fluid. The biomechanical fusion modeling module receives the stress gradient data of the crystallization-tissue interface, the crystallization surface energy distribution data, and the viscoelastic parameters of the synovial fluid output by the data acquisition module, and generates a three-dimensional biomechanical fusion model through the crystallization surface energy compensation algorithm. The three-dimensional biomechanical fusion model includes the crystallization adhesion risk coefficient and the fluid resistance prediction matrix. The dynamic path planning module receives the biomechanical fusion model generated by the biomechanical fusion modeling module, and calculates the following parameters by combining it with real-time device pose data: Device motion compensation vector based on adhesion risk coefficient; Optimization parameters for the suction path based on the fluid resistance matrix; Output a dynamic navigation instruction set, which includes instrument pose correction amount and suction flow control curve; The multiphysics field execution control module receives the output dynamic navigation instruction set and synchronously drives the magnetic navigation puncture needle, the adaptive negative pressure aspiration device, and the laser ablation unit.

2. The multimodal image navigation system for minimally invasive gout treatment according to claim 1, characterized in that, The data acquisition module includes: Initial configuration: The dual-frequency ultrasound probe is set to alternate between high-frequency and low-frequency modes. The high-frequency mode is used for elastic imaging to acquire stress gradient data at the crystal-tissue interface, and the low-frequency mode is used to penetrate deep tissues. The energy dispersive CT scanner is configured with a tube voltage of 80-140kVp and a multi-energy spectral separation mode to generate energy dissipation curves of energy distribution data on the crystal surface. The micro-optical coherence tomography device was set to a center wavelength of 1300 nm and an axial resolution of 5 μm for dynamic tomographic scanning of the viscoelastic parameters of synovial fluid. Synchronous data acquisition: The dual-frequency ultrasound probe, energy spectrum CT scanner, and microscopic optical coherence tomography device are synchronously activated by hardware trigger signals; A radio frequency signal sequence output by a dual-frequency ultrasound probe to provide stress gradient data at the crystallization-tissue interface; The projected energy spectrum matrix of the energy distribution data on the crystal surface output by the energy spectroscopy CT scanner; The microscopic optical coherence tomography device outputs a three-dimensional interference signal of the viscoelastic parameters of the synovial fluid; Extraction and fusion of stress gradient data at the crystallization-tissue interface: The radio frequency signal sequence output by the dual-frequency ultrasound probe is input into the elastic inversion algorithm to generate a two-dimensional elastic map of the stress gradient data at the crystallization-tissue interface; The two-dimensional elastic map is converted into a three-dimensional stress gradient distribution tensor aligned with the coordinate system of the energy spectrum CT scanner using a spatial interpolation algorithm. Reconstruction and calibration of surface energy distribution data of crystals: The projected energy spectrum matrix output by the energy spectrum CT scanner is input into the base material decomposition algorithm to separate the energy spectrum characteristics of uric acid crystals from those of the surrounding tissue; Based on the decomposed energy spectrum characteristics, an energy density cloud map of the crystal surface energy distribution data is generated through a surface energy calculation model. The refractive index of the energy density cloud map was calibrated using the viscoelastic parameters of the synovial fluid output by the micro-optical coherence tomography device, and the refractive compensation data of the crystal surface energy distribution were obtained. Multimodal registration of synovial fluid viscoelastic parameters: The three-dimensional interference signal output from the microscopic optical coherence tomography device is input into the viscoelastic deconvolution algorithm to generate the frequency-strain rate response curve of the synovial fluid viscoelastic parameters; The frequency-strain rate response curve is spatiotemporally registered with the three-dimensional stress gradient distribution tensor and the refractive compensated crystal surface energy distribution data to form a three-dimensional viscoelastic field of synovial fluid viscoelastic parameters. Data fusion and output: The three-dimensional stress gradient distribution tensor, the refractive compensated crystal surface energy distribution data, and the three-dimensional viscoelastic field are fused in multiple modes to generate a synchronously acquired data package.

3. The multimodal image navigation system for minimally invasive gout treatment according to claim 2, characterized in that, The data acquisition module also includes: Quality verification and anomaly marking of synchronous acquisition data packets: The generated synchronous acquisition data packets are input into the data quality assessment algorithm, which calculates the quality score of the synchronous acquisition data packets based on the signal-to-noise ratio of the three-dimensional stress gradient distribution tensor, the standard deviation of the energy density of the crystal surface energy distribution data after refraction compensation, and the spatial continuity parameters of the three-dimensional viscoelastic field. If the quality score is lower than the preset threshold, it is marked as an abnormal data packet and an abnormal type label is output. The abnormal type label includes stress gradient distortion, surface energy refraction shift or viscoelastic field fracture. Dynamic parameter reconfiguration based on anomaly type labels: Matching parameter adjustment rules for dual-frequency ultrasound probes, energy dispersive CT scanners, and microscopic optical coherence tomography devices according to anomaly type labels. If the label indicates stress gradient distortion, adjust the high-frequency mode transmission power of the dual-frequency ultrasound probe to 150% of the reference value and shorten the low-frequency mode pulse interval to 10μs. If the label is surface energy refraction offset, then reload the uric acid crystal energy spectrum feature library in the base material decomposition algorithm of the energy spectrum CT scanner and start the refractive index dynamic compensation mode of the micro-optical coherence tomography device. If the label indicates viscoelastic field fracture, increase the scanning frame rate of the micro-optical coherence tomography device to 200 Hz and expand the sampling bandwidth of the frequency-strain rate response curve of the synovial fluid viscoelastic parameters. Local reacquisition and replacement of abnormal data packets: Based on the adjusted parameters of the dual-frequency ultrasound probe, energy dispersive CT scanner, and micro-optical coherence tomography (EDT), local reacquisition is initiated only for the anatomical region corresponding to the abnormality type label. The dual-frequency ultrasonic probe outputs a local radio frequency signal sequence to replace the abnormal data block in the original three-dimensional stress gradient distribution tensor. The energy spectrum CT scanner outputs a local projection energy spectrum matrix, which replaces the abnormal energy spectrum region in the original refractive compensated crystal surface energy distribution data. The micro-optical coherence tomography device outputs a local three-dimensional interference signal, which replaces the fracture viscoelastic parameters in the original three-dimensional viscoelastic field. Fusion verification of the corrected synchronous acquisition data packet: The three-dimensional stress gradient distribution tensor after local reacquisition, the crystal surface energy distribution data after refraction compensation, and the three-dimensional viscoelastic field are fused in a second multi-modal manner to generate the corrected synchronous acquisition data packet. A data quality assessment algorithm is executed on the corrected synchronously acquired data packets. If the quality score reaches the preset threshold, a verification pass signal is output; otherwise, an iterative correction loop of dynamic parameter reconfiguration → local reacquisition → fusion verification is triggered until the verification passes. Dynamic data loading and biomechanical model update: The validated and corrected data packets are synchronously collected and sent to the biomechanical fusion modeling module.

4. The multimodal image navigation system for minimally invasive gout treatment according to claim 3, characterized in that, The biomechanical fusion modeling module includes: Preprocessing and feature extraction of synchronous acquisition data packets: Receive synchronous acquisition data packets output by the data acquisition module, which include three-dimensional stress gradient distribution tensor, refractive compensated crystal surface energy distribution data and three-dimensional viscoelastic field; Anisotropic noise reduction filtering is performed on the three-dimensional stress gradient distribution tensor to generate a noise-reduced three-dimensional stress gradient distribution tensor. Energy density thresholding was performed on the refractive-compensated crystal surface energy distribution data to extract energy densities greater than 0.8 J / m². 2 High-density regions of crystalline surface energy; Fluid-structure interaction registration was performed on the three-dimensional viscoelastic field to generate the displacement coupling matrix between the synovial fluid viscoelastic parameters and the tissue interface; Finite element modeling of crystal adhesion risk coefficient: The denoised three-dimensional stress gradient distribution tensor is input into the crystal-tissue contact mechanics model, and the contact stress distribution at the crystal-soft tissue interface is calculated based on Hertz contact theory. By superimposing the high-density region of crystalline surface energy with the contact stress distribution, a three-dimensional probability distribution map of the initial crystalline adhesion risk coefficient is generated using the energy correction term formula in the crystalline surface energy compensation algorithm.

5. The multimodal image navigation system for minimally invasive gout treatment according to claim 4, characterized in that, The biomechanical fusion modeling module also includes: Fluid-structure interaction calculation of fluid resistance prediction matrix: Input the synovial fluid viscoelastic parameters and the displacement coupling matrix of the tissue interface into the Navier-Stokes equation, and combine the noise-reduced three-dimensional stress gradient distribution tensor to calculate the flow velocity field and shear stress field of synovial fluid on the crystallized surface. The initial weight distribution of the fluid resistance prediction matrix is ​​constructed based on the spatial gradient of the velocity field and the shear stress field. A three-dimensional dynamic response model of the fluid resistance prediction matrix is ​​generated by dynamically correcting the attenuation factor of the fluid resistance prediction matrix using viscoelastic parameters. Iterative calibration of the three-dimensional biomechanical fusion model: The three-dimensional probability distribution map of the initial crystallization adhesion risk coefficient is coupled and iterated with the three-dimensional dynamic response model of the fluid resistance prediction matrix. Within the high-density region of crystalline surface energy, the energy density difference of the crystalline surface energy distribution data after refraction compensation is used as a benchmark to calculate the crystalline adhesion risk coefficient. If the energy density difference is greater than 0.5 J / m 2 Then adjust the surface energy compensation coefficient in the crystallization surface energy compensation algorithm until the energy density difference is less than or equal to 0.5 J / m³. 2 ; The calibrated crystal adhesion risk coefficient and fluid resistance prediction matrix are spatially weighted and fused to generate an initial version of the three-dimensional biomechanical fusion model; Model output and dynamic path planning interface: The initial version of the 3D biomechanical fusion model is sent to the dynamic path planning module as the input source for the device motion compensation vector and suction path optimization parameters; The three-dimensional biomechanical fusion model includes a three-dimensional probability distribution map of the final crystallization adhesion risk coefficient and a three-dimensional dynamic response model of the fluid resistance prediction matrix.

6. The multimodal image navigation system for minimally invasive gout treatment according to claim 5, characterized in that, The dynamic path planning module includes: Registration of the three-dimensional biomechanical fusion model with real-time instrument pose data: receiving the three-dimensional biomechanical fusion model output by the biomechanical fusion modeling module, wherein the three-dimensional biomechanical fusion model includes a three-dimensional probability distribution map of the crystal adhesion risk coefficient and a three-dimensional dynamic response model of the fluid resistance prediction matrix; The three-dimensional probability distribution map of the crystal adhesion risk coefficient is spatially registered with the real-time instrument pose data to generate an instrument tip safety margin index, which is defined as the inverse product of the crystal adhesion risk coefficient and the distance to the instrument tip. Dynamic calculation of instrument motion compensation vector: Based on the three-dimensional dynamic response model of the instrument tip safety margin index and fluid resistance prediction matrix, the following parallel calculations are performed: Adhesion risk avoidance path planning: In the region where the crystal adhesion risk coefficient is >0.7, the gradient descent method is used to calculate the motion compensation vector of the device to avoid the high-risk region and output the six-degree-of-freedom compensation amount; Fluid resistance optimization path generation: Extract fluid blockage areas with resistance values ​​> 1.5 kPa from the fluid resistance prediction matrix, and generate suction path optimization parameters through the minimum resistance path algorithm; Synthesis and verification of dynamic navigation instruction sets: Inputting six-degree-of-freedom compensation quantities and suction path optimization parameters into the instruction synthesis engine: The device pose correction amount is generated based on the six-degree-of-freedom compensation amount, which includes a linear displacement vector and rotational Euler angles; Based on the optimal suction angle and critical suction flow rate, combined with the real-time updated data of the fluid resistance prediction matrix, a cubic spline function of the suction flow rate control curve is generated. By integrating the instrument pose correction amount and the suction flow control curve, an initial version of the dynamic navigation instruction set is formed.

7. The multimodal image navigation system for minimally invasive gout treatment according to claim 6, characterized in that, The dynamic path planning module also includes: Real-time prediction of instrument-crystallization collision risk: Input the initial version of the dynamic navigation instruction set into the collision prediction model: Based on the three-dimensional probability distribution diagram of the crystal adhesion risk coefficient, the motion trajectory of the puncture needle after the instrument pose correction is simulated. If the trajectory of the motion intersects with a region where the crystal adhesion risk coefficient is greater than 0.9, it is marked as a high-risk collision zone, and the collision risk level is output. The collision risk level is divided into low, medium and high. Iterative correction of collision avoidance instructions: If the collision risk level is high, path replanning is triggered, including: In the dynamic calculation of the motion compensation vector of the device, a collision constraint condition is added to the adhesion risk avoidance path planning: the threshold of the crystal adhesion risk coefficient is increased from 0.7 to 0.6; Recalculate the motion compensation vector of the device to generate updated six-degree-of-freedom compensation quantities; The updated six-degree-of-freedom compensation values ​​are input into the synthesis and verification of the dynamic navigation instruction set and the real-time prediction of the instrument-crystallization collision risk for secondary verification until the collision risk level is reduced to medium or low. Output and execution interface of dynamic navigation instruction set: The verified dynamic navigation instruction set is sent to the multiphysics execution control module. The dynamic navigation instruction set includes the final instrument pose correction amount and the final suction flow control curve.

8. The multimodal image navigation system for minimally invasive gout treatment according to claim 7, characterized in that, The multiphysics execution control module includes: Parsing and execution allocation of dynamic navigation instruction set: Receives the dynamic navigation instruction set output by the dynamic path planning module, the dynamic navigation instruction set including the final instrument pose correction amount and the final suction flow control curve; The final instrument pose correction amount is allocated to the magnetic navigation puncture needle control channel to generate a six-degree-of-freedom motion compensation command for the magnetic navigation puncture needle. The final suction flow control curve is assigned to the control channel of the adaptive negative pressure suction device to generate a suction power modulation waveform. Simultaneously, the surface energy distribution data of the crystallized crystals is extracted and allocated to the control channel of the laser ablation unit; Six-degree-of-freedom motion compensation execution of the magnetic navigation puncture needle: The magnetic navigation puncture needle is driven according to six-degree-of-freedom motion compensation commands, including: The linear displacement vector is analyzed, and a displacement magnetic field is generated by the triaxial gradient coil of the magnetic navigation system to drive the tip of the puncture needle to move with an accuracy of 0.1 mm / ms. The Euler angles of rotation are analyzed, and the puncture needle attitude angle is adjusted by a rotating torque magnetic field generator, with an angular resolution of 0.5°. Real-time feedback of the actual position data of the puncture needle to the dynamic path planning module forms a closed-loop position control; Dynamic flow control of the adaptive negative pressure suction device: The adaptive negative pressure suction device is driven according to the suction power modulation waveform, including: The final suction flow control curve is converted into a pulse width modulation duty cycle sequence to control the motor speed of the negative pressure pump; Based on real-time updated data from the fluid resistance prediction matrix, when the resistance value suddenly changes by more than 10%, the coefficients of the suction flow control curve are dynamically readjusted.

9. The multimodal image navigation system for minimally invasive gout treatment according to claim 8, characterized in that, The multiphysics execution control module also includes: Dynamic switching of the focal morphology of the laser ablation unit: The laser ablation unit is driven based on the surface energy distribution data of the crystal, including: When the energy density in the surface energy distribution data of the crystal is >1.2 J / m 2 At that time, the ellipsoid focal shape is activated; When the energy density is between 0.8 and 1.2 J / m³ 2 At that time, the columnar focal point pattern is activated; When energy density < 0.8 J / m 2 Furthermore, when the crystal distribution dispersion is greater than 40%, the multifocal morphology is activated; Spatiotemporal synchronization control executed by multiphysics: Establishing the action timing protocol for the magnetic navigation puncture needle, adaptive negative pressure aspiration device, and laser ablation unit, including: Phase 1: After the magnetically guided puncture needle moves to the target position, the adaptive negative pressure aspiration device is activated after a 50ms delay; Phase 2: When the suction flow rate reaches 90% of the critical suction flow rate, the laser ablation unit is triggered to emit. Phase 3: During the laser ablation period, the flow rate of the adaptive negative pressure suction device is increased to remove ablation products; Real-time verification and iterative triggering of execution results: Real-time acquisition of residual crystallization data after ablation using a microscopic optical coherence tomography device in the data acquisition module. If the residual crystal surface energy distribution data is greater than 20% of the initial value, it is marked as an area of ​​insufficient ablation; The coordinates of the insufficiently ablated area are sent to the dynamic path planning module, triggering a local replanning of the dynamic navigation instruction set until the residual crystal surface energy distribution data is ≤ 5% of the initial value.

Images