Multimodal perception based on interoception integrated orthopedic surgery navigation positioning method

Abstract

This application relates to the field of orthopedic surgical navigation, and more particularly to a multimodal sensing-based orthopedic surgical navigation and positioning method based on integrated sensing. The method includes: simultaneously acquiring UWB ranging data and laser ranging data; evaluating the communication link status; classifying the UWB ranging data into levels based on the evaluation results to obtain effective UWB ranging data after a first screening; applying a perception-driven adaptive weighted Chan algorithm to the effective UWB ranging data after the first screening to obtain an initial position estimate; calculating the residual based on the initial position estimate to obtain effective UWB ranging data after a second screening; and inputting the initial position estimate and the effective UWB ranging data after the second screening into an AMCA-EKF filtering algorithm to output high-precision position coordinates of the surgical instruments. This achieves stable, continuous, and high-precision three-dimensional positioning capabilities.

Description

Technical Field

[0001] This application relates to the field of orthopedic surgical navigation, and in particular to a multimodal sensing orthopedic surgical navigation and positioning method based on synesthesia integration. Background Technology

[0002] With the continuous development of technologies such as computer vision, spatial positioning, and biomechanical modeling, orthopedic surgical navigation and positioning systems are playing an increasingly important role in assisting doctors to achieve precise and minimally invasive operations. Traditional orthopedic surgery relies heavily on the doctor's experience and intraoperative visual positioning, resulting in problems such as large trauma, high positioning errors, and slow postoperative recovery. Orthopedic surgical navigation and positioning systems integrate medical imaging, spatial positioning, and computer-aided computing technologies to achieve real-time display and comparison of preoperative image acquisition and 3D reconstruction, as well as intraoperative navigation, providing doctors with intuitive and visual operational guidance. By spatially registering preoperative CT images with the patient's skeletal structure during surgery, combined with optical, wireless positioning, and inertial sensing technologies, the system tracks the position of surgical instruments in real time and displays it synchronously in the navigation interface. Doctors can plan and execute operations in a virtual and real 3D environment. Surgical navigation systems not only improve the positioning accuracy and operational safety of surgery but also reduce surgical trauma and operation time, promote rapid postoperative recovery, and thus improve surgical safety and outcomes.

[0003] With the rapid development of photoelectric sensing, spatial ranging, and image recognition technologies, optical positioning technology is increasingly widely used in the field of medical navigation. Optical positioning surgical navigation systems, as a key extension of computer-assisted surgery (CAS), use infrared or visible light cameras to capture optical markers mounted on the surfaces of surgical instruments and the patient's bones. Combined with spatial geometric reconstruction algorithms, they calculate the three-dimensional coordinates in real time, enabling high-precision real-time positioning and tracking of bone structures during surgery. However, optical positioning systems have limited adaptive correction capabilities for optical positioning errors, making them prone to positioning loss or low positioning accuracy.

[0004] With the development of precision medicine and intelligent surgery, and due to its advantages of high precision and strong anti-interference capabilities, UWB positioning technology has been introduced into the field of medical navigation. UWB systems achieve millimeter-level spatial positioning accuracy by emitting extremely short pulse signals and measuring their propagation time, making them suitable for the complex environment of operating rooms. However, most existing positioning systems are not optimized for the complex electromagnetic environment of operating rooms, making them susceptible to factors such as metal reflection and human body obstruction, leading to fluctuations or loss of positioning data. Furthermore, existing systems often employ a single algorithm, making it difficult to achieve continuous, high-precision positioning in dynamic surgical scenarios.

[0005] In orthopedics, the treatment of conditions such as spinal deformities, fracture reduction, and joint replacement places extremely high demands on spatial positioning accuracy. Taking scoliosis correction and vertebral screw implantation as examples, surgeons need to precisely determine the position and angle of the pedicles within complex anatomical structures during surgery; even the slightest deviation can lead to nerve damage or structural instability. However, current orthopedic navigation and positioning systems still have certain limitations. Obstructions in the operating room often lead to signal loss or misinterpretation; environments with dense metal surgical instruments are prone to interference, affecting positioning accuracy and making it difficult to maintain high precision over long periods; automatic registration of image data and intraoperative coordinate systems is still imperfect; and some systems are bulky and complex to deploy, limiting their widespread application in complex orthopedic surgical scenarios. Summary of the Invention

[0006] To address the problems in the prior art, this application provides a multimodal sensing orthopedic surgical navigation and positioning method based on integrated sensing.

[0007] This application provides a multimodal sensing orthopedic surgical navigation and positioning method based on synesthesia integration, including: Simultaneously acquire UWB ranging data and laser ranging data; The time jitter perception factor and energy first path ratio perception factor are extracted from UWB ranging data, and the communication link status is evaluated based on the comparison results of the time jitter perception factor and energy first path ratio perception factor with preset thresholds. Based on the evaluation results, the UWB ranging data are classified into levels, and the participation level of the UWB ranging data is determined according to the level, resulting in the effective UWB ranging data after the first screening. The selected valid UWB ranging data are used to obtain the initial position estimate by the perception-driven adaptive weighted Chan algorithm. The residual is calculated based on the initial position estimate, which is the cumulative value of the squared difference between the actual measured distance difference and the theoretically calculated distance difference. The UWB ranging data is then classified into levels based on a comparison of the residual with a residual threshold. The level determines the degree of participation of the UWB ranging data, resulting in effective UWB ranging data after a second round of filtering. The initial position estimate and the effective UWB ranging data after the second round of filtering are input into the AMCA-EKF filtering algorithm. The measurement noise covariance is dynamically adjusted based on the time jitter sensing factor and the energy first diameter ratio sensing factor, ultimately outputting the high-precision position coordinates of the surgical instrument.

[0008] Furthermore, the extraction of the time jitter perception factor includes: within a preset time window, performing multiple consecutive two-way ranging communications with the i-th base station, and recording the round-trip propagation delay measured in each communication process; Calculate the average of all round-trip propagation delays within the time window; Calculate the absolute value of the difference between the average propagation delay for each round trip; sum these absolute differences and divide by the number of communications to obtain the time jitter perception factor.

[0009] Furthermore, the extraction of the energy first path ratio sensing factor includes: for the communication signal received by the i-th base station, extracting the received energy of the first path signal and the total received energy of all path signals in the entire communication frame; The ratio of the received energy of the first path signal to the total received energy is the energy first path ratio sensing factor.

[0010] Furthermore, the levels include high-confidence data, low-confidence data, and abnormal data. High-confidence data directly participates in the location calculation, low-confidence data participates in the calculation with reduced weight, and abnormal data is suppressed or eliminated.

[0011] Furthermore, the method also includes determining the current environmental state based on the values ​​of the time jitter perception factor and the energy first-path ratio perception factor, and executing a differentiated laser information-assisted processing strategy accordingly, including: Reference threshold consistency verification under unobstructed and stable communication link conditions: Using laser ranging data as a reference threshold, the UWB ranging data of each base station is compared with the reference threshold; if the deviation between the UWB ranging data and the reference threshold is within a preset allowable range, the UWB ranging data is determined to pass the consistency verification and is marked as valid UWB ranging data; if the deviation between the UWB ranging data and the reference threshold exceeds the preset allowable range, the verification is determined to fail, and the UWB ranging data is discarded or its weight in positioning is reduced.

[0012] Furthermore: the current environmental state is determined based on the values ​​of the time jitter perception factor and the energy first path ratio perception factor, and a differentiated laser information-assisted processing strategy is executed accordingly. It also includes: when human body obstruction, metal reflection or multipath interference is detected, the deviation between the UWB ranging data of each base station and the reference threshold is calculated, and the UWB ranging data is divided into severely abnormal measurements, low-confidence measurements and valid measurements according to the magnitude of the deviation.

[0013] Furthermore: The filtered valid UWB ranging data are processed using a perception-driven adaptive weighted Chan algorithm to obtain initial position estimates, including: A system of nonlinear equations is established based on the distance from each base station to the positioning tag; After vectorizing the nonlinear equations, a linear equation is obtained. The unknowns in the linear equation include the three-dimensional coordinates of the positioning tag and distance-related terms. The coefficient matrix of the linear equation is calculated from the measured distance difference and the base station coordinates. The weights are obtained based on the relationship between the UWB ranging data of each base station and the positioning tag and the reference threshold; All weights are constructed into a weight diagonal matrix. This weight diagonal matrix is ​​introduced into the weighted least squares solution of the perception-driven adaptive weighted Chan algorithm. The first estimate is obtained by using the weighted least squares solution. The weighted matrix is ​​recalculated based on the first estimate. The weighted matrix is ​​then used to replace the weighted diagonal matrix. The second estimate is obtained by solving the linear least squares method after the second weighting.

[0014] Furthermore: the weights are: , in, Indicates the first UWB ranging results for each base station Indicates the laser ranging distance. For time jitter perception factor, The energy first-path ratio sensing factor. All are sensory regulation coefficients. This indicates the reference threshold.

[0015] Furthermore: The initial position estimate and the effective UWB ranging data after the second filtering are jointly input into the AMCA-EKF filtering algorithm, including: Establish the state prediction equation: predict the state at the current time based on the estimated state at the previous time step, and obtain the predicted state at time k. Establish measurement equations: Establish the mathematical relationship between the predicted state and the measurement value at the current time, and determine the measurement matrix; Calculate the penalty amount: Compare the UWB ranging data of each base station with the reference threshold, and calculate the penalty amount based on the difference between the two. Dynamic modulation measurement noise covariance: When it is determined that the UWB ranging data exceeds the parameter threshold, the measurement noise covariance is dynamically increased according to the time jitter sensing factor and energy first path ratio sensing factor in the communication sensing parameters; when it does not exceed the parameter threshold, the basic measurement noise covariance remains unchanged. Calculate the Kalman gain: Calculate the Kalman gain based on the measured noise covariance and prior covariance; Based on the calculated Kalman gain and prior covariance, the posterior covariance is calculated, and the posterior covariance is used as the prior covariance for the state prediction at the next time step. Based on the state prediction equation, the predicted state is corrected using Kalman gain, and the estimated state vector at time k is obtained as the final positioning coordinates.

[0016] Compared with the prior art, the advantages of this application are: This application enables UWB wireless communication to simultaneously perform both spatial sensing and link status assessment functions. Based on this concept, UWB measurement data is filtered using a laser ranging threshold, and combined with the rapid initial localization of the perception-driven adaptive weighted Chan (PAW-Chan) algorithm and the dynamic correction of nonlinear errors by extended Kalman filtering, stable, continuous, and high-precision three-dimensional localization capabilities are achieved in surgical scenarios with dense metal instruments and frequent obstructions. It can perceive changes in ranging reliability in real time in complex environments such as surgical obstructions and metal reflections, and dynamically adjust the localization calculation strategy, giving the system excellent continuity and stability, as well as good environmental adaptability and instrument scalability. It can be widely applied in high-precision orthopedic surgeries such as spinal correction, fracture reduction, and joint replacement. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall system framework used in the embodiments of this application; Figure 2 A flowchart illustrating a multimodal sensing orthopedic surgical navigation and positioning method based on synesthesia integration, provided in this application embodiment; Figure 3 This is a schematic diagram showing the comparison of positioning errors provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. It should be noted that the described embodiments are only a part of this application, and not all embodiments. Other embodiments obtained by those skilled in the art based on the technical solutions of this application without creative effort are all within the protection scope of this application.

[0019] See Figure 1 As shown, this application deploys four base stations, each equipped with a UWB wireless positioning module (ultra-wideband wireless communication technology). Each UWB wireless positioning module has a corresponding power supply unit to power it. The UWB wireless positioning modules installed on the base stations adopt the same structure. The surgical instruments are equipped with positioning tags with built-in inertial measurement units that communicate wirelessly with the base stations and laser ranging modules are deployed in the environment. The specific deployment needs to be based on the actual surgical procedure.

[0020] Both the base station and the positioning tag use the same UWB wireless positioning module, with the base station... As a reference base station, a wireless synchronization mechanism is used to synchronize other base stations with the base station. The system is calibrated to the same time reference, ensuring that all base stations start at the same time and remain synchronized. The surgical instruments are equipped with positioning tags that communicate with the base stations. The UWB wireless positioning module includes a signal processing module and a main control MCU.

[0021] The UWB wireless positioning module is configured to simultaneously output communication parameters for link sensing in the bidirectional ranging communication frame, which meets the requirements of the required wireless communication. The electrical connection between the main control MCU and the signal processing module is a bidirectional electrical connection. The main control MCU runs a firmware module to realize the statistics and discrimination of communication sensing parameters, ranging reliability mapping, dynamic weight generation, and online modulation of measurement noise covariance.

[0022] The positioning tags on the surgical instruments can communicate with the base station in real time. The motion state information collected by the inertial measurement unit and the communication sensing parameters are used together to distinguish between time delay fluctuations caused by instrument movement and communication link anomalies caused by blockage or multipath.

[0023] The laser ranging module based on the ATK-MS53L2M chip works in conjunction with the UWB wireless positioning module through a 1*6 ribbon cable. The distance measured by the laser ranging module is used as the spatial constraint threshold driven by communication sensing. The distance information measured by the laser ranging module is the straight-line distance from the laser ranging module to the positioning tag, and the laser ranging module and the UWB wireless positioning module are at the same position and height.

[0024] Please see Figure 1 This application provides a multimodal sensing-based orthopedic surgical navigation and positioning method, comprising the following steps: Simultaneously acquire UWB ranging data and laser ranging data; The time jitter perception factor and energy first path ratio perception factor are extracted from UWB ranging data, and the communication link status is evaluated based on the comparison results of the time jitter perception factor and energy first path ratio perception factor with preset thresholds. Based on the evaluation results, the UWB ranging data are classified into levels, and the participation level of the UWB ranging data is determined according to the level, resulting in the effective UWB ranging data after the first screening. The effective UWB ranging data after the first screening are used to obtain the initial position estimate by the perception-driven adaptive weighted Chan algorithm. The residual is calculated based on the initial position estimate, which is the cumulative value of the squared difference between the actual measured distance difference and the theoretically calculated distance difference. The UWB ranging data is then classified into levels based on a comparison of the residual with a residual threshold. The level determines the degree of participation of the UWB ranging data, resulting in effective UWB ranging data after a second round of filtering. The initial position estimate and the effective UWB ranging data after the second round of filtering are input into the AMCA-EKF filtering algorithm. The measurement noise covariance is dynamically adjusted based on the time jitter sensing factor and the energy first diameter ratio sensing factor, ultimately outputting the high-precision position coordinates of the surgical instrument.

[0025] In one embodiment, the positioning tag and the base station interact bidirectionally through a unified UWB wireless communication frame. While transmitting positioning data and control information, the communication frame simultaneously acquires timestamp information for time ranging and link status awareness. The distance information measured by the laser ranging module is the straight-line distance from the laser ranging module to the positioning tag, and the laser ranging module and the UWB wireless positioning module are at the same position and height. During UWB positioning, the base station and the positioning tag conduct bidirectional wireless communication and ranging to obtain UWB ranging data. The base station also acquires timestamps, round-trip propagation delays, and link quality parameters generated during communication in real time, enabling online sensing and evaluation of ranging reliability. Specifically, within continuous communication frames, the [number of frames] are statistically analyzed. Round-trip delay sequence of each base station And construct a time jitter perception factor. The extraction of the time jitter perception factor includes: within a preset time window, perform multiple consecutive two-way ranging communications with the i-th base station and record the round-trip propagation delay measured in each communication process. Calculate the average of all round-trip propagation delays within the time window; Calculate the absolute value of the difference between the average propagation delay for each round trip; sum these absolute differences and divide by the number of communications to obtain the time jitter perception factor.

[0026] The formula is expressed as: , in, Indicates the number of times within the preset time window The average round-trip propagation delay of each base station. Indicates the number of communications within the time window. Indicates the number of times within the preset time window The first base station Round-trip propagation delay for each communication, This represents the time jitter perception factor.

[0027] The extraction of the energy first-path ratio sensing factor includes: for the communication signal received by the i-th base station, extracting the received energy of the first-path signal and the total received energy of all path signals within the entire communication frame. The energy first-path ratio sensing factor is expressed as: , in, The energy first-path ratio sensing factor. The received energy of the first-path signal. For total received energy.

[0028] The communication sensing parameters composed of the aforementioned energy first path ratio sensing factor and energy first path ratio sensing factor are used together to characterize the stability, obstruction status, and multipath interference level of the UWB communication link.

[0029] In one embodiment, a time jitter sensing factor and an energy first-path ratio sensing factor are extracted from UWB ranging data, and the communication link status is evaluated based on the comparison results of the time jitter sensing factor and the energy first-path ratio sensing factor with preset thresholds, including: Set preset thresholds. Preset the time jitter perception factor threshold and the energy first path ratio perception factor threshold as reference standards for judging the status of the communication link.

[0030] The extracted time jitter perception factor is compared with the time jitter perception factor threshold, and the energy first path ratio perception factor is compared with the energy first path ratio perception factor threshold. Based on the comparison results, the communication link status is divided into the following levels: the level includes high-confidence data, low-confidence data and abnormal data. High-confidence data directly participates in the positioning calculation, low-confidence data participates in the calculation by reducing the weight, and abnormal data is suppressed or removed.

[0031] Specifically, when the time jitter perception factor is not greater than the time jitter perception factor threshold and the energy first path ratio perception factor is not less than the energy first path ratio perception factor threshold, it is determined to be a stable and reliable state, and the corresponding data is highly reliable data; when the time jitter perception factor is not greater than the time jitter perception factor threshold but the energy first path ratio perception factor is less than the energy first path ratio perception factor threshold, it is determined to be a slight multipath state; when the time jitter perception factor is greater than the time jitter perception factor threshold but the energy first path ratio perception factor is not less than the energy first path ratio perception factor threshold, it is determined to be a time-unstable state; when the time jitter perception factor is greater than the time jitter perception factor threshold and the energy first path ratio perception factor is less than the energy first path ratio perception factor threshold, it is determined to be a severe occlusion or strong multipath state.

[0032] Slightly multipath states or time-instantaneous states are considered low-confidence data, while severely occluded or strong multipath states are considered anomalous data.

[0033] By dynamically evaluating the reliability of UWB ranging data from different base stations, the participation mode of UWB ranging data can be adaptively adjusted according to the communication link status, thereby realizing the perception-driven positioning process by communication signals.

[0034] Furthermore, it also includes determining the current environmental state based on the values ​​of the time jitter perception factor and the energy first-path ratio perception factor, and executing differentiated laser information-assisted processing strategies accordingly, including: Reference threshold consistency verification under unobstructed and stable communication link conditions: Using laser ranging data as a reference threshold, the UWB ranging data of each base station is compared with the reference threshold; if the deviation between the UWB ranging data and the reference threshold is within a preset allowable range, the UWB ranging data is determined to pass the consistency verification and is marked as valid UWB ranging data; if the deviation between the UWB ranging data and the reference threshold exceeds the preset allowable range, the verification is determined to fail, and the UWB ranging data is discarded or its weight in positioning is reduced.

[0035] Furthermore, when human occlusion, metal reflection, or multipath interference is detected, the deviation between the UWB ranging data of each base station and the reference threshold is calculated. Based on the magnitude of the deviation, the UWB ranging data is categorized into severely anomalous measurements, low-confidence measurements, and valid measurements. The specific implementation method for classifying and evaluating UWB ranging data is as follows: A reference threshold is set. The system pre-sets laser ranging data as a reference threshold, which is considered a true distance reference under unobstructed and interference-free conditions. Simultaneously, a first deviation threshold and a second deviation threshold are set, where the first deviation threshold is less than the second deviation threshold.

[0036] The system calculates the deviation for each base station. It compares the UWB ranging data of each base station with a reference threshold and calculates the absolute deviation, which is the absolute value of the UWB ranging value minus the reference threshold. The system also records the positive and negative directions of the deviation, where a positive deviation indicates that the UWB ranging value is greater than the actual distance, and a negative deviation indicates that the UWB ranging value is less than the actual distance.

[0037] When the deviation is not greater than the first deviation threshold, the UWB ranging data of the base station is determined to be a valid measurement. The link status corresponding to this type of measurement is good, the ranging error is within an acceptable range, and it can be used in the positioning calculation with high reliability.

[0038] For measurements deemed severely anomalous, the system sets their weight to a near-zero minimum, or in extreme cases, removes them entirely, making their contribution to subsequent positioning calculations negligible. For measurements deemed low-confidence, the system dynamically calculates their weight based on the magnitude of the deviation and communication sensing parameters, causing the weight to decay exponentially with increasing deviation, while moderately increasing the corresponding measurement noise covariance in the AMCA-EKF algorithm, thereby reducing the impact of this measurement on filter updates. For measurements deemed valid, the system maintains a high weight and a low measurement noise covariance, allowing them to play a dominant role in positioning calculations.

[0039] The above steps are repeated in each ranging update cycle. The hierarchical status and processing parameters of the UWB ranging data of each base station are dynamically adjusted according to the real-time detected occlusion and interference, so as to achieve adaptive response to complex communication environments.

[0040] In one embodiment, the selected valid UWB ranging data is processed using a perception-driven adaptive weighted Chan algorithm to obtain an initial position estimate, including: A system of nonlinear equations is established based on the distance from each base station to the positioning tag; After vectorizing the nonlinear equations, a linear equation is obtained. The unknowns in the linear equation include the three-dimensional coordinates of the positioning tag and distance-related terms. The coefficient matrix of the linear equation is calculated from the measured distance difference and the base station coordinates. The weights are obtained based on the relationship between the UWB ranging data of each base station and the positioning tag and the reference threshold; All weights are constructed into a weight diagonal matrix. This weight diagonal matrix is ​​introduced into the linear least squares solution of the perception-driven adaptive weighted Chan algorithm. The first estimate is obtained by using the weighted linear least squares method. The weighted matrix is ​​recalculated based on the first estimate. The weighted matrix is ​​then used to replace the weighted diagonal matrix. The second estimate is obtained by solving the linear least squares method after the second weighting.

[0041] In one example, the coordinates of four base stations deployed in the environment are set in a clockwise order as follows: , , , And set the location label to Then the nonlinear equation is: , This is expressed as the distance between the location tag and the base station. The distance to be calculated is expressed as... The calculation is as follows: , in, Represented as ,and The speed at which electromagnetic waves propagate in the air. This represents the time difference between the location tag sending the ranging request message M1 and receiving the response message M2. This represents the time difference between the base station receiving message M1 and sending response message M2. This represents the time difference between the tag receiving the response message M2 and sending the acknowledgment message M3. This represents the time difference between the base station sending the response message M2 and receiving the acknowledgment message M3; As one specific implementation, the nonlinear equation system can be vectorized to obtain linear equations. , For unknown quantities, , It is calculated from the measured distance difference and the base station coordinates. For the observation vector, It is a coefficient matrix.

[0042] Based on the UWB ranging values ​​of each base station and tag Compared with reference threshold Relationship, define weight To enhance the reliability of the feedback data and further improve the adaptability of the positioning algorithm to complex communication environments, this application introduces a communication-aware driven dynamic weight function. This function directly maps the link-aware parameters extracted during UWB communication to the weights in the positioning solution, setting: , in, Indicates the first UWB ranging results for each base station Indicates the laser ranging distance. For time jitter perception factor, The energy first-path ratio sensing factor. All are sensing regulation coefficients, when Time weight The range is drastically reduced, and its contribution to the positioning solution is approximately ignored. All range weights are then constructed into a weight diagonal matrix. In the weighted least squares solution of the perception-driven adaptive weighted Chan (PAW-Chan) algorithm, a weight diagonal matrix is ​​introduced. This means using weighted least squares to handle abnormal measurements that exceed the reference threshold.

[0043] As a specific implementation method, the first estimate is obtained using the weighted least squares method: , in, This is the first estimate. The weighted matrix can be obtained by recalculating using the position coordinates obtained from the first estimate, as the transpose of the coefficient matrix: ,and , We can obtain the following using the double-weighted least squares method: , in, , , , Then we obtain the second estimate, where For weighted matrices, For Each component is a diagonal matrix consisting of diagonal elements. This is the second estimate. It is a quadratic coefficient matrix. For the second observation vector, , , , Let be the squares of the first, second, third, and fourth components of the quadratic observation vector, respectively. Then the initial position estimate is... .

[0044] After obtaining the initial position estimate, the residual is calculated according to the following formula: , in, Represents the residual. Indicates the base station index. The UWB ranging value from the i-th base station to the positioning tag and the base station The difference between the UWB ranging values ​​to the positioning tag, ( , , ( ) represents the initial position estimate coordinates. , , ) is the first The coordinates of each base station , , (base station) The coordinates of the base station. This is a reference base station.

[0045] The residual is equal to the sum of the residuals of each base station.

[0046] When the system is unobstructed and the link is stable, the reference measurement value of the laser ranging module is used as the residual squared threshold. Combined with a low-time jitter sensing factor and an energy first-path ratio sensing factor, the UWB ranging data of the corresponding base station is classified as a high-reliability measurement. This type of data will participate in subsequent positioning calculations with a higher weight.

[0047] When the system detects occlusion or multipath interference in the communication sensing parameters, which include the time jitter sensing factor and the energy first path ratio sensing factor, it automatically adopts one or more of the following processing methods based on the specific values ​​of the communication sensing parameters: automatically adjust the residual threshold to adapt to the current link quality; reduce the weight of the corresponding UWB ranging data to reduce its contribution to the positioning solution; mark the corresponding UWB ranging data as abnormal and suppress it so that it is approximately not involved in the positioning solution.

[0048] The UWB ranging data, after being filtered again, and the initial position estimate are then fed into the AMCA-EKF filter for further filtering to obtain the final positioning coordinates.

[0049] The aforementioned adaptive weighted Chan algorithm introduces a weighting mechanism based on a reference threshold for laser ranging. It assigns smaller weights or removes ranging data with abnormal distances, thereby improving the system's stability in occluded environments. Specifically, it uses the existing reference threshold for laser ranging to determine whether the UWB ranging value exceeds the line-of-sight distance. When the UWB ranging value is large, its weight in the adaptive weighted Chan algorithm is dynamically reduced to filter out non-line-of-sight errors and significantly enhance the algorithm's anti-occlusion capability.

[0050] In one embodiment, the initial position estimate and the filtered UWB ranging data are jointly input into the AMCA-EKF filtering algorithm (Adaptive Measurement Covariance Adjusted Extended Kalman Filter), which includes: Establish the state prediction equation: predict the state at the current time based on the estimated state at the previous time step, and obtain the predicted state at time k. Establish measurement equations: Establish the mathematical relationship between the predicted state and the measurement value at the current time, and determine the measurement matrix; Calculate the penalty amount: Compare the UWB ranging data of each base station with the reference threshold, and calculate the penalty amount based on the difference between the two. Dynamic modulation measurement noise covariance: When it is determined that the UWB ranging data exceeds the parameter threshold, the measurement noise covariance is dynamically increased according to the time jitter sensing factor and energy first path ratio sensing factor in the communication sensing parameters; when it does not exceed the parameter threshold, the basic measurement noise covariance remains unchanged. Calculate the Kalman gain: Calculate the Kalman gain based on the measured noise covariance and prior covariance; Based on the calculated Kalman gain and prior covariance, the posterior covariance is calculated, and the posterior covariance is used as the prior covariance for the state prediction at the next time step. Based on the state prediction equation, the predicted state is corrected using Kalman gain, and the estimated state vector at time k is obtained as the final positioning coordinates.

[0051] First, establish the state prediction equation: , Establish the measurement equation: , in, yes The predicted state vector at time t. It is the state transition matrix. It is a control input matrix. It is a measurement matrix. It is measurement noise. yes The predicted state vector at time t. It is the control input vector. for The measured value at a given time.

[0052] For each update step, the first step is to refer to and set thresholds based on the various UWB ranging data. The difference is used to calculate the penalty. To enable the filtering process to adaptively reflect changes in the communication link state, this application introduces a dynamic modulation mechanism based on the measurement noise covariance of communication sensing parameters. hour, For UWB ranging values, it is assumed that the base station ranging is obstructed, which increases the measurement error. Therefore, the system's basic measurement noise covariance is set. , construct with Measurement noise covariance of the modulation coefficients: ,by Replace the original Calculate the Kalman gain: , , in, It is a measurement matrix. for Prior covariance at time 1 for The prior covariance at any given time, then when a certain UWB ranging value... When the value is greater than the reference threshold, the corresponding measurement noise covariance As the Kalman gain is increased, the weight of the UWB ranging value in the update decreases, thereby suppressing the abnormal errors introduced by occlusion. It is Kalman gain. express Measurement noise covariance at time.

[0053] Therefore, the state prediction equation is:

[0054] Among them, Kalman gain Based on prior covariance calculations, this prediction-update loop continues, enabling smooth and accurate state estimation by fusing multiple measurements and the system dynamic model, even when a single measurement is noisy. The final filtered result... This refers to the dynamic estimated location of the positioning tag.

[0055] The measurement noise covariance is dynamically increased based on the deviation of UWB ranging data or UWB ranging values ​​from a reference threshold, enabling the Kalman filter to automatically weaken measurement information affected by obstruction. When communication sensing information indicates a stable link status, a small measurement noise covariance is maintained to enhance the contribution of the corresponding UWB ranging data to filter updates. When communication sensing information indicates link obstruction, multipath propagation, or signal fluctuations, the system automatically increases the measurement noise covariance of the corresponding UWB ranging data, thereby reducing the impact of that UWB ranging data on filter updates.

[0056] The positioning error of the method in this application is compared with the positioning error of other algorithms, as shown in the figure below. Figure 3 The figure shows a comparison of positioning errors for the RSSI method (RSSI is a wireless ranging and positioning method based on received signal strength indication), the method of this application, and the bilateral ranging joint least squares method.

[0057] It can be seen that the method of this application has a significant advantage in positioning accuracy, with the error controlled within 2 meters, while the errors of the two comparative algorithms are about 4 meters and 10 meters, respectively. The positioning error fluctuation of the method of this application is very small, indicating that it has good adaptability to base stations at different distances. This verifies the effectiveness of the technical solution of this application, especially the improvement measures of introducing communication sensing parameters for dynamic weight adjustment and adaptive filtering, which indeed bring about a significant improvement in positioning accuracy.

[0058] Although embodiments of this application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multimodal sensing-based orthopedic surgical navigation and positioning method, characterized in that: include: Simultaneously acquire UWB ranging data and laser ranging data; The time jitter perception factor and energy first path ratio perception factor are extracted from UWB ranging data, and the communication link status is evaluated based on the comparison results of the time jitter perception factor and energy first path ratio perception factor with preset thresholds. Based on the evaluation results, the UWB ranging data are classified into levels, and the participation level of the UWB ranging data is determined according to the level, resulting in the effective UWB ranging data after the first screening. The effective UWB ranging data after the first screening are used to obtain the initial position estimate by the perception-driven adaptive weighted Chan algorithm. The residual is calculated based on the initial position estimate. The residual is the cumulative value of the squared difference between the actual measured distance difference and the theoretical calculated distance difference. Based on the comparison between the residual and the residual threshold, the UWB ranging data is classified into levels. The participation degree of the UWB ranging data is determined according to the level, and the effective UWB ranging data after the second screening is obtained. The initial position estimate and the effective UWB ranging data after the second screening are input into the AMCA-EKF filtering algorithm. The measurement noise covariance is dynamically adjusted according to the time jitter sensing factor and the energy first diameter ratio sensing factor, and finally the high-precision position coordinates of the surgical instrument are output.

2. The method for navigation and positioning in orthopedic surgery based on multimodal sensing with integrated synesthesia as described in claim 1, characterized in that: The extraction of time jitter perception factor includes: within a preset time window, performing multiple consecutive two-way ranging communications with the i-th base station and recording the round-trip propagation delay measured in each communication process; Calculate the average of all round-trip propagation delays within the time window; Calculate the absolute value of the difference between the average propagation delay for each round trip; sum these absolute differences and divide by the number of communications to obtain the time jitter perception factor.

3. The method for navigation and positioning in orthopedic surgery based on multimodal sensing according to claim 2, characterized in that: The extraction of the energy first path ratio sensing factor includes: for the communication signal received by the i-th base station, extracting the received energy of the first path signal and the total received energy of all path signals in the entire communication frame; The ratio of the received energy of the first path signal to the total received energy is the energy first path ratio sensing factor.

4. The method for navigation and positioning in orthopedic surgery based on multimodal sensing with integrated synesthesia as described in claim 2, characterized in that: The classification includes high-confidence data, low-confidence data, and abnormal data. High-confidence data directly participates in the location calculation, low-confidence data participates in the calculation with reduced weight, and abnormal data is suppressed or eliminated.

5. The method for navigation and positioning in orthopedic surgery based on multimodal sensing according to claim 2, characterized in that: The method further includes determining the current environmental state based on the values ​​of the time jitter perception factor and the energy first-path ratio perception factor, and executing differentiated laser information-assisted processing strategies accordingly, including: Reference threshold consistency verification under unobstructed and stable communication link conditions: Using laser ranging data as a reference threshold, the UWB ranging data of each base station is compared with the reference threshold; if the deviation between the UWB ranging data and the reference threshold is within a preset allowable range, the UWB ranging data is determined to pass the consistency verification and is marked as valid UWB ranging data; if the deviation between the UWB ranging data and the reference threshold exceeds the preset allowable range, the verification is determined to fail, and the UWB ranging data is discarded or its weight in positioning is reduced.

6. The method for navigation and positioning in orthopedic surgery based on multimodal sensing according to claim 5, characterized in that: The current environmental state is determined based on the values ​​of the time jitter perception factor and the energy first path ratio perception factor, and a differentiated laser information-assisted processing strategy is executed accordingly. It also includes: when human body obstruction, metal reflection or multipath interference is detected, the deviation between the UWB ranging data of each base station and the reference threshold is calculated, and the UWB ranging data is divided into severely abnormal measurements, low-confidence measurements and valid measurements according to the magnitude of the deviation.

7. The method for navigation and positioning in orthopedic surgery based on multimodal sensing according to claim 1, characterized in that: The filtered valid UWB ranging data are used to obtain initial position estimates using a perception-driven adaptive weighted Chan algorithm, including: A system of nonlinear equations is established based on the distance from each base station to the positioning tag; After vectorizing the nonlinear equations, a linear equation is obtained. The unknowns in the linear equation include the three-dimensional coordinates of the positioning tag and distance-related terms. The coefficient matrix of the linear equation is calculated from the measured distance difference and the base station coordinates. The weights are obtained based on the relationship between the UWB ranging data of each base station and the positioning tag and the reference threshold; All weights are constructed into a weight diagonal matrix. This weight diagonal matrix is ​​introduced into the weighted least squares solution of the perception-driven adaptive weighted Chan algorithm. The first estimate is obtained by using the weighted least squares solution. The weighted matrix is ​​recalculated based on the first estimate. The weighted matrix is ​​then used to replace the weighted diagonal matrix. The second estimate is obtained by solving the linear least squares method after the second weighting.

8. The orthopedic surgical navigation and positioning method based on synesthesia integration according to claim 1, characterized in that: The weights are: ， in, Indicates the first UWB ranging results for each base station Indicates the laser ranging distance. For time jitter perception factor, The energy first-path ratio sensing factor. All are sensory regulation coefficients. This indicates the reference threshold.

9. The multimodal sensing orthopedic surgical navigation and positioning method based on synesthesia integration according to claim 1, characterized in that: The initial position estimate and the effective UWB ranging data after the second filtering are input into the AMCA-EKF filtering algorithm, including: Establish the state prediction equation: predict the state at the current time based on the estimated state at the previous time step, and obtain the predicted state at time k. Establish measurement equations: Establish the mathematical relationship between the predicted state and the measurement value at the current time, and determine the measurement matrix; Calculate the penalty amount: Compare the UWB ranging data of each base station with the reference threshold, and calculate the penalty amount based on the difference between the two. Dynamic modulation measurement noise covariance: When it is determined that the UWB ranging data exceeds the parameter threshold, the measurement noise covariance is dynamically increased according to the time jitter sensing factor and energy first path ratio sensing factor in the communication sensing parameters; when it does not exceed the parameter threshold, the basic measurement noise covariance remains unchanged. Calculate the Kalman gain: Calculate the Kalman gain based on the measured noise covariance and prior covariance; Based on the calculated Kalman gain and prior covariance, the posterior covariance is calculated, and the posterior covariance is used as the prior covariance for the state prediction at the next time step. Based on the state prediction equation, the predicted state is corrected using Kalman gain, and the estimated state vector at time k is obtained as the final positioning coordinates.

Images