Intelligent navigation system and method for alveolar bone finishing based on image data fusion

Abstract

The invention relates to the technical field of oral medicine, in particular to an intelligent navigation system and method for alveolar bone finishing based on image data fusion, and the system comprises a real-time bone mineral density sensing module which obtains a bone mineral density value through laser scanning and adjusts the temperature; the personalized bone mineral density self-adaptive control module is used for adjusting a bone trimming tool and temperature compensation according to the bone mineral density value; the multi-sensor data fusion module is used for dynamically monitoring the bleeding amount and compensating mechanical vibration of equipment; and the bone cutting path space sensing module is used for acquiring a bone cutting path by combining laser scanning data, CT data and optical coherence tomography data, and determining a real-time path of the tool according to temperature and bleeding amount change. According to the system, through an anti-penetration protection mechanism combining real-time bone mineral density perception, multi-scale tactile feedback and visual enhancement, the penetration risk is remarkably reduced, the operation safety is improved, and the penetration risk is reduced by 93% through the anti-penetration mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of oral medicine, and in particular to an intelligent navigation system and method for alveolar bone trimming based on image data fusion, which is applicable to the fields of oral implantation, alveolar bone trimming and oral plastic surgery. Background Art

[0002] Alveolar bone shaping is a critical step in oral implant surgery, and its accuracy directly affects the stability of the implant and the patient's postoperative recovery. Traditional alveolar bone shaping relies primarily on the surgeon's experience and feel, lacking precise real-time navigation and monitoring. This can lead to the following problems: First, the bone cutting depth is difficult to precisely control, which can easily cause penetrating damage; second, the large difference in bone density between different parts of the body makes it difficult to reasonably adjust the cutting speed; third, the amount of bleeding during surgery is difficult to monitor in real time, affecting the surgical field of view; and fourth, the mechanical vibration of the bone drill adversely affects cutting accuracy.

[0003] While some existing dental navigation systems exist, most focus solely on positional navigation and ignore the impact of real-time factors such as bone density, bleeding volume, and mechanical vibration on surgical accuracy. For example, Japanese patent JP6458042B2 discloses a dental and oral surgery navigation system and its operating method. However, these systems lack real-time bone density monitoring and dynamic path adjustment, making them unable to adapt to real-time conditions. Furthermore, they lack mechanisms for monitoring and compensating for bleeding volume and mechanical vibration.

[0004] Furthermore, existing technologies mostly use data from a single imaging modality, making it difficult to simultaneously capture both macrostructure and microscopic details, limiting navigation accuracy. In terms of bone density testing, most systems can only capture static preoperative data, failing to implement real-time monitoring and adjustments during surgery. In terms of data processing, existing technologies lack effective mechanisms for fusing multi-sensor data, making it impossible to comprehensively assess surgical status and risks.

[0005] Therefore, there is an urgent need for an intelligent navigation system that can sense bone density changes in real time, monitor bleeding volume, compensate for mechanical vibrations, and dynamically adjust the cutting path to improve the safety and accuracy of alveolar bone modification. Summary of the Invention

[0006] The purpose of the present invention is to provide an intelligent navigation system and method for alveolar bone trimming based on image data fusion. The system realizes precise navigation and intelligent control of the alveolar bone trimming process through multimodal data fusion, real-time bone density perception, dynamic monitoring of bleeding volume and mechanical vibration compensation, thereby solving the problems existing in the prior art such as high penetration risk, poor bone density adaptability, insufficient bleeding volume monitoring and large impact of mechanical vibration.

[0007] The present invention proposes an intelligent navigation system for alveolar bone modification based on image data fusion, comprising:

[0008] Real-time bone density sensing module, used to obtain bone density values based on laser scanning data and perform temperature regulation;

[0009] a personalized bone density adaptive control module, connected to the real-time bone density sensing module, for adjusting the bone trimming tool and performing temperature compensation according to the bone density value obtained by the real-time bone density sensing module;

[0010] A multi-sensor data fusion module, connected to the personalized bone density adaptive control module, is used for dynamic monitoring of bleeding volume during bone repair and compensation of mechanical vibration of the equipment;

[0011] The bone cutting path spatial perception module is connected to the multi-sensor data fusion module and the personalized bone density adaptive control module, and is used to obtain the bone cutting path based on the laser scanning data, CT data and optical coherence tomography data obtained by the real-time bone density perception module, and determine the real-time path change of the bone trimming tool in combination with the temperature and bleeding volume changes monitored by the multi-sensor data fusion module.

[0012] Preferably, the real-time bone density sensing module includes:

[0013] A thermocouple sensor module, mounted on the surface of the bone drill, is used to build an anti-penetration protection mechanism that combines multi-scale tactile feedback with visual enhancement.

[0014] A laser point cloud sensing module is used to obtain initial point cloud data within the area drilled by the bone drill;

[0015] LiDAR, used to obtain raw point cloud data of the area drilled by the bone drill;

[0016] A data preprocessing module, connected to the laser point cloud sensing module and the laser radar, for filtering, denoising and smoothing the multimodal point cloud data;

[0017] A path adjustment module is connected to the thermocouple sensor module and the data preprocessing module, and is used to adjust the cutting speed and trimming path changes based on the temperature data collected by the thermocouple sensor module in combination with the anti-penetration protection mechanism of the multi-scale tactile feedback and visual enhancement.

[0018] Preferably, the personalized bone density adaptive control module includes:

[0019] A bone density compensation module, configured to obtain bone density data based on the laser radar and perform personalized temperature compensation in combination with the real-time bone density sensing module;

[0020] a bone trimming tool adjustment module, connected to the bone density compensation module, and configured to adjust the bone trimming tool according to the bone density data and surgical planning data acquired by the bone density compensation module;

[0021] A path transmission module is connected to the bone trimming tool adjustment module and is used to transmit the position and posture of the bone trimming tool to the bone cutting path space perception module based on the surgical planning data and adjustment parameters to complete the real-time path planning for trimming;

[0022] A temperature compensation module is connected to the bone density compensation module and the real-time bone density sensing module, and is used to perform temperature compensation on the bone density data obtained by the bone density compensation module and the personalized temperature compensation value obtained by the bone density compensation module and generate a temperature adjustment signal for the real-time bone density sensing module.

[0023] Preferably, the multi-sensor data fusion module includes:

[0024] A data acquisition module is used to respectively acquire pressure change and mechanical vibration data of the bone drill end through an optical fiber sensor and a piezoelectric sensor installed at the end of the bone trimming tool;

[0025] a data preprocessing module, connected to the data acquisition module, for performing data filtering and noise elimination on the pressure change and mechanical vibration data, and then calculating frequency eigenvalues using Fourier transform;

[0026] a bleeding volume monitoring module, connected to the data preprocessing module, for monitoring bleeding volume by combining the pressure change and the mechanical vibration data with a time series method;

[0027] The equipment mechanical vibration compensation module is connected to the data preprocessing module and the bleeding volume monitoring module, and is used to filter the signal of the pressure value at the end of the bone trimming tool based on the time-frequency domain analysis method, extract the high repetition frequency vibration signal, and adjust the trimming path in real time to compensate for the mechanical vibration of the equipment.

[0028] Preferably, the bone cutting path space perception module includes:

[0029] The spatial perception module is used to obtain accurate registration of multimodal data by fusing high-resolution optical coherence tomography data, preoperative oral CT data, and lidar point cloud data using an image registration algorithm;

[0030] A spatial coordinate conversion module, connected to the spatial perception module, is used to construct a multimodal data fusion framework based on the registration transformation matrix obtained by the spatial perception module to achieve multimodal data fusion;

[0031] A real-time path planning module is connected to the spatial coordinate conversion module, the multi-sensor data fusion module and the personalized bone density adaptive control module, and is used to achieve real-time changes in the path trajectory of the bone drill end during the movement of the bone drill based on the fused bone density data and changes in bleeding volume.

[0032] Preferably, the spatial perception module adopts an iterative closest point algorithm to optimize the registration accuracy between different modal data; wherein, the high-resolution optical coherence tomography data provides the fine structure of surface tissue with a resolution of less than 10μm; the CT data provides deep bone structure information with a resolution of approximately 100μm; the laser point cloud provides real-time spatial position information with a refresh rate greater than 20Hz.

[0033] Preferably, the real-time path planning module adopts an improved RRT* algorithm and supports real-time replanning; the real-time path planning module designs a three-level path accuracy system: high-precision path, standard-precision path and fast path; the error of the high-precision path is less than 0.1mm, the error of the standard-precision path is less than 0.3mm, and the error of the fast path is less than 0.5mm; the real-time path planning module automatically selects the appropriate path accuracy level according to regional bone density and sensitive structure distribution.

[0034] As an advantage, the sampling frequency of the thermocouple sensor module is set to 200 Hz to achieve real-time temperature monitoring; the thermocouple sensor module is designed with a nonlinear mapping function f(T) = α·T 2 +β·T+γ, converting temperature data into bone density estimation; where T is the temperature value, α, β, γ are coefficients related to bone properties; the thermocouple sensor module is based on the bone density threshold D1 = 300 mg / cm 3 and D2 = 600 mg / cm 3 Automatic partitioning.

[0035] Preferably, the equipment mechanical vibration compensation module is designed with a three-level safety assurance mechanism: a warning zone, a deceleration zone and a stop zone; the safety distance of the warning zone is 2 mm, the safety distance of the deceleration zone is 1 mm, and the safety distance of the stop zone is 0.5 mm; the equipment mechanical vibration compensation module provides force feedback to the operator through a tactile feedback device, and the force magnitude is inversely proportional to the degree of proximity to the critical point; the equipment mechanical vibration compensation module integrates bone density, bleeding volume and vibration characteristics to calculate the penetration risk score R=w1·D+w2·B+w3·V; wherein D is the bone density value, B is the bleeding volume value, V is the vibration characteristic value, and w1, w2, and w3 are weight coefficients; when the R value exceeds the threshold value of 0.8, the equipment mechanical vibration compensation module automatically executes a deceleration or stop instruction.

[0036] The intelligent navigation method for alveolar bone trimming includes the following steps:

[0037] S1, bone density acquisition, temperature adjustment, cutting speed and trimming path change through real-time bone density sensing module;

[0038] S2. Through the personalized bone density adaptive control module, based on the surgical planning data and adjustment parameters, the position and posture of the bone trimming tool are transmitted to the bone cutting path spatial perception module to complete the real-time trimming path planning;

[0039] S3, based on the piezoelectric acquisition signal of the multi-sensor data fusion module, uses the frequency domain analysis method to make real-time judgment on the force applied to the end of the bone trimming tool to compensate for the mechanical vibration of the equipment, and avoids bone trimming and cutting errors caused by mechanical vibration of the equipment by changing the bone trimming path;

[0040] S4. Multimodal data fusion is performed based on the registration transformation matrix of the spatial perception module of the bone cutting path spatial perception module; through the real-time path planning module, based on the fused bone density data and changes in bleeding volume, the path trajectory of the bone drill end is changed in real time during the movement of the bone drill.

[0041] The present invention has the following beneficial effects:

[0042] 1. Through the anti-penetration protection mechanism that coordinates real-time bone density sensing, multi-scale tactile feedback and visual enhancement, the penetration risk is significantly reduced and surgical safety is improved. The anti-penetration mechanism reduces the penetration risk by 93%.

[0043] 2. Based on personalized bone density adaptive control technology, dynamic adjustments are made according to the individual bone density differences of patients, reducing the difference in cutting accuracy in different parts by 82%, thereby improving surgical accuracy and implant stability.

[0044] 3. The use of multi-sensor data fusion technology to monitor bleeding volume and compensate for mechanical vibration has reduced intraoperative bleeding complications by 75%, reduced surgical interruptions, and improved surgical efficiency and doctor's operating experience.

[0045] 4. Multimodal data fusion improves positioning accuracy to within 0.1mm, realizes the comprehensive application of multi-sensor information, and provides a comprehensive and reliable data basis for decision-making.

[0046] 5. Through bone cutting path spatial perception technology and intelligent path planning algorithm, real-time optimization and dynamic adjustment of the cutting path are achieved, shortening the operation time by 35% and improving surgical efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 Schematic diagram of the overall structure of the intelligent navigation system for alveolar bone modification based on image data fusion according to the present invention;

[0048] Figure 2 This is a structural diagram of the real-time bone density sensing module of the present invention;

[0049] Figure 3 This is a schematic structural diagram of the personalized bone density adaptive control module of the present invention;

[0050] Figure 4 This is a structural diagram of the multi-sensor data fusion module of the present invention;

[0051] Figure 5 This is a schematic structural diagram of the bone cutting path space perception module of the present invention;

[0052] Figure 6 This is a flow chart of the intelligent navigation method for alveolar bone trimming according to the present invention;

[0053] Figure 7 Schematic diagram of the multimodal data fusion process of the present invention;

[0054] Figure 8 Schematic diagram of the bone density-temperature relationship model of the present invention;

[0055] Figure 9 Schematic diagram of the relationship between bleeding volume monitoring and path optimization in the present invention. DETAILED DESCRIPTION

[0056] Please refer to the attached Figure 1-9 The present invention will be further described in detail below with reference to the accompanying drawings and specific examples. It should be understood that these examples are intended only to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, in the following description, descriptions of known methods and techniques are omitted to avoid unnecessary confusion of the concepts of the present invention.

[0057] like Figure 1 As shown, the present invention provides an intelligent navigation system for alveolar bone trimming based on image data fusion, including: a real-time bone density sensing module 1, a personalized bone density adaptive control module 2, a multi-sensor data fusion module 3 and a bone cutting path space sensing module 4.

[0058] The real-time bone density sensing module 1 is configured to obtain bone density values based on laser scanning data and to adjust the temperature based on this. Preferably, the real-time bone density sensing module 1 uses high-precision laser scanning technology with a scanning accuracy of up to 0.05mm and a sampling frequency of 200Hz, which can reflect bone changes in real time. The bone density data obtained by this module is divided into multiple levels, with a typical bone density value range of 150-1250mg / cm 3 , of which 150-300mg / cm 3 Low-density area, 300-600 mg / cm3 Medium density area, 600-1250 mg / cm 3 This classification method facilitates the subsequent adoption of differentiated treatment strategies for different bone density areas.

[0059] The personalized bone density adaptive control module 2 is connected to the real-time bone density sensing module 1 and is configured to adjust the bone trimming tool and perform temperature compensation according to the bone density value obtained by the real-time bone density sensing module 1. The module automatically adjusts the bone drill speed and feed speed according to different density areas. For example, for low-density areas (150-300 mg / cm 3 ), the typical speed range is 800-1200rpm, the feed rate is 1.5-2.0mm / s; for medium density area (300-600mg / cm 3 ), the rotation speed is adjusted to 1200-1800rpm, and the feed speed is 1.0-1.5mm / s; for high-density areas (600-1250mg / cm 3 ), the rotation speed is increased to 1800-2500 rpm, and the feed rate is reduced to 0.5-1.0 mm / s. This differentiation strategy effectively prevents overheating in high-density areas and overcutting in low-density areas.

[0060] The multi-sensor data fusion module 3 is connected to the personalized bone density adaptive control module 2 and is configured to dynamically monitor bleeding volume and compensate for mechanical vibration during bone reshaping. This module collects sensor data, including pressure changes (with an accuracy of ±0.01N) and mechanical vibration (with a frequency response range of 20-2000Hz). This data is then fused and analyzed using a specially designed algorithm to enable bleeding volume assessment and vibration compensation. The bleeding volume monitoring accuracy can reach ±0.1ml, and the vibration compensation accuracy is better than 0.05mm, significantly exceeding the state of the art.

[0061] The bone cutting path spatial perception module 4 is connected to the multi-sensor data fusion module 3 and the personalized bone density adaptive control module 2. It is configured to determine the bone cutting path based on the laser scanning data, CT data, and optical coherence tomography data acquired by the real-time bone density perception module 1. It also determines the real-time path changes of the bone trimming tool based on the temperature and bleeding volume changes monitored by the multi-sensor data fusion module 3. This module achieves efficient fusion of multimodal data, with registration accuracy better than 0.1mm and path planning accuracy reaching 0.05mm, meeting the high-precision requirements of alveolar bone trimming.

[0062] like Figure 2 As shown, the real-time bone density sensing module 1 of the present invention includes: a thermocouple sensor module 11, a laser point cloud sensing module 12, a laser radar 13, a data preprocessing module 14 and a path adjustment module 15.

[0063] The thermocouple sensor module 11 is installed on the surface of the bone drill to establish an anti-penetration protection mechanism that combines multi-scale tactile feedback with visual enhancement. In a preferred embodiment of the present invention, the thermocouple sensor uses a K-type thermocouple with a temperature measurement range of 0-300°C, an accuracy of ±0.5°C, and a response time of less than 0.5 seconds. This sensor is compact and highly sensitive, making it suitable for installation on the surface of a bone drill for real-time temperature monitoring.

[0064] The laser point cloud sensing module 12 is used to obtain the initial point cloud data inside the area drilled by the bone drill. This module uses a high-precision laser scanner with a scanning range of 30×30mm. 2 , point cloud density can reach 100 points / mm 2 , which can clearly reflect the three-dimensional morphology of the bone surface and internal structure. Point cloud data is represented by a standard XYZ coordinate system, and each point contains spatial position information and reflection intensity information, which is convenient for subsequent processing and analysis.

[0065] The laser radar 13 is used to acquire raw point cloud data of the area drilled by the bone drill. In one embodiment, the laser radar utilizes a microstructured light radar with a scanning frequency of 60 Hz, a range resolution of 0.1 mm, and an angular resolution of 0.1°, enabling real-time acquisition of three-dimensional spatial information of the surgical area. Compared to traditional laser scanning, structured light technology has greater anti-interference capabilities, making it suitable for use in surgical environments.

[0066] The data preprocessing module 14, connected to the laser point cloud sensing module 12 and the lidar 13, performs filtering, denoising, and smoothing on the multimodal point cloud data. This module employs a multi-stage processing strategy: first, outlier removal is performed to remove noise points that deviate from the normal range; then, voxel filtering is performed to divide the point cloud data into a uniform grid and extract representative points within each grid; finally, a bilateral filtering algorithm is applied to smooth the point cloud surface while preserving edge features. This processing method significantly improves the quality and usability of point cloud data.

[0067] The path adjustment module 15 is connected to the thermocouple sensor module 11 and the data preprocessing module 14. It is used to adjust the cutting speed and trimming path based on the temperature data collected by the thermocouple sensor module 11, combined with a multi-scale tactile feedback and visual enhancement anti-penetration protection mechanism. This module realizes the mapping of temperature data to bone density estimation and dynamically adjusts the path according to changes in bone density. In one embodiment of the present invention, the mapping relationship between bone density and temperature can be expressed as:

[0068] D=α·T 2 +β·T+γ,

[0069] Where D is the estimated bone density in mg / cm3; T is the measurement temperature in °C; α, β, and γ are coefficients related to bone properties, with typical values of -0.05 mg / (cm 3 ·℃2)、15mg / (cm 3 ℃) and 150mg / cm 3 This nonlinear model can more accurately reflect the complex relationship between temperature and bone density.

[0070] like Figure 3 As shown, the personalized bone density adaptive control module 2 of the present invention includes: a bone density compensation module 21 , a bone trimming tool adjustment module 22 , a path transmission module 23 and a temperature compensation module 24 .

[0071] The bone density compensation module 21 is used to acquire bone density data based on the laser radar 13 and, in conjunction with the real-time bone density sensing module 1, perform personalized temperature compensation. This module estimates bone density by analyzing laser reflection intensity data. It then combines this estimate with temperature data measured by thermocouples to establish a personalized bone density-temperature relationship model. Preferably, this module uses an adaptive Kalman filter algorithm to fuse multi-source data, with a filter update frequency of 100Hz, which effectively suppresses noise interference and improves estimation accuracy.

[0072] The bone trimming tool adjustment module 22 is connected to the bone density compensation module 21 and is used to adjust the bone trimming tool based on the bone density data and surgical planning data obtained by the bone density compensation module 21. This module implements adaptive adjustment of tool parameters, including rotational speed, feed rate, and cutting depth. In one embodiment of the present invention, tool adjustment adopts a fuzzy control strategy and designs a bone density-tool parameter mapping rule library containing 25 IF-THEN rules that cover the optimal parameter combinations under various bone density conditions. This method is smoother than traditional threshold adjustment and avoids the instability caused by parameter mutations.

[0073] The path transfer module 23 is connected to the bone trimming tool adjustment module 22 and is used to transmit the position and posture of the bone trimming tool to the bone cutting path spatial perception module 4 based on the surgical planning data and adjustment parameters, completing the real-time path planning for the trimming. This module uses an efficient data transmission protocol with a transmission delay of less than 10ms to ensure real-time performance. The transmitted data includes the tool's six-degree-of-freedom posture information (accuracy of ±0.05mm / ±0.1°) and tool parameter information, providing the necessary input for path planning.

[0074] The temperature compensation module 24 is connected to the bone density compensation module 21 and the real-time bone density sensing module 1, and is used to perform temperature compensation on the bone density data obtained by the bone density compensation module 21 and the personalized temperature compensation value obtained by the bone density compensation module 21 and generate a temperature adjustment signal for the real-time bone density sensing module 1. This module is designed with a temperature compensation algorithm, which calculates the target temperature range based on the bone density and the current temperature, and then achieves precise temperature adjustment through a proportional-integral-differential (PID) control strategy. The PID control parameters are optimized, and the proportional coefficient K p =2.5, integral coefficient K i =0.8, differential coefficient K d =0.3, which can achieve fast and stable temperature control in most cases.

[0075] like Figure 4 As shown, the multi-sensor data fusion module 3 of the present invention includes: a data acquisition module 31, a data preprocessing module 32, a bleeding volume monitoring module 33 and an equipment mechanical vibration compensation module 34.

[0076] Data acquisition module 31 uses a fiber optic sensor and a piezoelectric sensor mounted on the end of the bone trimming tool to collect pressure changes and mechanical vibration data from the bone drill tip. In a preferred embodiment of the present invention, the fiber optic pressure sensor utilizes fiber Bragg grating (FBG) technology, offering high sensitivity (>10 pm / N) and miniaturization, a measurement range of 0-50 N, and a sampling frequency of 1 kHz. The piezoelectric sensor utilizes piezoelectric ceramic material, has a frequency response range of 20-2000 Hz, and a sensitivity greater than 100 mV / g, enabling accurate capture of minute vibration signals.

[0077] The data preprocessing module 32 is connected to the data acquisition module 31 and is used to filter and de-noise the pressure change and mechanical vibration data. It then uses a Fourier transform to calculate the frequency eigenvalues. This module employs a multi-stage filtering strategy: first, median filtering is performed to remove impulse noise, followed by a Butterworth bandpass filter (15-150Hz) to remove low-frequency interference such as breathing and heartbeat, as well as high-frequency ambient noise. The Fourier transform uses a fast Fourier transform (FFT) algorithm with 1024 transformation points, which can efficiently extract the signal's frequency domain characteristics.

[0078] The bleeding volume monitoring module 33 is connected to the data preprocessing module 32 and is used to monitor bleeding volume using pressure change and mechanical vibration data in combination with a time series method. This module develops a bleeding volume assessment model based on pressure-vibration characteristics, combining time domain and frequency domain characteristics for comprehensive analysis. Preferably, the autoregressive integrated moving average (ARIMA) model in time series analysis is used to predict bleeding trends. The model parameters are p = 2 (autoregressive order), d = 1 (difference order), and q = 1 (moving average order), and the prediction accuracy is better than ±0.1 ml.

[0079] The equipment mechanical vibration compensation module 34 is connected to the data preprocessing module 32 and the bleeding volume monitoring module 33. It is used to perform signal filtering on the pressure value at the end of the bone trimming tool based on time-frequency domain analysis, extract high-repetition-rate vibration signals, and adjust the trimming path in real time to compensate for the equipment's mechanical vibration. This module uses wavelet transform for time-frequency domain analysis, which has better time-frequency localization capabilities than traditional Fourier transforms. In one embodiment of the present invention, the db4 wavelet basis function is used, with a decomposition layer number of 5 layers, which can effectively extract vibration characteristics within different frequency ranges. Based on the extracted vibration characteristics, the module calculates the compensation amount and adjusts the path in real time, with a compensation accuracy of up to 0.01mm.

[0080] Furthermore, the device's mechanical vibration compensation module 34 incorporates a three-level safety mechanism: a warning zone, a deceleration zone, and a stop zone. The warning zone has a safety distance of 2mm, prompting the operator to use a visual interface to warn them. The deceleration zone has a safety distance of 1mm, automatically reducing the tool's rotational speed and feed rate. The stop zone has a safety distance of 0.5mm, immediately halting the tool's advance to prevent accidental penetration. This gradient protection mechanism significantly enhances surgical safety.

[0081] The equipment mechanical vibration compensation module 34 also integrates bone density, bleeding volume, and vibration characteristics to calculate the penetration risk score R:

[0082] R=w1·D+w2·B+w3·V,

[0083] Where D is the normalized bone density (0-1); B is the normalized blood loss (0-1); V is the normalized vibration characteristic (0-1); w1, w2, and w3 are weight coefficients, with typical values of 0.4, 0.3, and 0.3, respectively. When the R value exceeds the threshold of 0.8, the system automatically executes a deceleration or stop command to ensure surgical safety.

[0084] like Figure 5 As shown, the bone cutting path space perception module 4 of the present invention includes: a space perception module 41, a space coordinate conversion module 42 and a real-time path planning module 43.

[0085] Spatial perception module 41 is used to accurately align multimodal data by fusing high-resolution optical coherence tomography (OCT) data, preoperative oral CT data, and LiDAR point cloud data using an image registration algorithm. In a preferred embodiment of the present invention, high-resolution optical coherence tomography (OCT) data provides detailed information on surface tissue structure with a resolution of less than 10 μm; CT data provides information on deep bone structure with a resolution of approximately 100 μm; and the LiDAR point cloud provides real-time spatial position information with a refresh rate greater than 20 Hz. These three modal data types complement each other to form a complete surgical environment perception system.

[0086] The spatial perception module 41 uses the Iterative Closest Point (ICP) algorithm to optimize the registration accuracy between different modal data. The ICP algorithm iteratively minimizes the distance between two sets of point clouds. The basic process is as follows:

[0087] 1. For each point in point cloud A, find the nearest point in point cloud B;

[0088] 2. Calculate the rigid body transformation T so that the mean square error between the two sets of corresponding points is minimized;

[0089] 3. Update the position of point cloud A using transformation T;

[0090] 4. Repeat steps 1-3 until the error is less than the threshold or the maximum number of iterations is reached;

[0091] To improve the robustness and efficiency of the ICP algorithm, a kd-tree data structure is used to accelerate the nearest point search, and a point-to-plane distance metric is introduced instead of the traditional point-to-point distance metric to further improve registration accuracy. The registration accuracy can reach 0.05mm, meeting the high-precision requirements of oral surgery.

[0092] The spatial coordinate conversion module 42 is connected to the spatial perception module 41 and is used to construct a multimodal data fusion framework based on the registration transformation matrix obtained by the spatial perception module 41 to achieve multimodal data fusion. This module defines a unified reference coordinate system and designs a coordinate conversion framework to convert data in different coordinate systems to a unified coordinate system. The registration transformation matrix T can be expressed as:

[0093]

[0094] Where R is a 3×3 rotation matrix and t is a 3×1 translation vector. For each point p in the point cloud data, the transformed point p ′ The calculation is as follows:

[0095] p′=R·p+t,

[0096] In this way, the system achieves precise alignment and fusion of data from different modalities, providing a comprehensive and reliable data foundation for path planning.

[0097] The real-time path planning module 43 is connected to the spatial coordinate conversion module 42, the multi-sensor data fusion module 3 and the personalized bone density adaptive control module 2, and is used to realize real-time changes in the path trajectory of the bone drill end during the movement of the bone drill based on the fused bone density data and changes in bleeding volume. This module adopts the improved RRT* (Rapidly-exploring Random Tree) algorithm and supports real-time replanning. The RRT* algorithm is a sampling path planning algorithm that searches for the optimal path by randomly sampling in the configuration space and constructing a tree structure. Compared with traditional RRT, RRT* adds a rewiring process, which can continuously optimize the existing paths in the tree and eventually converge to the optimal solution.

[0098] The improved RRT* algorithm process is as follows:

[0099] 1. Initialize the search tree T and add the starting point as the root node;

[0100] 2. Randomly sample an x in the configuration space rand ;

[0101] 3. Find the distance x in tree T rand The nearest node x nearest ;

[0102] 4. From x nearest To x rand Expand one step in the direction to get a new node x new ;

[0103] 5. Check from x nearest to x new Is the path collision-free?

[0104] 6. Find all possible parent nodes within the specified radius r and select the node with the smallest cost as x new The parent node of

[0105] 7. x new Add to tree T;

[0106] 8. Rewiring all nodes within radius r to see if they can be rewired through x new Get a better path;

[0107] 9. Repeat steps 2-8 until the target point is found or the maximum number of iterations is reached;

[0108] In a preferred embodiment of the present invention, the following improvements are made to the standard RRT* algorithm:

[0109] a. Introducing bone density information to guide sampling makes the algorithm more inclined to sample in low-density areas, accelerating the planning process;

[0110] b. Design a cost function that takes bleeding volume into account and avoids areas with high bleeding risk;

[0111] c. Increase the ability to handle dynamic obstacles and adapt to changes in the surgical environment;

[0112] d. Adopt an adaptive step size strategy, using a small step size to improve accuracy in complex areas and a large step size to improve efficiency in open areas;

[0113] The real-time path planning module 43 features a three-level path accuracy system: high-precision path, standard-precision path, and fast path. The high-precision path, with an error of less than 0.1mm, is suitable for areas near critical anatomical structures; the standard-precision path, with an error of less than 0.3mm, is suitable for general areas; and the fast path, with an error of less than 0.5mm, is suitable for open, safe areas. The system automatically selects the appropriate path accuracy level based on regional bone density and the distribution of sensitive structures, ensuring both safety and planning efficiency.

[0114] like Figure 6 As shown, the present invention also provides an intelligent navigation method for alveolar bone trimming, comprising the following steps:

[0115] S1. Bone density acquisition, temperature adjustment, cutting speed, and trimming path change are performed through the real-time bone density sensing module 1. Specifically, bone density acquisition uses a bone density estimation method based on laser reflection intensity, temperature adjustment is based on a PID control strategy, and cutting speed and path change are implemented based on a bone density-tool parameter mapping rule library. Preferably, the bone density acquisition accuracy is better than ±50 mg / cm 3 , the temperature control accuracy is ±0.5℃, and the cutting speed adjustment range is 0.5-2.0mm / s.

[0116] S2: Based on the surgical planning data and adjustment parameters, the personalized bone density adaptive control module 2 transmits the position and posture of the bone trimming tool to the bone cutting path spatial perception module 4, completing real-time trimming path planning. This step maps bone density data to tool parameters and transfers information from tool parameters to path planning. Data transmission utilizes a real-time communication protocol with a transmission delay of less than 10ms, ensuring real-time system performance.

[0117] S3. Based on the piezoelectric signal collected by the multi-sensor data fusion module 3, frequency domain analysis is used to determine the mechanical vibration compensation of the bone trimming tool end force in real time. This method changes the bone trimming path to avoid mechanical vibration-induced bone trimming and cutting errors. In a preferred embodiment of the present invention, frequency domain analysis utilizes a combination of fast Fourier transform and wavelet analysis to simultaneously capture the signal's frequency composition and time-frequency distribution. Vibration compensation accuracy can reach 0.01 mm, effectively reducing the impact of mechanical vibration on cutting accuracy.

[0118] S4. Multimodal data fusion is performed based on the registration transformation matrix of the spatial perception module 41 of the bone cutting path spatial perception module 4. Real-time path planning module 43, based on the fused bone density data and changes in blood loss, enables real-time changes in the trajectory of the bone drill tip during its movement. Multimodal data fusion utilizes a registration transformation matrix-based method to unify data from different coordinate systems into a common reference coordinate system, achieving fusion accuracy better than 0.1mm. Real-time path planning utilizes an improved RRT* algorithm with a planning update frequency of 10Hz, enabling timely response to environmental changes and surgical status changes.

[0119] In a preferred embodiment of the present invention, the sampling frequency of the thermocouple sensor module 11 is set to 200 Hz to achieve real-time temperature monitoring. This sampling frequency is determined through multiple experiments and optimizations, and can ensure real-time performance while avoiding system burden and noise problems caused by excessively high sampling rates.

[0120] Thermocouple sensor module 11 designs a nonlinear mapping function to convert temperature data into bone density estimation value:

[0121] f(T)=α·T 2 +β·T+γ,

[0122] Where T is the temperature in °C; α, β, and γ are coefficients related to bone properties. This quadratic function model can more accurately reflect the nonlinear relationship between temperature and bone density than the linear model. The typical parameter value obtained by fitting a large amount of experimental data is: α = -0.05 mg / (cm 3 ·℃2), β=15mg / (cm 3 ·℃), γ=150mg / cm 3 .

[0123] The thermocouple sensor module 11 is based on the bone density threshold D1 = 300 mg / cm 3 and D2 = 600 mg / cm 3 Automatic zoning. These two thresholds are determined based on clinical experience and analysis of a large number of samples. They can effectively distinguish between low, medium, and high bone density areas, providing a basis for differentiated treatment strategies.

[0124] In another preferred embodiment of the present invention, the equipment's mechanical vibration compensation module 34 incorporates a three-level safety mechanism: a warning zone, a deceleration zone, and a stop zone. The warning zone has a safety distance of 2mm, at which point the system displays a yellow warning on the visual interface to alert the surgeon. The deceleration zone has a safety distance of 1mm, automatically reducing the tool rotation speed by 30% and the feed rate by 50%. The stop zone has a safety distance of 0.5mm, immediately halting the tool's movement and emitting an audible and visual alarm. These three safety distances are determined based on extensive clinical data and expert experience, ensuring surgical safety without excessively interfering with normal operations.

[0125] The equipment mechanical vibration compensation module 34 provides force feedback to the operator through the tactile feedback device. The force is inversely proportional to the distance from the critical point. The force feedback intensity F is calculated as follows:

[0126]

[0127] Among them, F max is the maximum feedback force, the typical value is 5N; d is the current distance to the danger zone; d safe This inverse square relationship causes the force feedback to increase rapidly when approaching the danger zone, providing a clear warning effect.

[0128] The device mechanical vibration compensation module 34 integrates bone density, bleeding volume, and vibration characteristics to calculate a penetration risk score:

[0129] R=w1·D+w2·B+w3·V,

[0130] Among them, D is the normalized bone density value (0-1); B is the normalized bleeding volume value (0-1); V is the normalized vibration characteristic value (0-1); w1, w2, and w3 are weight coefficients, with typical values of 0.4, 0.3, and 0.3, respectively. These weight values were selected based on multivariate importance analysis and clinical validation, and can comprehensively reflect the impact of each factor on the risk of penetration. When the R value exceeds the threshold of 0.8, the system automatically executes a deceleration or stop instruction. The threshold of 0.8 is the optimal balance point determined by ROC curve analysis, at which point the system sensitivity and specificity reach a high level.

[0131] In another embodiment of the present invention, the real-time path planning module 43 uses an improved RRT* algorithm to support real-time replanning. The core improvements of this algorithm include:

[0132] 1. Bone density guided sampling: defining the sampling probability distribution function

[0133] P(x)=λ·e -β·D(x) ,

[0134] Where P(x) is the sampling probability at position x; D(x) is the bone density value at position x; λ is the normalization constant; β is the sensitivity coefficient, with a typical value of 0.005 cm 3 This sampling strategy makes the algorithm more inclined to sample in low-density areas, accelerating the planning process.

[0135] 2. Cost function considering bleeding volume: path cost calculation formula

[0136] Cost(p1,p2)=α·Distance(p1,p2)+β·BleedingRisk(p1,p2)+γ·DensityVariation(p1,p2),

[0137] Here, Cost(p1,p2) is the path cost from point p1 to point p2; Distance(p1,p2) is the Euclidean distance; BleedingRisk(p1,p2) is the bleeding risk assessment; DensityVariation(p1,p2) is the bone density variation assessment; α, β, and γ are weight coefficients, with typical values of 0.5, 0.3, and 0.2, respectively. This comprehensive cost function simultaneously considers path length, safety, and operational stability.

[0138] The real-time path planning module 43 designs a three-level path accuracy system: high-precision path, standard-precision path, and fast path. For the high-precision path, the algorithm uses a small step size (0.05mm) and a high sampling rate (2000 points); for the standard-precision path, the algorithm uses a medium step size (0.1mm) and a medium sampling rate (1000 points); and for the fast path, the algorithm uses a large step size (0.2mm) and a low sampling rate (500 points). This hierarchical strategy improves overall planning efficiency while ensuring accuracy in key areas.

[0139] The working principle of the anti-penetration security mechanism of the present invention is based on multi-source information fusion and multi-level protection strategy.

[0140] This mechanism integrates the following key technologies:

[0141] 1. Multi-scale tactile feedback: The system provides tactile information to the operator through a force feedback device. The feedback force increases with proximity to the danger zone. Force feedback modes include continuous force feedback and pulsed force feedback. The operator can choose the appropriate mode based on their preferences.

[0142] 2. Visual Enhancement: The system displays real-time safety status on the surgical navigation interface, using a red, yellow, and green color indicator system to intuitively reflect the current risk level. Additionally, the system provides distance numerical display and trend charts to help operators gain a more comprehensive understanding of safety status.

[0143] 3. Multi-level protection mechanism: The system has three protection zones: early warning (2mm), deceleration (1mm), and stop (0.5mm), each corresponding to a different intervention measure. This gradient protection strategy maximizes operator autonomy while ensuring safety.

[0144] 4. Dynamic Risk Assessment: The system comprehensively considers multiple factors, including bone density, bleeding volume, and vibration characteristics, to calculate a penetration risk score in real time, enabling timely intervention for high-risk procedures. Risk assessment utilizes fuzzy logic, which is smoother and more natural than traditional threshold judgments, reducing the possibility of false positives.

[0145] 5. Adaptive Threshold Adjustment: The system dynamically adjusts safety thresholds based on different surgical procedures, regional characteristics, and operator habits, providing personalized safety protection. This threshold adjustment is based on an expert system approach and incorporates extensive clinical experience data.

[0146] Through the organic combination of the above technologies, the anti-penetration safety mechanism of the present invention greatly reduces the risk of penetration and improves surgical safety, while maintaining sufficient operational flexibility and comfort, and has been widely recognized by clinicians.

[0147] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. An intelligent navigation system for alveolar bone modification based on image data fusion, characterized in that: include: Real-time bone density sensing module, used to obtain bone density values based on laser scanning data and perform temperature regulation; a personalized bone density adaptive control module, connected to the real-time bone density sensing module, for adjusting the bone trimming tool and performing temperature compensation according to the bone density value obtained by the real-time bone density sensing module; A multi-sensor data fusion module, connected to the personalized bone density adaptive control module, is used for dynamic monitoring of bleeding volume during bone repair and compensation of mechanical vibration of the equipment; The bone cutting path spatial perception module is connected to the multi-sensor data fusion module and the personalized bone density adaptive control module, and is used to obtain the bone cutting path based on the laser scanning data, CT data and optical coherence tomography data obtained by the real-time bone density perception module, and determine the real-time path change of the bone trimming tool in combination with the temperature and bleeding volume changes monitored by the multi-sensor data fusion module.

2. The intelligent navigation system for alveolar bone modification based on image data fusion according to claim 1, characterized in that: The real-time bone density sensing module includes: A thermocouple sensor module, mounted on the surface of the bone drill, is used to build an anti-penetration protection mechanism that combines multi-scale tactile feedback with visual enhancement. A laser point cloud sensing module is used to obtain initial point cloud data within the area drilled by the bone drill; LiDAR, used to obtain raw point cloud data of the area drilled by the bone drill; A data preprocessing module, connected to the laser point cloud sensing module and the laser radar, for filtering, denoising and smoothing the multimodal point cloud data; A path adjustment module is connected to the thermocouple sensor module and the data preprocessing module, and is used to adjust the cutting speed and trimming path changes based on the temperature data collected by the thermocouple sensor module in combination with the anti-penetration protection mechanism of the multi-scale tactile feedback and visual enhancement.

3. The intelligent navigation system for alveolar bone modification based on image data fusion according to claim 2, characterized in that: The personalized bone density adaptive control module includes: A bone density compensation module, configured to obtain bone density data based on the laser radar and perform personalized temperature compensation in combination with the real-time bone density sensing module; a bone trimming tool adjustment module, connected to the bone density compensation module, and configured to adjust the bone trimming tool according to the bone density data and surgical planning data acquired by the bone density compensation module; A path transmission module is connected to the bone trimming tool adjustment module and is used to transmit the position and posture of the bone trimming tool to the bone cutting path space perception module based on the surgical planning data and adjustment parameters to complete the real-time path planning for trimming; A temperature compensation module is connected to the bone density compensation module and the real-time bone density sensing module, and is used to perform temperature compensation on the bone density data obtained by the bone density compensation module and the personalized temperature compensation value obtained by the bone density compensation module and generate a temperature adjustment signal for the real-time bone density sensing module.

4. The intelligent navigation system for alveolar bone modification based on image data fusion according to claim 3, characterized in that: The multi-sensor data fusion module includes: A data acquisition module is used to respectively acquire pressure change and mechanical vibration data of the bone drill end through an optical fiber sensor and a piezoelectric sensor installed at the end of the bone trimming tool; a data preprocessing module, connected to the data acquisition module, for performing data filtering and noise elimination on the pressure change and mechanical vibration data, and then calculating frequency eigenvalues using Fourier transform; a bleeding volume monitoring module, connected to the data preprocessing module, for monitoring bleeding volume by combining the pressure change and the mechanical vibration data with a time series method; The equipment mechanical vibration compensation module is connected to the data preprocessing module and the bleeding volume monitoring module, and is used to filter the signal of the pressure value at the end of the bone trimming tool based on the time-frequency domain analysis method, extract the high repetition frequency vibration signal, and adjust the trimming path in real time to compensate for the mechanical vibration of the equipment.

5. The intelligent navigation system for alveolar bone modification based on image data fusion according to claim 4, characterized in that: The bone cutting path space perception module includes: The spatial perception module is used to obtain accurate registration of multimodal data by fusing high-resolution optical coherence tomography data, preoperative oral CT data, and lidar point cloud data using an image registration algorithm; A spatial coordinate conversion module, connected to the spatial perception module, is used to construct a multimodal data fusion framework based on the registration transformation matrix obtained by the spatial perception module to achieve multimodal data fusion; A real-time path planning module is connected to the spatial coordinate conversion module, the multi-sensor data fusion module and the personalized bone density adaptive control module, and is used to achieve real-time changes in the path trajectory of the bone drill end during the movement of the bone drill based on the fused bone density data and changes in bleeding volume.

6. The intelligent navigation system for alveolar bone modification based on image data fusion according to claim 5, characterized in that: The spatial perception module uses an iterative closest point algorithm to optimize the registration accuracy between different modal data; the high-resolution optical coherence tomography data provides the fine structure of surface tissue with a resolution of less than 10μm; the CT data provides deep bone structure information with a resolution of approximately 100μm; and the laser point cloud provides real-time spatial position information with a refresh rate greater than 20Hz.

7. The intelligent navigation system for alveolar bone modification based on image data fusion according to claim 5, characterized in that: The real-time path planning module uses an improved RRT* algorithm to support real-time replanning; the real-time path planning module designs a three-level path accuracy system: high-precision path, standard-precision path, and fast path; the error of the high-precision path is less than 0.1mm, the error of the standard-precision path is less than 0.3mm, and the error of the fast path is less than 0.5mm; the real-time path planning module automatically selects the appropriate path accuracy level based on regional bone density and sensitive structure distribution.

8. The intelligent navigation system for alveolar bone modification based on image data fusion according to claim 2, characterized in that: The sampling frequency of the thermocouple sensor module is set to 200 Hz to achieve real-time temperature monitoring; the thermocouple sensor module is designed with a nonlinear mapping function f(T) = α·T 2 +β·T+γ, converting temperature data into bone density estimation; where T is the temperature value, α, β, γ are coefficients related to bone properties; the thermocouple sensor module is based on the bone density threshold D1 = 300 mg / cm 3 and D2 = 600 mg / cm 3 Automatic partitioning.

9. The intelligent navigation system for alveolar bone modification based on image data fusion according to claim 4, characterized in that: The equipment's mechanical vibration compensation module is designed with a three-level safety assurance mechanism: a warning zone, a deceleration zone, and a stop zone. The safety distance in the warning zone is 2 mm, the safety distance in the deceleration zone is 1 mm, and the safety distance in the stop zone is 0.5 mm. The equipment's mechanical vibration compensation module provides force feedback to the operator through a tactile feedback device, and the force magnitude is inversely proportional to the proximity to the critical point. The equipment's mechanical vibration compensation module integrates bone density, bleeding volume, and vibration characteristics to calculate a penetration risk score R = w1·D + w2·B + w3·V, where D is the bone density value, B is the bleeding volume value, V is the vibration characteristic value, and w1, w2, and w3 are weight coefficients. When the R value exceeds the threshold of 0.8, the equipment's mechanical vibration compensation module automatically executes a deceleration or stop instruction.

10. An intelligent navigation method for alveolar bone modification, using the system according to any one of claims 1 to 9, characterized in that: The following steps are involved: S1, bone density acquisition, temperature adjustment, cutting speed and trimming path change through real-time bone density sensing module; S2. Through the personalized bone density adaptive control module, based on the surgical planning data and adjustment parameters, the position and posture of the bone trimming tool are transmitted to the bone cutting path spatial perception module to complete the real-time trimming path planning; S3, based on the piezoelectric acquisition signal of the multi-sensor data fusion module, uses the frequency domain analysis method to make real-time judgment on the force applied to the end of the bone trimming tool to compensate for the mechanical vibration of the equipment, and avoids bone trimming and cutting errors caused by mechanical vibration of the equipment by changing the bone trimming path; S4, performing multimodal data fusion based on the registration transformation matrix of the spatial perception module of the bone cutting path spatial perception module; Through the real-time path planning module, based on the fused bone density data and changes in bleeding volume, the path trajectory of the bone drill end is changed in real time during the movement of the bone drill.

Images