Dual-mode registration method and system for optical positioning and electromagnetic navigation of surgical robots

Abstract

The application belongs to the technical field of medical apparatus and instruments, and discloses a surgical robot optical positioning and electromagnetic navigation dual-mode registration method, and the specific steps of the surgical robot optical positioning and electromagnetic navigation dual-mode registration method are as follows: S1, data acquisition and time synchronization; S2, dynamic coupling registration; S3, environment perception driving weight adjustment; S4, multi-level interference closed loop self-correction; and S5, output and dynamic deformation compensation. Through the dual-mode registration method of optical positioning and electromagnetic navigation, the application effectively integrates high-frequency local pose data and low-frequency global pose data, avoids the measurement failure caused by the influence of single mode under the condition of shielding or magnetic interference, adopts a time stamp synchronization unit to realize time alignment of multi-source data, combines a sliding fusion and Kalman filtering algorithm to dynamically couple short-time and long-time window data, so that the registration result maintains high precision and stability in a dynamic surgical environment, and is obviously superior to the existing single mode or simple fusion method.

Description

Technical Field

[0001] This invention belongs to the field of medical device technology, specifically a method and system for dual-mode registration of optical positioning and electromagnetic navigation for surgical robots. Background Technology

[0002] With the development of minimally invasive surgery, surgical robots have been widely used in neurosurgery, laparoscopic surgery and other fields. The accuracy requirements for positioning and navigation during surgery are extremely high. Existing positioning technologies are mainly divided into two categories: optical positioning and electromagnetic navigation.

[0003] Optical positioning relies on infrared or visible light reflection point arrays, which have advantages such as high frequency and high precision. However, during surgery, factors such as obstruction, blood, smoke, or instrument reflection can easily lead to signal loss or measurement errors. In contrast, electromagnetic navigation is based on magnetic field sensing and can provide global pose information without line of sight. However, it is limited by interference from metal equipment in the operating room and the stability of the magnetic field, resulting in lower data frequency and a larger error range.

[0004] In existing technologies, single-modal approaches often fail to achieve both high local accuracy and global stability. Especially in dynamic soft tissue environments, positioning errors can easily accumulate, leading to deviations in navigation and manipulation, and affecting surgical safety and outcomes. Some studies have attempted to introduce multimodal data fusion, but they still face challenges such as difficulties in time alignment, rigid modal weights, sensitivity to environmental interference, and insufficient adaptive adjustment capabilities. A mature and robust dual-modal dynamic registration and real-time compensation scheme is lacking. Therefore, how to achieve efficient and accurate optical and electromagnetic dual-modal registration in complex, dynamic, and multi-interference surgical scenarios has become a key technical problem that needs to be solved. Summary of the Invention

[0005] The purpose of this invention is to provide a dual-mode registration method and system for optical positioning and electromagnetic navigation of surgical robots, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a dual-mode registration method for optical positioning and electromagnetic navigation of a surgical robot, the specific steps of which are as follows:

[0007] S1 Data Acquisition and Time Synchronization: The optical positioning module acquires high-frequency three-dimensional pose data, and the electromagnetic navigation module acquires low-frequency global pose data. The registration calculation module calculates the time difference between the optical and electromagnetic data through the timestamp synchronization unit and aligns and synchronizes them to ensure the time consistency of the dual-modal data stream.

[0008] S2 Dynamic Coupling Registration: The multi-time window sliding fusion unit of the registration calculation module processes optical data based on a short time window and electromagnetic data based on a long time window. It combines the Kalman filter algorithm to adjust the gain of the predicted state and the observed state, dynamically couples the two-mode data, and generates a preliminary registration result.

[0009] S3 Environmental Perception Driven Weight Adjustment: The environmental perception and adaptive interference correction module monitors optical occlusion, magnetic field interference and reflectivity changes in real time, calculates the confidence level of each mode, and feeds it back to the sliding fusion unit to dynamically adjust the weighting ratio of optical and electromagnetic data and improve fusion stability.

[0010] S4 Multi-level Interference Closed-Loop Self-Correction: When the registration calculation module detects that the residual error exceeds the preset threshold, it starts multi-level closed-loop self-correction, including sensor-level outlier removal, algorithm-level adaptive adjustment of filter measurement noise covariance matrix, and system-level mode dominance switching, to generate optimized registration results.

[0011] S5 Output and Dynamic Deformation Compensation: The optimized registration pose data is transmitted to the surgical robot control system through the output interface. The deformation compensation module, based on the preoperative anatomical model and intraoperative pose data, uses non-rigid deformation field estimation and iterative matching correction algorithm to correct the registration error caused by the dynamic deformation of soft tissue in real time, thereby achieving accurate dynamic compensation.

[0012] Preferably, the specific steps for data acquisition and time synchronization in S1 are as follows:

[0013] S11 High-Frequency and Low-Frequency Pose Data Acquisition: The optical positioning module acquires high-frequency local three-dimensional pose information, while the electromagnetic navigation module acquires low-frequency global pose data, ensuring coverage of visible and invisible areas in complex surgical environments, and providing a multi-scale, multi-source data foundation for subsequent dual-modal registration;

[0014] S12 Time Synchronization and Consistency Calibration: The registration calculation module has a built-in timestamp synchronization unit. Based on the time synchronization error calculation formula, it evaluates the time difference between optical and electromagnetic data, completes time alignment, and ensures that multimodal data streams are processed collaboratively under the same time reference, laying the foundation for consistency conditions for dynamic coupling registration.

[0015] The formula for calculating time synchronization error is as follows:

[0016]

[0017] In the formula, Optical positioning data acquisition timestamp Electromagnetic navigation data acquisition timestamp Time difference between optical and electromagnetic data acquisition;

[0018] Principle Explanation: Based on classical time synchronization theory, it calculates the time between high- and low-frequency data sources.

[0019] Align the differences;

[0020] Technical effect: Ensures temporal consistency of dual-modal data, providing a unified time reference for subsequent dynamic registration;

[0021] Source: Multi-sensor synchronization (typical source: Robot Operating System (ROS) messagefilters module).

[0022] Preferably, the specific steps of dynamic coupling registration in S2 are as follows:

[0023] S21 Multi-Time Window Layered Processing: The multi-time window sliding fusion unit uses a short time window to process optical data to capture high-frequency dynamic changes, and a long time window to process electromagnetic data to extract steady-state global pose, forming a layered processing framework that lays the data foundation for the stable fusion of subsequent multimodal data.

[0024] S22 Prediction Gain-Driven Dynamic Coupling: During the fusion process, the module adjusts the weights by comparing the predicted and observed states and using the Kalman gain calculation formula to achieve dynamic coupling of the two types of data. This mechanism helps to improve the prediction correction capability and makes the fusion results robust when facing high-frequency disturbances and low-frequency offsets.

[0025] The formula for calculating Kalman gain is as follows:

[0026]

[0027] Kalman gain, Predict the covariance matrix. , Measure the noise covariance matrix;

[0028] Principle Explanation: Based on the prediction-observation residual, the gain weight is adjusted to improve the short-time performance of multimodal observations.

[0029] Stability of long-term data fusion;

[0030] Technical effect: Reduces oscillations and delays in dynamic registration, resulting in smooth and continuous pose output;

[0031] Source: Classical Kalman filter algorithm (Kalman, 1960).

[0032] Preferably, the specific steps for adjusting the environmental perception-driven weights in S3 are as follows:

[0033] S31 Environmental Status Monitoring and Interference Identification: The environmental sensing module monitors in real time the obstruction, magnetic field fluctuations and surface reflectivity changes in the optical path. By analyzing these parameter changes, it assesses the possible interference to optical and electromagnetic data, providing a dynamic evaluation basis for modal data quality.

[0034] S32 Confidence-Based Dynamic Weight Adjustment: Based on the interference assessment results, the module calculates the current confidence level of each mode and feeds the calculation results back to the sliding fusion unit to dynamically adjust the weighting ratio of optical and electromagnetic data in the fusion process, so as to improve the overall registration accuracy and system robustness.

[0035] Preferably, the specific steps of the multi-level interference closed-loop self-calibration in S4 are as follows:

[0036] S41 Residual Monitoring and Outlier Removal: The registration calculation module continuously monitors the registration residual. When the error exceeds the preset threshold, it first identifies and removes outliers at the sensor level to prevent local distorted data from amplifying and interfering with the overall registration calculation, thus ensuring the quality of the data source.

[0037] S42 Covariance Adaptation and Mode Switching: At the algorithm and system level, the module uses the covariance matrix adaptive update formula to dynamically adjust the filter measurement noise parameters and switch the main data mode when necessary to enhance the system's self-correction capability in interference environments and optimize the final registration output.

[0038] The adaptive update formula for the covariance matrix is:

[0039]

[0040] In the formula, Updated measurement noise covariance matrix. Original measurement noise covariance matrix, Observe the residual vector. Smoothing coefficient (0~1);

[0041] Principle explanation: When a surge in error is detected, the noise model is automatically adjusted and the dominant modal source is switched to improve the system's self-recovery capability;

[0042] Technical benefits: Adaptively responds to sudden environmental changes or sensor malfunctions, ensuring system robustness;

[0043] Source: Adaptive Kalman Filter.

[0044] Preferably, the specific steps for output and dynamic deformation compensation in S5 are as follows:

[0045] S51 Optimized Pose Data Output and Transmission: The system transmits the registration pose data, which has been processed and optimized through multiple levels, to the surgical robot control system through a low-latency output interface, ensuring that the pose information can be accurately and efficiently synchronized with the main control system in a highly dynamic surgical environment.

[0046] S52 Deformation Compensation Based on Anatomical Model: The deformation compensation module combines the preoperative anatomical model with the pose data collected during the operation. It uses non-rigid deformation field estimation and iterative matching correction algorithm to effectively correct the error caused by the dynamic deformation of soft tissue, and realize real-time optimization and dynamic accurate compensation of the registration results.

[0047] Preferably, a dual-mode registration system for optical positioning and electromagnetic navigation of a surgical robot is provided, the system being based on the above-described method, and comprising:

[0048] Dual-modal data acquisition and dynamic registration module: including optical positioning device (sampling frequency ≥60Hz), electromagnetic navigation device (sampling frequency ≤20Hz), timestamp synchronization unit, sliding fusion and Kalman filter components, used to realize synchronous alignment and dynamic coupling registration of high and low frequency pose data;

[0049] Environmental perception and adaptive interference correction module: including occlusion detection, magnetic field interference detection, reflectivity monitoring, trust level calculator, and covariance matrix adaptive adjustment unit, used to dynamically adjust mode weights and close-loop error correction;

[0050] Deformation compensation and system integration module: including non-rigid deformation compensation unit, B-spline-based free deformation field estimation module, and output interface (delay ≤10ms), used to correct soft tissue dynamic deformation errors and transmit optimized pose data to the surgical robot control system with low latency.

[0051] Preferably, the dual-modal data acquisition and dynamic registration module includes:

[0052] (1) Dual-modal positioning and data acquisition: The system includes an optical positioning device and an electromagnetic navigation device, which are responsible for acquiring high-frequency local pose data and low-frequency global pose information, respectively. The two work together to cover visible and invisible areas in the surgical environment, providing a comprehensive and complementary data foundation for subsequent registration.

[0053] (2) Time synchronization and dynamic coupling registration: The integrated timestamp synchronization unit, combined with sliding fusion and Kalman filtering components, first completes the time alignment of multi-source data, and then generates real-time optimized pose registration results through dynamic coupling of prediction and observation information, thereby improving the registration accuracy and stability of the system in complex scenarios.

[0054] Preferably, the environmental perception and adaptive interference correction module includes:

[0055] (1) Environmental monitoring and modal trust assessment: The module includes occlusion detection, magnetic field interference detection and reflectivity monitoring units. By collecting environmental status in real time and combining it with the modal trust calculation formula, the current reliability of optical and electromagnetic data is assessed, providing a dynamic and quantitative trust basis for subsequent weighted adjustments.

[0056] The formula for calculating modal trust is:

[0057]

[0058] In the formula, :Modal Trust level (0~1). :Modal Current residual error, :Preset maximum allowable error;

[0059] Principle explanation: The weight of each mode is dynamically adjusted based on its real-time error, with priority given to stable modes;

[0060] Technical effect: Automatically avoids modal data with severe occlusion or interference, ensuring registration accuracy;

[0061] Source: Multi-sensor fusion weighting methods (e.g., Madgwick, Mahony filter weight allocation strategies).

[0062] (2) Dynamic weight adjustment and error correction: Based on the trust evaluation results, the system dynamically modifies the fusion weight of each mode through the covariance matrix adaptive adjustment unit and performs closed-loop error correction to achieve adaptive response to complex environmental interference and ensure that the registration output has high stability and high accuracy.

[0063] Preferably, the deformation compensation and system integration module includes:

[0064] (1) Deformation field estimation based on B-spline: The non-rigid deformation compensation unit and the free deformation field estimation module in the system are used together. By using the free deformation field interpolation formula based on B-spline, the preoperative anatomical model and intraoperative pose data are fitted to accurately estimate the local and global deformation of soft tissue in the dynamic process.

[0065] The interpolation formula for the free deformation field is as follows (based on 3D B-splines):

[0066]

[0067] In the formula, Deformation displacement at any location , , B-spline basis functions (along the u, v, and w directions respectively). Control point offset, Normalized spatial coordinates;

[0068] Principle explanation: By fitting a non-rigid deformation field through a three-dimensional B-spline mesh, high-precision compensation for dynamic changes in soft tissue can be achieved.

[0069] Technical effect: Improves the ability to correct posture errors caused by dynamic deformation during surgery;

[0070] Source: Medical image registration, free-form deformation method in computer graphics (Bookstein, 1989);

[0071] (2) Optimize pose output and low-latency transmission: After deformation correction is completed, the system transmits the optimized pose data to the surgical robot control system in a low-latency form through the output interface to realize real-time response and control during the operation, ensuring high precision, stability and safety of the surgical operation.

[0072] The beneficial effects of this invention are as follows:

[0073] 1. This invention effectively integrates high-frequency local pose data and low-frequency global pose data through a dual-modal registration method combining optical positioning and electromagnetic navigation. This avoids measurement failures caused by occlusion or magnetic interference in a single modality. The invention uses a timestamp synchronization unit to achieve time alignment of multi-source data. By combining sliding fusion and Kalman filtering algorithms to dynamically couple short-time and long-time window data, the registration results maintain high accuracy and stability in dynamic surgical environments, which is significantly better than existing single-modal or simple fusion methods.

[0074] 2. This invention, through its designed environmental perception and adaptive interference correction module, can monitor optical obstruction, magnetic field interference, and reflectivity changes in real time, calculate the confidence level of each mode, and dynamically adjust the data weighting ratio. When the system detects that the residual exceeds the threshold, it has a multi-level closed-loop self-correction function, including outlier removal, covariance adaptive adjustment, and mode dominance switching. Robustness optimization is achieved from the sensor level to the system level, which greatly improves the adaptability and error suppression capability in complex environments.

[0075] 3. This invention uses a non-rigid deformation compensation module, combined with preoperative anatomical models and intraoperative pose data, and a free deformation field estimation method based on B-splines to correct registration errors caused by dynamic deformation of soft tissue in real time. The optimized pose data is transmitted to the surgical robot control system through a low-latency output interface, achieving high-precision, low-latency dynamic compensation during surgery, which significantly improves the safety, reliability and clinical application value of robot operation. Attached Figure Description

[0076] Figure 1This is a flowchart of the dual-mode registration method for optical positioning and electromagnetic navigation of the surgical robot of the present invention;

[0077] Figure 2 This is a flowchart of the dual-mode registration system for optical positioning and electromagnetic navigation of the surgical robot of the present invention. Detailed Implementation

[0078] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0079] like Figures 1 to 2 As shown in the figure, this embodiment of the invention provides a dual-mode registration method for optical positioning and electromagnetic navigation of a surgical robot. The specific steps of this dual-mode registration method are as follows:

[0080] S1 Data Acquisition and Time Synchronization: The optical positioning module acquires high-frequency three-dimensional pose data, the electromagnetic navigation module acquires low-frequency global pose data, and the registration calculation module calculates the time difference between the optical and electromagnetic data through the timestamp synchronization unit and aligns and synchronizes them to ensure the time consistency of the dual-modal data stream.

[0081] S2 Dynamic Coupled Registration: The multi-time window sliding fusion unit of the registration calculation module processes optical data based on a short time window and electromagnetic data based on a long time window. It combines the Kalman filter algorithm to adjust the gain of the predicted state and the observed state, dynamically couples the two-mode data, and generates preliminary registration results.

[0082] S3 Environmental Perception Drives Weight Adjustment: The environmental perception and adaptive interference correction module monitors optical occlusion, magnetic field interference and reflectivity changes in real time, calculates the confidence level of each mode, and feeds it back to the sliding fusion unit to dynamically adjust the weighting ratio of optical and electromagnetic data and improve fusion stability.

[0083] S4 Multi-level Interference Closed-Loop Self-Correction: When the registration calculation module detects that the residual error exceeds the preset threshold, it starts multi-level closed-loop self-correction, including sensor-level outlier removal, algorithm-level adaptive adjustment of filter measurement noise covariance matrix, and system-level mode dominance switching, to generate optimized registration results.

[0084] S5 Output and Dynamic Deformation Compensation: The optimized registration pose data is transmitted to the surgical robot control system through the output interface. The deformation compensation module, based on the preoperative anatomical model and intraoperative pose data, uses non-rigid deformation field estimation and iterative matching correction algorithm to correct the registration error caused by the dynamic deformation of soft tissue in real time, thereby achieving accurate dynamic compensation.

[0085] Example 1: Dual-modal registration in a standard surgical scenario

[0086] In neurosurgery, an optical positioning device is fixed to the surgical instruments, while an electromagnetic navigation device is placed at a specific reference point within the patient's body. The system acquires high-frequency optical pose data and low-frequency electromagnetic pose data, aligning the time stream through a timestamp synchronization unit. A sliding fusion unit processes different modal data based on short-time and long-time windows, employing Kalman filtering for dynamic coupling to output high-precision preliminary registration results. An environmental perception module monitors occlusion and magnetic field fluctuations within the surgical area, adjusting modal weights according to confidence levels. When the system detects excessive residuals, it automatically performs outlier removal and adaptive adjustment of covariance parameters, transmitting the optimization results to the main control system. This, combined with a deformation compensation module, corrects dynamic deformation errors in soft tissue.

[0087] Example 2: Adaptive Registration Optimization in Complex Environments

[0088] In thoracoscopic minimally invasive surgery, the optical module intermittently malfunctions due to surgical smoke and reflections, while electromagnetic navigation fluctuates due to interference from surrounding metal. The system relies on an environmental perception module to assess the confidence level of each mode in real time and dynamically adjust the weighting ratios, prioritizing electromagnetic data with less interference. Upon detecting excessive errors, the system activates multi-level closed-loop self-calibration, eliminating abnormal optical data, adaptively adjusting filter measurement noise parameters, and switching to electromagnetic mode dominance when necessary. The corrected pose data is transmitted to the robot control system via the output interface, where the deformation compensation module performs non-rigid registration correction in conjunction with the preoperative model, ensuring surgical precision.

[0089] Example 3: Dynamic Deformation Compensation for Different Soft Tissue Sites

[0090] In laparoscopic surgery, the system first loads the anatomical model established by preoperative CT. During the operation, it combines pose data acquired by optical and electromagnetic means to establish a free deformation field through a non-rigid deformation field estimation module. The deformation compensation module uses B-spline control mesh and control point offset to achieve local dynamic deformation correction. The compensation result is transmitted in real time to the main control system of the surgical robot through a low-latency output interface, enabling precise correction of registration errors caused by soft tissue respiratory movements, pulsation, etc., and meeting the high-precision navigation requirements in complex scenarios.

[0091] In S1, data acquisition and time synchronization refer to using the optical positioning module to acquire high-frequency local three-dimensional pose information, while the electromagnetic navigation module acquires low-frequency global pose data. This ensures coverage of visible and invisible areas in complex surgical environments, providing a multi-scale, multi-source data foundation for subsequent dual-modal registration. The registration calculation module has a built-in timestamp synchronization unit that evaluates the time difference between optical and electromagnetic data based on the time synchronization error calculation formula, completing time alignment and ensuring that multi-modal data streams are processed collaboratively under the same time reference, laying a consistent foundation for dynamic coupling registration.

[0092] In S2, dynamic coupling registration refers to the multi-time-window sliding fusion unit using a short-time window to process optical data to capture high-frequency dynamic changes and a long-time window to process electromagnetic data to extract steady-state global pose, forming a hierarchical processing framework that lays the data foundation for the stable fusion of subsequent multimodal data. During the fusion process, the module adjusts the weights by comparing the predicted and observed states and using the Kalman gain calculation formula to achieve dynamic coupling of the two types of data. This mechanism helps to improve the prediction correction capability and makes the fusion results robust in the face of high-frequency disturbances and low-frequency offsets.

[0093] In S3, the environmental perception-driven weight adjustment refers to the environmental perception module monitoring the occlusion, magnetic field fluctuations, and surface reflectivity changes in the optical path in real time. By analyzing these parameter changes, it assesses the potential interference to optical and electromagnetic data, providing a dynamic evaluation basis for modal data quality. Based on the interference evaluation results, the module calculates the current confidence level of each modality and feeds the calculation results back to the sliding fusion unit to dynamically adjust the weighting ratio of optical and electromagnetic data in the fusion process, thereby improving the overall registration accuracy and system robustness.

[0094] In S4, the multi-level interference closed-loop self-correction refers to the registration calculation module continuously monitoring the registration residual. When the error exceeds the preset threshold, it first identifies and removes outliers at the sensor level to prevent local distorted data from amplifying the interference to the overall registration calculation and ensuring the quality of the data source. At the algorithm and system level, the module uses the covariance matrix adaptive update formula to dynamically adjust the filter measurement noise parameters and switch the main data mode when necessary to enhance the system's self-correction capability under interference environment and optimize the final registration output.

[0095] In S5, output and dynamic deformation compensation refer to the system transmitting the registration pose data, which has been processed and optimized through multiple levels, to the surgical robot control system through a low-latency output interface. This ensures that the pose information can be accurately and efficiently synchronized with the main control system in a highly dynamic surgical environment. The deformation compensation module combines the preoperative anatomical model with the pose data collected during the operation, and uses non-rigid deformation field estimation and iterative matching correction algorithms to effectively correct errors caused by the dynamic deformation of soft tissue, thereby achieving real-time optimization and dynamic accurate compensation of the registration results.

[0096] This embodiment provides a dual-mode registration system for optical positioning and electromagnetic navigation of a surgical robot based on the above method, including:

[0097] Dual-modal data acquisition and dynamic registration module: including optical positioning device, electromagnetic navigation device, timestamp synchronization unit, sliding fusion and Kalman filter components, used to realize synchronous alignment and dynamic coupling registration of high and low frequency pose data;

[0098] Environmental perception and adaptive interference correction module: including occlusion detection, magnetic field interference detection, reflectivity monitoring, trust level calculator, and covariance matrix adaptive adjustment unit, used to dynamically adjust mode weights and close-loop error correction;

[0099] Deformation compensation and system integration module: including non-rigid deformation compensation unit, B-spline-based free deformation field estimation module, and output interface, used to correct soft tissue dynamic deformation errors and transmit optimized pose data to the surgical robot control system with low latency.

[0100] The dual-modal data acquisition and dynamic registration module refers to the system including an optical positioning device and an electromagnetic navigation device, which are responsible for acquiring high-frequency local pose data and low-frequency global pose information, respectively. The two work together to cover visible and invisible areas in the surgical environment, providing a comprehensive and complementary data foundation for subsequent registration. The integrated timestamp synchronization unit, combined with sliding fusion and Kalman filtering components, first completes the time alignment of multi-source data, and then generates real-time optimized pose registration results through dynamic coupling of prediction and observation information, improving the registration accuracy and stability of the system in complex scenarios.

[0101] The environmental perception and adaptive interference correction module includes occlusion detection, magnetic field interference detection, and reflectivity monitoring units. By collecting environmental data in real time and combining it with the modal confidence calculation formula, it assesses the current reliability of optical and electromagnetic data, providing a dynamic and quantitative basis for subsequent weighted adjustments. Based on the confidence assessment results, the system dynamically modifies the fusion weights of each mode through the covariance matrix adaptive adjustment unit and performs closed-loop error correction, achieving an adaptive response to complex environmental interference and ensuring that the registration output has high stability and high accuracy.

[0102] The deformation compensation and system integration module refers to the combined use of the non-rigid deformation compensation unit and the free deformation field estimation module in the system. By using the free deformation field interpolation formula based on B-splines, the preoperative anatomical model and intraoperative pose data are fitted to accurately estimate the local and global deformation of soft tissue during the dynamic process. After completing the deformation correction, the system transmits the optimized pose data to the surgical robot control system in a low-latency form through the output interface, realizing real-time response and control during the operation and ensuring high precision, stability and safety of the surgical operation.

[0103] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0104] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A dual-mode registration method for optical positioning and electromagnetic navigation of a surgical robot, characterized in that: The specific steps of the dual-mode registration method for optical positioning and electromagnetic navigation of this surgical robot are as follows: S1 Data Acquisition and Time Synchronization: The system acquires high-frequency three-dimensional pose data through the optical positioning module and low-frequency global pose data through the electromagnetic navigation module. It also uses the timestamp synchronization unit to calculate the time difference, thereby achieving time alignment of the two types of data and ensuring the synchronization and consistency of the dual-modal data stream. S2 Dynamic Coupled Registration: The multi-time-window sliding fusion unit uses a short-time window to process optical data and a long-time window to process electromagnetic data. It combines the Kalman filter algorithm to adjust the gain of the prediction and observation states, dynamically couples the two-mode data, and outputs the preliminary registration results. S3 Environmental Perception Drives Weight Adjustment: The environmental perception and interference correction module monitors optical occlusion, magnetic field interference, and reflectivity changes in real time, calculates the confidence level of optical and electromagnetic modes, and feeds it back to the sliding fusion unit to dynamically adjust the weight of each mode data and improve registration stability. S4 Multi-level Interference Closed-loop Self-correction: When the registration residual is detected to exceed the threshold, the system starts closed-loop self-correction, including sensor-level outlier removal, algorithm-level noise parameter adaptive adjustment, and system-level mode-dominant switching, to generate optimized high-precision registration results. S5 Output and Dynamic Deformation Compensation: The optimized registration pose data is transmitted to the surgical robot control system through the output interface. The deformation compensation module performs non-rigid deformation estimation and correction based on the preoperative anatomical model and intraoperative pose, realizing dynamic compensation and precise control of soft tissue.

2. The dual-mode registration method for optical positioning and electromagnetic navigation of a surgical robot according to claim 1, characterized in that: The specific steps for data acquisition and time synchronization in S1 are as follows: S11 high-frequency and low-frequency pose data acquisition: acquires optical three-dimensional pose data with a frequency of greater than or equal to 60Hz and electromagnetic global pose data with a frequency of less than or equal to 20Hz, and aligns them through a timestamp synchronization unit to control the synchronization error to no more than 20ms, ensuring coverage of visible and invisible areas in complex surgical environments. S12 Time Synchronization and Consistency Calibration: The registration calculation module has a built-in timestamp synchronization unit. Based on the time synchronization error calculation formula, it evaluates the time difference between optical and electromagnetic data, completes time alignment, and ensures that multimodal data streams are processed collaboratively under the same time reference, laying the foundation for consistency conditions for dynamic coupling registration.

3. The dual-mode registration method for optical positioning and electromagnetic navigation of a surgical robot according to claim 1, characterized in that: The specific steps of dynamic coupling registration in S2 are as follows: S21 Multi-Time Window Layered Processing: The multi-time window sliding fusion unit uses a short time window to process optical data to capture high-frequency dynamic changes, and a long time window to process electromagnetic data to extract steady-state global pose, forming a layered processing framework that lays the data foundation for the stable fusion of subsequent multimodal data. S22 Predictive Gain-Driven Dynamic Coupling: During the fusion process, the module compares the predicted and observed states, and based on the short-term window optical data and the long-term window electromagnetic data, combines a Kalman filter and uses a residual-driven covariance adaptive update formula to dynamically adjust the fusion gain, generating preliminary registration results.

4. The dual-mode registration method for optical positioning and electromagnetic navigation of a surgical robot according to claim 1, characterized in that: The specific steps for adjusting the environmental perception-driven weights in S3 are as follows: S31 Environmental Status Monitoring and Interference Identification: The environmental sensing module monitors in real time the obstruction, magnetic field fluctuations and surface reflectivity changes in the optical path. By analyzing these parameter changes, it assesses the interference to optical and electromagnetic data. The S32 monitors optical occlusion rate, magnetic field disturbance amplitude and modal residuals in real time, calculates modal confidence level, and dynamically adjusts the weighting ratio of modal data based on the confidence level.

5. The dual-mode registration method for optical positioning and electromagnetic navigation of a surgical robot according to claim 1, characterized in that: The specific steps of the multi-level interference closed-loop self-calibration in S4 are as follows: S41 Residual Monitoring and Outlier Removal: The registration calculation module continuously monitors the registration residual. When the error exceeds the preset threshold, it first identifies and removes outliers at the sensor level to prevent local distorted data from amplifying and interfering with the overall registration calculation, thus ensuring the quality of the data source. When the registration residual exceeds the preset threshold three times in a row, S42 sequentially performs outlier removal at the sensor layer, adaptive update of the measurement covariance matrix at the algorithm layer, and mode-dominant switching at the system layer to complete closed-loop multi-level self-calibration.

6. The dual-mode registration method for optical positioning and electromagnetic navigation of a surgical robot according to claim 1, characterized in that: The specific steps for output and dynamic deformation compensation in S5 are as follows: S51 Optimized Pose Data Output and Transmission: The system transmits the registration pose data, which has been processed and optimized through multiple levels, to the surgical robot control system through a low-latency output interface, ensuring that the pose information can be accurately and efficiently synchronized with the main control system in a highly dynamic surgical environment. S52 Deformation Compensation Based on Anatomical Model: The deformation compensation module combines the preoperative anatomical model with the pose data acquired during the operation. Based on the preoperative CT anatomical model and the pose data during the operation, it constructs a B-spline deformation field to estimate and compensate for the non-rigid dynamic deformation of soft tissue. The corrected registration pose is then output to the surgical robot control system with a delay of less than or equal to 10ms.

7. A dual-mode registration system for surgical robots, combining optical positioning and electromagnetic navigation, characterized in that: The surgical robot's dual-mode registration system, combining optical positioning and electromagnetic navigation, is based on the method described in claims 1-6, and the system comprises: Dual-modal data acquisition and dynamic registration module: including optical positioning device, electromagnetic navigation device, timestamp synchronization unit, sliding fusion and Kalman filter components, used to realize synchronous alignment and dynamic coupling registration of high and low frequency pose data; Environmental perception and adaptive interference correction module: including occlusion detection, magnetic field interference detection, reflectivity monitoring, trust level calculator, and covariance matrix adaptive adjustment unit, used to dynamically adjust mode weights and close-loop error correction; Deformation compensation and system integration module: including non-rigid deformation compensation unit, B-spline-based free deformation field estimation module, and output interface, used to correct soft tissue dynamic deformation errors and transmit optimized pose data to the surgical robot control system with low latency.

8. The dual-mode registration system for optical positioning and electromagnetic navigation of a surgical robot according to claim 7, characterized in that: The dual-modal data acquisition and dynamic registration module includes: (1) Dual-modal positioning and data acquisition: The system includes an optical positioning device and an electromagnetic navigation device, which are responsible for acquiring high-frequency local pose data and low-frequency global pose information, respectively. The two work together to cover visible and invisible areas in the surgical environment, providing a comprehensive and complementary data foundation for subsequent registration. (2) Time synchronization and dynamic coupling registration: The integrated timestamp synchronization unit, combined with sliding fusion and Kalman filtering components, first completes the time alignment of multi-source data, and then generates real-time optimized pose registration results through dynamic coupling of prediction and observation information, thereby improving the registration accuracy and stability of the system in complex scenarios.

9. The dual-mode registration system for optical positioning and electromagnetic navigation of a surgical robot according to claim 7, characterized in that: The environmental perception and adaptive interference correction module includes: (1) Environmental monitoring and modal trust assessment: The module includes occlusion detection, magnetic field interference detection and reflectivity monitoring units. By collecting environmental status in real time and combining it with the modal trust calculation formula, the current reliability of optical and electromagnetic data is assessed, providing a dynamic and quantitative trust basis for subsequent weighted adjustments. (2) Dynamic weight adjustment and error correction: Based on the trust evaluation results, the system dynamically modifies the fusion weight of each mode through the covariance matrix adaptive adjustment unit and performs closed-loop error correction to achieve adaptive response to complex environmental interference and ensure that the registration output has high stability and high accuracy.

10. The dual-mode registration system for optical positioning and electromagnetic navigation of a surgical robot according to claim 7, characterized in that: The deformation compensation and system integration module includes: (1) Deformation field estimation based on B-spline: The non-rigid deformation compensation unit and the free deformation field estimation module in the system are used together. By using the free deformation field interpolation formula based on B-spline, the preoperative anatomical model and intraoperative pose data are fitted to accurately estimate the local and global deformation of soft tissue in the dynamic process. (2) Optimize pose output and low-latency transmission: After deformation correction is completed, the system transmits the optimized pose data to the surgical robot control system in a low-latency form through the output interface to realize real-time response and control during the operation, ensuring high precision, stability and safety of the surgical operation.

Images