Oral dental pulp cleaning and repairing method and dental pulp diagnosis and treatment system thereof

Abstract

The invention provides an oral cavity dental pulp cleaning and repairing method and a dental pulp diagnosis and treatment system thereof, and relates to the technical field of oral industry, and the method comprises the following steps: fusing CBCT, OCT and other multi-modal images to construct a four-dimensional risk model to guide path planning; self-adaptive precise debridement is carried out under multi-sensing closed-loop feedback of pressure, acoustics, chemistry and the like; designing and 3D printing a personalized restoration adaptive to biomechanics through a topological optimization algorithm; and finally, long-term stability monitoring and early failure early warning of the diseased teeth are realized through the implantable passive sensor. The invention further discloses an intelligent diagnosis and treatment system for implementing the method. Through the integrated intelligent diagnosis and treatment process, the safety and success rate of complex root canal treatment and the long-term survival rate of diseased teeth are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of the oral industry, and specifically to an oral pulp cleaning and repair method and its pulp diagnosis and treatment system. Background Art

[0002] Root canal treatment is currently the main treatment method for preserving teeth with pulp necrosis or infection caused by pulp diseases, periapical diseases, etc. in the field of oral medicine. Its core goal is to thoroughly remove the infectious substances in the root canal system, including necrotic pulp tissue, bacteria and their metabolites, and then perform a tight three-dimensional filling and sealing of the root canal system to prevent reinfection and ultimately restore the chewing function of the affected tooth. There are still many limitations in existing conventional root canal treatment techniques in multiple key aspects such as diagnosis, debridement, repair, and prognosis monitoring, which directly affect the long-term success rate of treatment:

[0003] Current diagnosis mainly relies on two-dimensional periapical X-ray films and three-dimensional cone beam CT (CBCT). There are problems of image overlap in X-ray films, and although CBCT can provide three-dimensional structures, its spatial resolution is still insufficient to clearly display the microscopic anatomical structures in the root canal system, such as lateral accessory canals, root canal isthmuses, and the complex morphology of C-shaped root canals, and it is even more impossible to directly observe the microcracks on the root canal wall or the actual attachment of biofilms. In addition, existing diagnoses are all static evaluations and cannot dynamically predict the future development trend and invasiveness of the lesions, resulting in the lack of foresight in the formulation of treatment plans.

[0004] After the treatment is completed, the long-term effect evaluation of the affected tooth mainly depends on whether the patient has clinical symptoms and regular X-ray reexaminations. This monitoring mode is passive. When clinical symptoms appear or obvious bone defects are observed on the X-ray film, it often means that the infection has recurred or the repair has failed quite seriously, and it is easy to miss the best remedial opportunity. Summary of the Invention

[0005] Technical Problems to be Solved

[0006] In view of the deficiencies of the prior art, the present invention provides an oral pulp cleaning and repair method and its pulp diagnosis and treatment system, which solve the problems of the prior art.

[0007] Technical Solutions

[0008] To achieve the above objectives, the present invention is realized through the following technical solutions: An oral pulp cleaning and repair method includes the following steps:

[0009] Sp1. Multi-modal digital twin construction and risk prediction:

[0010] Fuse the three-dimensional structural data of the hard dental tissues obtained by cone-beam computed tomography and the microscopic structural and blood perfusion data of the internal soft tissues of the pulp chamber obtained by optical coherence tomography to construct a multi-scale high-fidelity digital twin model of the dentin-pulp complex; input the imaging features of the lesion area and the blood flow stagnation index measured by optical coherence tomography, predictively calibrate the invasiveness and liquefaction boundary of the lesion, and dynamically optimize the initial tensor field with the prediction results to form a four-dimensional risk assessment tensor field containing pathological trends.

[0011] Sp2, Adaptive debridement path planning and priority decision-making:

[0012] Based on the four-dimensional risk assessment tensor field, automatically plan a three-dimensional debridement path through a multi-objective optimization algorithm that combines minimum biological perturbation and maximum debridement efficiency.

[0013] Sp3, Closed-loop precise debridement based on acoustic streaming force coupling:

[0014] Adopt an ultrasonic cutting frequency to adaptively match the resonance frequency based on the dentin hardness and thickness distribution in the digital twin model to achieve the highest cutting efficiency on the premise of minimizing the risk of microcrack generation; then perform irrigation.

[0015] Sp4, Function-oriented repair and stress reconstruction:

[0016] After cleaning, the system calculates the ideal stress distribution model of the tooth after repair by simulating the tooth stress distribution under different chewing forces and combining the cumulative fatigue damage map of the dentin wall during the debridement process. The function-oriented repair and stress reconstruction include material gradient distribution repair and embedded structure repair. During the material gradient distribution repair, when it is found that the structural change in the repair section will cause the local stress to exceed 150 MPa of the dentin, search for and plan a secondary stress conduction path in the adjacent healthy tooth tissue, and guide the main load to the secondary path by adjusting the geometric shape of the prosthesis. The embedded structure repair uses a bionic dental pulp post printed by PEEK material 3D printing.

[0017] Sp5, Multi-modal long-term stability monitoring and self-consistent verification:

[0018] After the repair is completed, implant a passive micro-sensor array on the tooth crown surface or in the prosthesis for long-term monitoring of multi-source physiological parameters including bioelectrical impedance spectroscopy, temperature gradient, and acoustic emission signals; establish a personalized health baseline associated with the patient's digital twin model, and continuously analyze the data transmitted back by the sensors through cloud algorithms.

[0019] Preferably, for the four-dimensional risk assessment tensor field in Sp1, its time dimension is constructed through a pathological evolution model based on cellular automata. The pathological evolution model based on cellular automata uses the initial blood perfusion map and lesion boundary measured by optical coherence tomography as initial conditions to simulate the prediction path and rate of pulp tissue necrosis and inflammation spread.

[0020] Preferably, the multi-objective optimization algorithm in Sp2 is an improved ant colony algorithm, where the evaporation and update of pheromone depend on the path length, as well as the risk assessment tensor value of the area passed by the path, the predicted debris generation amount, and the proximity to the neurovascular bundle.

[0021] Preferably, in Sp3, it also includes a feedback sub based on the acoustic fingerprint of cavitation bubbles. By analyzing the sound signal spectrum when cavitation bubbles generated during ultrasonic debridement burst, it identifies the states of effective cleaning and ineffective or dangerous cleaning, and adjusts the ultrasonic power and frequency in real time to maximize the cleaning efficiency and inhibit excessive erosion of the root canal wall.

[0022] Preferably, in Sp3, when the Raman spectroscopy sensor detects the appearance of specific protein markers related to nerve tissue in the irrigation fluid, the system immediately pauses the mechanical cutting in this area and switches to a low-pressure and energy-free irrigation mode to protect the nerve bundle.

[0023] Preferably, the distribution strategy of the restoration in Sp4 is generated by a topology optimization algorithm, which takes maximizing the overall stiffness of the restored tooth and minimizing stress concentration as the optimization objectives and the physiological performance parameters of dentin as the constraint conditions.

[0024] Preferably, the bioelectrical impedance spectroscopy analysis in Sp5 is processed using a deep convolutional neural network. The deep convolutional neural network is trained with data to automatically identify weak but characteristic patterns related to microleakage, secondary caries, and root fractures from complex impedance spectra.

[0025] Preferably, the pulp diagnosis and treatment system of the oral pulp cleaning and restoration method includes:

[0026] A data acquisition module, including a cone beam computed tomography interface and an optical coherence tomography probe interface;

[0027] A computing and modeling server, pre-installed with digital twin construction software, a pathological evolution model, a path planning algorithm, and restoration design software;

[0028] An intelligent debridement execution unit, including a multi-degree-of-freedom robotic arm, a PEEK end effector integrated with an ultrasonic transducer and a fluid channel, and a control system and a fluid pump connected to the end effector;

[0029] The real-time sensing and feedback module includes a Raman spectrometer for analyzing the reflux liquid, a micro pressure sensor for monitoring the apical pressure, and a hydrophone for collecting the acoustic signals of air bubbles;

[0030] The prosthesis manufacturing and implantation module includes a multi-material DLP 3D printer and a navigation system for guiding the implantation of the prosthesis;

[0031] The long-term monitoring and warning terminal includes a signal reading device for receiving and processing the data of passive sensors, and a user interface for displaying the analysis results and warning information.

[0032] Preferably, a hard real-time control closed loop with a delay less than 10 milliseconds is formed among the computing and modeling server, the intelligent debridement execution unit, and the real-time sensing and feedback module to ensure the immediacy and safety of the system response during the high-speed debridement process.

[0033] Beneficial effects

[0034] The present invention provides an oral pulp cleaning and repair method and its pulp diagnosis and treatment system. It has the following beneficial effects:

[0035] 1. By integrating the macroscopic structure of CBCT, the microscopic structure of OCT, and the blood flow data, the present invention constructs a high-fidelity digital twin model, realizing the visualization diagnosis at the micron level inside the root canal for the first time, and being able to accurately identify microcracks and infection dead corners. Based on the chemical feedback closed loop of Raman spectroscopy analysis, the end point of debridement is upgraded from the traditional "mechanical shaping completion" to "biological cleaning completion". By real-time monitoring the marker concentrations of necrotic tissues and bacterial metabolites, it provides an objective and quantitative "cleaning" index for the operation, ensuring the thorough removal of all infected areas. This full-process precision from diagnosis to the judgment of the treatment end point overcomes the blindness of traditional technologies, can significantly improve the first success rate of root canal treatment, and reduce the protracted illness and secondary treatment caused by incomplete debridement.

[0036] 2. Through the real-time feedback closed loop of multiple sensors, the present invention places the treatment process under strict safety monitoring: Based on the PID closed loop control of the apical micro pressure sensor, the fluid pressure in the apical region can always be maintained below the safety threshold, effectively preventing the chemical and physical stimulation of the surrounding tissues of the apex; Based on the analysis of the acoustic fingerprint of cavitation bubbles collected by the hydrophone, the ultrasonic power can be optimized in real time, avoiding excessive cutting and damage to dentin while ensuring the debridement efficiency. Description of the drawings

[0037] Figure 1 It is the system diagram of the present invention;

[0038] Figure 2 It is the stage schematic diagram of the present invention;

[0039] Figure 3 It is the first case diagram of the present invention;

[0040] Figure 4 It is the second case diagram of the present invention. Specific implementation manners

[0041] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention. Specific embodiment one:

[0043] As Figures 1 to 2 shown, an oral pulp cleaning and restoration method includes the following steps:

[0044] Sp1. Multimodal digital twin construction and risk prediction:

[0045] Fuse the three-dimensional structural data of dental hard tissues with a voxel resolution of 75 microns obtained from cone-beam computed tomography and the microscopic structural and blood perfusion data of the internal soft tissues of the pulp cavity with a depth resolution of 10 microns obtained from optical coherence tomography. Construct a multi-scale high-fidelity digital twin model of the dentin-pulp complex through a multi-resolution image registration process. This registration process first performs coarse registration using a global affine transformation based on normalized mutual information, then applies a non-rigid registration algorithm with B-spline interpolation based on free-form deformation for fine alignment, and uses an L-BFGS optimizer to minimize the registration cost function. Based on a preset structural tensor analysis algorithm, by calculating the gradient covariance matrix within the neighborhood of each voxel in the CBCT image, determine its eigenvectors and eigenvalues, thereby calibrating the three-dimensional orientation and anisotropy intensity of dentinal tubules, and combining a neural network based on the U-Net architecture to learn and segment the probabilistic distribution path of neurovascular bundles from an anatomical atlas database. Finally, generate an initial tensor field representing the anisotropy of energy transfer and mass diffusion. Input the multi-dimensional feature vector of the lesion area, which is composed of data in 12 dimensions such as the CBCT gray value of the voxel, the texture features of the gray-level co-occurrence matrix, the OCT signal intensity, and the Doppler shift value, into a pre-trained model containing two stacked long short-term memory network layers with 128 hidden units each. This model makes a probability prediction of the invasiveness and liquefaction boundary of the lesion through a Softmax output layer, and dynamically optimizes the initial tensor field with the prediction results to form a four-dimensional risk assessment tensor field containing pathological trends. In the four-dimensional risk assessment tensor field in Sp1, the time dimension is constructed through a pathological evolution model based on cellular automata. The cell space of this model is the discretized grid of the digital twin model, and the cell states include healthy, inflamed, necrotic, and fibrotic. Its state transition rules are defined as follows: The transition probability P(H→I) of a healthy cell to the inflamed state is proportional to the number of inflamed cells in its neighborhood and the concentration of inflammatory mediators simulated by the discrete Laplace operator; the transition probability P(I→N) of an inflamed cell to the necrotic state is based on a nutrient consumption model, in which the nutrient level is inversely proportional to the local inflammatory cell density and the distance to the blood supply source calibrated by OCT. This model uses the initial blood perfusion map and lesion boundary measured by optical coherence tomography as the initial conditions to simulate the predicted path and rate of pulp tissue necrosis and inflammation spread within the next 24 to 72 hours.

[0046] Sp2, Adaptive debridement path planning and priority decision-making:

[0047] Based on a four-dimensional risk assessment tensor field, a multi-objective optimization algorithm that combines minimum biological perturbation and maximum debridement efficiency is used to automatically plan a three-dimensional debridement path. The path not only includes spatial coordinates but also the recommended operation timing sequence for each point along the path. The path is dynamically divided into a guiding intervention segment, a core lesion segment, and a dynamic reflux segment. The system generates a non-linear operation priority sequence based on the risk levels and operation timing of each segment, allowing the interruption of the current operation under specific conditions to prioritize the handling of higher-risk events. The multi-objective optimization algorithm in Sp2 is an improved ant colony algorithm. Its core is to minimize a cost function Cost(P) composed of four weighted terms: Cost(P)=w1·L(P)+w2·∫Risk(s)ds+w3·∫Debris(s)ds+w4·(1 / d_min), where L(P) is the total length of path P, ∫Risk(s)ds is the integral of the four-dimensional risk tensor values along the path, ∫Debris(s)ds is the total amount of debris generated predicted based on dentin hardness and instrument parameters, and d_min is the minimum Euclidean distance between the path and the neurovascular bundle. The pheromone update rule is Δτ_ij=(1-ρ)·τ_ij+ΣΔτ_ij^k, where ρ is the evaporation coefficient, and the pheromone contribution amount Δτ_ij^k of a single ant k is not only proportional to the reciprocal of the total path cost but also modulated by a local heuristic function η_ij, whose value is inversely proportional to the risk tensor value of the next node j, thus guiding the ant colony to perform local risk avoidance while globally optimizing.

[0048] Sp3. Closed-loop precise debridement based on the coupling of acoustic streaming force:

[0049] A debridement tool driven by ultrasonic energy and pulsed fluid is used, in which the ultrasonic cutting frequency is adaptively matched to the resonant frequency through an integrated phase-locked loop (PLL) circuit. The circuit compares the phase of the driving current with the vibration displacement of the instrument sensed by the piezoelectric sensor. The error signal generated is smoothed by a low-pass filter and input to the voltage-controlled oscillator to adjust the output frequency, so that the system is always locked at the resonance point with the smallest phase difference, that is, the highest energy transfer efficiency, so as to achieve the highest cutting efficiency while minimizing the risk of microcracks. After that, flushing is performed, and the flushing pressure and pulse mode are coupled with the geometric shape of the dynamic reflow section and the real-time fluid impedance, and closed-loop control is performed through a proportional-integral-differential controller running at a frequency of 100 Hz. The controller uses the measured value of the micro-pressure sensor in the apical area as input and uses u(t) =Kp·e(t)+Ki·∫e(t)dt+Kd·de(t) / dt formula is used to calculate the output value to directly control the rotation speed of the fluid pump, so as to maintain the pressure in the apical area always below the safety threshold of 15 mmHg; the biochemical components in the irrigation fluid are analyzed in real time by the Raman spectroscopy sensor integrated in the reflux path. The data processing flow includes: using a median filter to remove cosmic rays, using asymmetric least squares method for baseline correction, and then using a database containing pure spectra of various bacterial porphyrins, hydroxyproline, human serum albumin and other substances, through classical least squares regression analysis, the concentration of each marker is calculated in real time. When the concentration of a specific marker is lower than the clearance threshold for 5 consecutive sampling cycles, the system determines that the area has been cleared and automatically switches to the next path point, realizing intelligent judgment based on the chemical clearance endpoint. Sp3 also includes a feedback sub-step based on the acoustic fingerprint of cavitation bubbles. By performing a short-time Fourier transform on the acoustic signal collected by the hydrophone, the acoustic signal spectrum when the cavitation bubbles generated during ultrasonic debridement burst is analyzed. When the signal is detected to change from a stable harmonic peak to a high-intensity broadband noise, the transition from the stable cavitation state of effective cleaning to the ineffective or dangerous inertial cavitation state is identified, and the ultrasonic power is reduced by 5% to 10% in real time to maximize the cleaning efficiency and inhibit excessive erosion of the root canal wall. In Sp3, when the Raman spectroscopy sensor detects the presence of specific protein markers related to nerve tissue such as gangliosides or myelin basic protein in the irrigation fluid, the system immediately suspends mechanical cutting in the area and switches to a low-pressure, energy-free chemical irrigation mode with a flow rate of 0.5 ml per minute to protect the nerve bundle.

[0050] Sp4, function-oriented repair and stress reconstruction:

[0051] After cleaning, the system uses the finite element analysis method to simulate the stress distribution of the tooth body under different chewing forces in the range of 50 N to 500 N, and combines the cumulative fatigue damage map of the dentin wall caused during the debridement process calculated based on the Miner linear cumulative damage criterion to calculate the ideal stress distribution model of the restored tooth body; based on this ideal model, a functionally graded restoration plan is automatically designed, which includes material gradient distribution restoration and embedded structure restoration; a load-bearing path optimization mechanism is introduced. When it is calculated that any structural change in a restoration section will cause the local stress to exceed the physiological limit of 150 MPa of the dentin, the system automatically searches and plans a secondary stress conduction path in the adjacent healthy tooth tissue and guides the main load to this secondary path by adjusting the geometric shape of the restoration. The distribution strategy of the restoration material in Sp4 is generated by the solid isotropic material penalization topology optimization algorithm, which iteratively solves on the finite element mesh of the restoration space. Its objective function is to minimize the total strain energy of the structure, and the constraint condition is that the total volume of the restoration material shall not exceed the preset value. In each iteration, the pseudo-density ρ ∈ [0, 1] of each element is updated, and the Young's modulus of the element is defined as E(ρ) = E_min + ρ^p·(E_max - E_min), where the penalty factor p is usually taken as 3 to drive the pseudo-density to converge to 0 or 1. The finally obtained pseudo-density distribution map is converted into a CAD model for guiding the 3D printing of the functionally graded material. The embedded structure restoration method designed in Sp4 is a bionic dental pulp post printed by 3D using PEEK material, and its surface is modified with a bioactive coating that guides the directional differentiation of dental pulp stem cells. This coating is constructed by the electrostatic layer-by-layer self-assembly technique, alternately depositing positively charged poly-L-lysine and negatively charged sodium alginate pre-loaded with growth factors to form 20 bilayer structures, among which the inner 10 layers are loaded with vascular endothelial growth factor and the outer 10 layers are loaded with bone morphogenetic protein-2 to achieve controllable and sequential factor release, aiming to promote the regeneration and vascularization of the periapical tissue.

[0052] Sp5, Multi-modal Long-term Stability Monitoring and Self-consistent Verification:

[0053] After the repair is completed, a passive micro-sensor array is implanted on the surface of the dental crown or inside the restoration for long-term monitoring of multi-source physiological parameters including bioelectrical impedance spectroscopy with a frequency range from 1 Hz to 1 MHz, temperature gradient, and acoustic emission signals with frequencies above 100 kHz generated by the propagation of microcracks; a personalized health baseline associated with the patient's digital twin model is established, and the data transmitted back by the sensors is continuously analyzed through cloud algorithms. When a specific signal pattern indicating structural abnormalities is detected, the system issues early warnings to doctors and patients; when the system issues a warning, it automatically retrieves the historical data of the tooth for virtual stress testing. If the predicted five-year failure probability based on the probabilistic fracture mechanics model exceeds 5%, a follow-up visit is recommended and a detailed diagnostic report is provided. For the bioelectrical impedance spectroscopy analysis in Sp5, a one-dimensional deep convolutional neural network is used for processing. Its specific architecture is as follows: the input layer receives impedance data at 256 frequency points, followed by two "convolution-activation-pooling" modules, each containing 64 and 128 convolutional kernels of size 5, ReLU activation functions, and max pooling layers of size 2. Then, it is connected to a dense layer containing 100 neurons and a Dropout layer operating at a ratio of 0.5 through a flattening layer. Finally, the signal is classified as "healthy", "microleakage", or "root fracture" through a Softmax output layer. This deep convolutional neural network is trained using a dataset containing more than 50,000 samples, which has been verified and labeled through in vitro artificial defect models and clinical trials using Micro-CT and dye penetration methods, enabling it to automatically identify weak but characteristic patterns related to microleakage, secondary caries, and root fractures from complex impedance spectra.

[0054] Generally speaking, the above work process is as a whole as Figure 2 shown below:

[0055] Diagnostic planning for the first part:

[0056] 1. Data acquisition;

[0057] 1.1. The operator initiates a scan command;

[0058] 1.2. CBCT data is received, parsed, and verified through the DICOM interface;

[0059] 1.3. The OCT probe collects interference signals, and the GPU reconstructs the B-scan image sequence in real time;

[0060] 2. Four-dimensional digital twin construction;

[0061] 2.1. Perform multi-resolution precise registration on CBCT and OCT data;

[0062] 2.2. Generate the dentin structure tensor field and the probability map of neurovascular bundles;

[0063] 2.3. LSTM model predicts the initial lesion invasiveness and liquefaction boundary based on multimodal features;

[0064] 2.4. The cellular automaton model predicts future pathological trends and forms a 4D risk field;

[0065] 3. Intelligent path planning;

[0066] 3.1. CUDA-accelerated ant colony algorithm for multi-objective optimization based on 4D risk field;

[0067] 3.2. Generate a debridement pathway consisting of three segments: guidance, core, and reflow, and a nonlinear priority sequence;

[0068] Part II of Intelligent Debridement and Repair:

[0069] 4. Automated debridement execution;

[0070] 4.1. The robot moves the end effector along the planned path;

[0071] 4.2. Start the underlying hard real-time feedback loop (<5ms);

[0072] 4.3. Start the high-level chemical feedback loop (<100ms);

[0073] 5. Function-oriented repair;

[0074] 5.1. The server runs the topology optimization algorithm to design functional gradient restorations;

[0075] 5.2. Multi-material 3D printer manufactures bionic endodontic posts based on CAD models;

[0076] 5.3. Under the real-time guidance of the optical navigation system, the robotic arm accurately implants the restoration;

[0077] Part III Long-term Monitoring:

[0078] 6. Regular data backtesting;

[0079] 6.1. The patient uses an NFC reader device to collect data from the implanted passive sensor;

[0080] 6.2. Data is securely uploaded to the monitoring terminal software via Bluetooth;

[0081] 7. Intelligent analysis and early warning;

[0082] 7.1. The terminal software automatically updates and visualizes data trend charts;

[0083] 7.2. The pre-loaded CNN model analyzes the bioelectrical impedance spectrum and outputs the diagnostic probability of microleakage or root fracture;

[0084] 7.3. The system synthesizes all data, updates the five-year failure probability, and triggers a two-way warning for doctors and patients if it exceeds the preset threshold. Specific Embodiment 2:

[0086] As Figures 1 to 2 shown, the data acquisition module. The workflow of this module starts with the operator initiating a collection command. It includes a DICOMC-STORE service program that continuously listens on a specific port through the TCP / IP protocol. When receiving a C-STORERQ request from the in-hospital PACS system, the module will automatically verify the DICOM header information, parse and stream an image data set with up to 2048x2048x1920 voxels and 16-bit depth to the temporary storage area of the computing server, and verify the data integrity through the MD5 checksum after the transmission is completed; at the same time, it includes a forward-looking fiber-optic common-path optical coherence tomography probe with a diameter of 0.8 mm and a built-in MEMS scanning galvanometer. After being inserted into the root canal, its laser source is activated, and the returned interference spectral signal (A-scan) is collected in real time by a PCIe interface high-speed digitizer with a sampling rate of 200 MS / s and transmitted to the GPU video memory without delay through the direct memory access (DMA) technology. The dedicated CUDA kernel function running on the GPU will perform fast Fourier transform, dispersion compensation, and logarithmic compression on each A-scan data stream, reconstruct a B-scan cross-sectional image sequence in real time, and synchronize with the CBCT data stream to jointly form the original input for subsequent modeling.

[0087] Computing and Modeling Server, whose hardware configuration is a high-performance server equipped with dual Intel Xeon Platinum processors, 256GB of DDR4 error-correcting code memory, and four-way NVLink-interconnected NVIDIA A100 GPUs. Its core workflow automatically starts after detecting the completion of data acquisition: First, the digital twin construction software performs anisotropic diffusion filtering on CBCT data on the CPU to protect the edge details of dentinal tubules while reducing noise; Next, a pre-trained U-Net neural network performs semantic segmentation on the OCT image sequence on the GPU to accurately extract the three-dimensional surface of the pulp chamber. Then, a multi-resolution registration process is started. Using the segmented pulp chamber surface as a highly reliable target, a global affine transformation is first coarsely registered on the CPU, and then the result is used as the initial condition to parallelly compute a B-spline-based non-rigid deformation field on the GPU. By minimizing the normalized mutual information cost function through the L-BFGS optimizer, sub-voxel-level accurate alignment is achieved. After alignment, a GPU-accelerated structure tensor analysis module will calculate the anisotropy of the entire dentin region and perform weighted fusion with the probability map of the neurovascular bundle segmented from the anatomical atlas by another U-Net model to generate the initial tensor field. Subsequently, the 12-dimensional feature vector extracted from the multi-modal data is fed into a pre-trained model loaded onto the GPU, which contains two stacked long short-term memory network layers, to output the initial risk prediction. Finally, a cellular automaton simulation program implemented in C++ is started on the CPU. Using the prediction result of the LSTM and the OCT blood flow map as the initial conditions at t = 0, it iterates according to the reaction-diffusion equation containing the rules of inflammatory mediator diffusion and nutrient consumption, deduces a four-dimensional risk assessment tensor field containing future pathological trends, and structurally stores the complete digital twin and risk data.

[0088] The intelligent debridement execution unit, whose workflow is initiated by the priority path instruction sent by the server, includes a six-axis collaborative robotic arm with a repeatable positioning accuracy of ±0.02 mm, whose controller receives the path point sequence containing the spatial six-degree-of-freedom posture sent by the server at a frequency of 60Hz, and generates a smooth joint motion trajectory through inverse kinematics solution and cubic spline interpolation; a PEEK end effector made of PEEK material, which can be sterilized at high temperature and high pressure, and integrates a PZT-8 piezoelectric ceramic ultrasonic transducer and a coaxial dual fluid channel, wherein the transducer has an operating frequency range of 25-45 kHz, and the inner and outer fluid channels are used for irrigation and reflux suction respectively; and a real-time control system based on a field programmable gate array and a precision peristaltic pump connected to the actuator. When the robotic arm moves, the FPGA controller synchronously activates the peristaltic pump and the ultrasonic generator, and immediately starts a built-in phase-locked loop circuit, which automatically finds the optimal resonant frequency under the current dentin structure within milliseconds by comparing the phase difference between the driving current and the vibration signal fed back by the piezoelectric sensor, thereby maximizing the energy transfer efficiency. During the entire debridement process, the FPGA controller independently and quickly executes underlying safety and efficiency assurance tasks.

[0089] The real-time sensing and feedback module is the core of the hard real-time closed loop, and its data flow runs uninterruptedly. It includes a compact Raman spectrometer integrated with the end effector through optical fiber, using a 785-nanometer excitation light source. Its CCD sensor collects spectra at a frequency of 20 Hz. The collected raw data is processed by a dedicated digital signal processor to perform cosmic ray removal, baseline correction and classical least squares solution, and the solved chemical composition concentration vector is sent to the computing server via Ethernet for high-level decision-making; an optical fiber micro pressure sensor based on the Fabry-Perot interferometer principle integrated at the top of the PEEK end effector, whose optical interrogator outputs the decoded pressure value in the form of an analog voltage signal; and a micro needle hydrophone with a bandwidth of up to 20 MHz integrated in the reflux suction pipeline, whose output analog acoustic signal covers the full spectrum of cavitation activity. The analog outputs of pressure and acoustic signals are directly fed into the FPGA real-time control system of the intelligent debridement execution unit and digitized by its internal high-speed analog-to-digital converter. The digitized data are directly used as input for the PID pressure control algorithm and the FFT acoustic fingerprint analysis algorithm. The calculation results of these two algorithms directly generate control instructions for the peristaltic pump motor and ultrasonic generator power, forming a complete underlying hardware-level feedback loop with a delay of less than 5 milliseconds.

[0090] The prosthesis manufacturing and implantation module starts its manufacturing process after the server completes the restoration design. It includes a multi-material DLP 3D printer that uses multi-material digital light processing technology with a printing resolution of 15 microns. The server sends the STL model of the functionally graded prosthesis to the printer, and its built-in software slices the model and assigns an accurate material mixing ratio to each layer. During printing, the DLP projector dynamically adjusts the lighting pattern according to the material distribution map of each layer, selectively curing the mixtures of different resins, stacking layer by layer until the prosthesis is formed, and finally entering the automated cleaning and ultraviolet post-curing unit. Its implantation process is as follows: The patient wears a dynamic reference dental arch splint attached with a tracer ball. An optical space navigation system composed of a binocular infrared camera first clicks on the instrument tracking target points on the tooth surface through a probe to accurately register the patient's dental arch with the preoperative CBCT data. Subsequently, the printed prosthesis is installed on a special fixture at the end of the robotic arm, and the tracer ball on the fixture is identified and registered. At this time, the virtual position of the prosthesis relative to the three-dimensional tooth model can be displayed in real time on the system screen. Under the guidance of the navigation system, the operator manipulates the robotic arm along the preset implantation path to accurately place the prosthesis into the cleaned root canal with sub-millimeter precision.

[0091] The long-term monitoring and early warning terminal has a working process of regular or on-demand inspections. It includes an NFC handheld signal reading device that communicates wirelessly with the implantable passive sensor through the near-field communication protocol. During inspections, the antenna coil of the reading device generates an electromagnetic field to power the sensor. After the sensor is awakened, it sequentially performs bioelectrical impedance spectroscopy scanning, temperature measurement, and acoustic emission event monitoring, and transmits the encrypted data packets collected back to the reading device through load modulation technology; and a graphical user interface running on a tablet or workstation. The reading device synchronizes the data to this software via Bluetooth. The software first decrypts and verifies the data, then immediately updates the new data points to the historical trend graph, and at the same time inputs the latest impedance spectrum data into the loaded deep convolutional neural network model composed of three one-dimensional convolutional layers and two fully connected layers for analysis. The classification results output by the model and the five-year failure probability recalculated based on the new data will be immediately displayed on the dashboard of the terminal main interface to complete a full health status assessment and early warning process.

[0092] A hard real-time control closed-loop with a data transmission and processing delay less than 10 milliseconds is formed among the computing and modeling server, the intelligent debridement execution unit, and the real-time sensing and feedback module. The specific path of this closed-loop is as follows: Signals collected by the pressure sensor and the hydrophone are subjected to PID and phase-locked loop operations through the FPGA real-time control system to directly generate control instructions for the peristaltic pump and the ultrasonic generator. The response time of this underlying safety loop is ensured to be within 5 milliseconds. At the same time, the real-time sensing data is uploaded to the computing and modeling server in the form of data packets. The server makes macroscopic decisions based on the Raman spectroscopy analysis results and path planning, and issues the updated path instructions to the robotic arm controller. The cycle of the entire high-level feedback loop is strictly limited within 10 milliseconds to ensure the immediacy and absolute safety of the system response during high-speed debridement. Specific Embodiment Three:

[0094] Based on the technical solutions of Specific Embodiment One and Specific Embodiment Two, a case is further given for supplementary explanation:

[0095] Such as Figure 3 Case One shown in the figure: Complex debridement and restoration of the C-shaped root canal of the mandibular second molar. Chief complaint: Slight discomfort when eating hot food in the lower right posterior tooth, no spontaneous pain, and a history of occlusal discomfort. Figure 3 From left to right in the figure, (1), (2), and (3) are the CBCT, OCT, and visualization images after the technical solution of this application respectively. The left side shows the original images from multiple sources, and each image has its inherent advantages and disadvantages, jointly constituting a complex diagnosis.

[0096] The conventional X-ray film shows a blurred low-density shadow around the apex of the right lower second molar (#47), and the pulp vitality test is non-responsive. The preliminary diagnosis is chronic apical periodontitis. The clinical challenge lies in that the incidence of the C-shaped root canal of the mandibular second molar is high, and its anatomical structure such as the isthmus, fin-shaped communicating branches, and irregular depressions is extremely complex. Traditional instruments are difficult to thoroughly debride, which is a high-risk factor for the failure of conventional root canal treatment. The long-term monitoring and warning terminal also includes a graphical user interface software and an implantable passive multi-modal sensor, and the graphical user interface software includes a three-dimensional data visualization module, a one-dimensional CNN diagnostic analysis engine, and a warning and report generator.

[0097] Sp1, Precise Diagnosis and Four-Dimensional Modeling: The system first fuses CBCT and OCT data. CBCT clearly outlines that the root canal of #47 presents a typical "C"-shaped fusion, but the internal details are blurred. Subsequently, the operator inserts an 0.8-mm OCT probe into the pulp chamber, and the system scans and reconstructs the high-resolution three-dimensional image of the root canal system in real time. The OCT image clearly reveals a large amount of signals seemingly of necrotic tissue in the wide isthmus connecting the buccal and lingual main root canals, and calibrates two fin-shaped communication branches that are extremely difficult to reach with traditional instruments. Based on these fine structures, the four-dimensional risk assessment tensor field of the system marks these areas as "high-risk biofilm enrichment areas" and predicts that the liquefaction boundary has a tendency to extend towards the apex within the next 48 hours.

[0098] Sp2, Multi-Objective Path Planning: The ant colony algorithm on the computing server starts the planning. The objective function of the algorithm not only needs to minimize the path, but more importantly, it needs to maximize the coverage of the C-shaped isthmus and fin-shaped communication branches, while strictly avoiding the weakest areas of the root canal wall. Finally, the system generates not a single straight path, but a "sweeping" three-dimensional debridement path that covers the entire C-shaped system, and sets different operation priorities and recommended energy parameters for different regions (main root canal, isthmus, communication branch).

[0099] Sp3, Closed-Loop Precise Debridement: The intelligent debridement execution unit starts to work. The robotic arm guides the end effector to move along the "sweeping path". When entering the wide isthmus, the phase-locked loop (PLL) circuit automatically adjusts the ultrasonic frequency to the resonant frequency that can stimulate the maximum cavitation effect for efficient non-contact acoustic streaming debridement. At the same time, the Raman spectroscopic sensor in the reflux fluid continuously monitors the intensity of the porphyrin signal representing bacterial metabolites. The system finds that even after the instrument leaves, the porphyrin signal in the isthmus area still has a slight increase, indicating the existence of deep biofilms. The system then instructs the robotic arm to perform secondary, low-power acoustic streaming cleaning on this area until the porphyrin signal is below the clearance threshold within consecutive sampling periods, truly achieving "chemically clean".

[0100] Sp4, Function-Oriented Repair: After the cleaning is completed, the system enters the repair design stage. Considering the thin and irregular C-shaped root canal wall, the topology optimization algorithm starts to design a non-circular bionic dental pulp post that perfectly matches the complex root canal morphology with the goal of maximizing the fracture resistance. The post is solidified in the areas that need support, while being grid-shaped in the non-load-bearing areas to save materials and enhance the locking with the sealer. The multi-material 3D printer prints this repair body using a highly tough nanocomposite resin.

[0101] Results: The postoperative recovery was good, and the discomfort with hot food disappeared. The postoperative CBCT showed that the root canal system was tightly filled without dead ends.

[0102] Sp5, Stability Feedback: A passive sensor is implanted in the restoration. During the follow-up examinations at the 6th and 12th months after the operation, Mr. Zhang used an NFC handheld device to collect data. The terminal software showed that his bioelectrical impedance spectrum was highly consistent with the healthy baseline immediately after the operation, and the CNN diagnostic engine did not detect any characteristic patterns related to microleakage, objectively verifying the long-term sealing and stability of the restoration.

[0103] OCT-CBCT fusion modeling reveals complex anatomical details that cannot be shown by traditional imaging, while the chemical feedback closed-loop based on Raman spectroscopy ensures thorough debridement of these complex regions, fundamentally solving the problem of C-shaped root canal treatment.

[0104] As Figure 4 shown, Case 2: Diagnosis and internal retention restoration of incomplete root fracture after maxillary anterior tooth trauma. Chief complaint: The upper left central incisor (#21) was traumatized one year ago. After root canal treatment in an external hospital, occlusal pain and repeated gum abscesses have gradually occurred in the past six months. Figure 4 From left to right are the CBCT, OCT, and images processed by this technical solution. The CBCT image on the left shows a "J"-shaped bone defect of the root fracture. The middle OCT image penetrates into the root canal and clearly captures the longitudinal microcracks on the dentin wall with high resolution. The image on the right after processing, after fusing the data of CBCT and OCT, its built-in AI segmentation algorithm automatically and accurately depicts the real and clear anatomical structure on the low-quality CBCT image and automatically superimposes a semi-transparent color model.

[0105] CBCT shows a "J"-shaped bone defect at the apex of #21, which is a typical indirect sign of vertical root fracture (VRF), but the crack line itself is not visible. The conventional treatment method is to extract the affected tooth. The patient is young and hopes to retain the natural tooth to the greatest extent possible. The clinical challenge lies in how to clearly diagnose VRF during the operation and explore the possibility of a tooth-preserving treatment.

[0106] Sp1, Microscopic Diagnosis and Qualification: After removing the old root canal filling material, the OCT probe was introduced into the root canal for a 360-degree scan. In the middle section of the buccal side wall of the root canal, the OCT image clearly shows a microcrack line with a depth of about 0.4 mm and a width of about 20 microns. This crack does not penetrate the entire length of the root. This direct evidence confirms the diagnosis of incomplete vertical root fracture. The system's risk model marks the crack line area as an "extremely high mechanical risk area".

[0107] Sp2 & Sp3, Ultra - low disturbance debridement: The system plans a "zero - contact" debridement path, and the tip of the instrument always maintains a safe distance from the tube wall on the opposite side of the crack. During debridement, the acoustic fingerprint of cavitation bubbles collected by the hydrophone is used for real - time monitoring. Once the inertial cavitation characteristics (high - intensity broadband noise) that may exacerbate the crack are detected, the FPGA real - time control system will instantaneously (less than 5 milliseconds) reduce the ultrasonic power, ensuring effective disinfection while not causing secondary damage to the affected tooth.

[0108] Sp4, Crack internal retention repair: This is the core of the tooth - saving attempt. The topology optimization algorithm takes "restricting stress concentration at the crack tip" as the primary goal and designs a special repair plan. This plan is not a traditional post - core but an "internal retention splint": on the inner side of the restoration corresponding to the crack line, the algorithm specifies the use of a flexible composite material with high fracture toughness and low elastic modulus; while for other parts of the tooth, a high - strength rigid composite material is used. The 3D printer manufactures this functionally graded restoration, which can absorb and disperse the occlusal force like an internal "shock absorber" after implantation, preventing the crack from continuing to expand.

[0109] Sp5, Acoustic emission "sentinel" monitoring: Given the high risk of VRF, a monitoring array containing acoustic emission sensors is implanted in the crown restoration of #21. During each review, the system will require Xiao Li to perform a specified - intensity biting action. If even a micron - level expansion occurs at the crack tip, the stress wave generated will be captured by the AE sensor and recorded in the form of high - frequency acoustic emission events. Monitoring for one year after the operation shows that the AE event count is always zero, proving that the "internal retention splint" repair plan effectively controls the crack expansion.

[0110] The microscopic diagnostic ability of OCT changes VRF from "guesswork" to "definitive diagnosis", providing a decision - making basis for extreme tooth preservation. The long - term monitoring of the functionally graded restoration based on topology optimization and the acoustic emission sensor together constitute a complete "diagnosis - repair - monitoring" closed - loop, minimizing the risk of failure.

[0111] It should be noted that, in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising a reference structure" does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0112] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An oral pulp cleaning and restoration method, characterized in that: Including the following steps: Sp1. Multimodal digital twin construction and risk prediction: Fusing the three-dimensional structural data of dental hard tissues obtained by cone-beam computed tomography and the microscopic structural and blood perfusion data of the internal soft tissues of the pulp cavity obtained by optical coherence tomography to construct a multi-scale high-fidelity digital twin model of the dentin-pulp complex; inputting the imaging features of the lesion area and the blood flow stagnation index measured by optical coherence tomography to predictively calibrate the invasiveness and liquefaction boundary of the lesion, and dynamically optimizing the initial tensor field with the prediction results to form a four-dimensional risk assessment tensor field including pathological trends; Sp2. Adaptive debridement path planning and priority decision-making: Based on the four-dimensional risk assessment tensor field, automatically plan a three-dimensional debridement path through a multi-objective optimization algorithm that combines minimum biological perturbation and maximum debridement efficiency; Sp3. Closed-loop precise debridement based on acoustic streaming force coupling: Adopt an ultrasonic cutting frequency to adaptively match the resonance frequency based on the dentin hardness and thickness distribution in the digital twin model to achieve the highest cutting efficiency on the premise of minimizing the risk of microcrack generation; then perform irrigation; Sp4. Function-oriented repair and stress reconstruction: After cleaning, the system calculates the ideal stress distribution model of the tooth after repair by simulating the stress distribution of the tooth under different chewing forces and combining the cumulative fatigue damage map of the dentin wall caused during the debridement process. The function-oriented repair and stress reconstruction include material gradient distribution repair and embedded structure repair. During the material gradient distribution repair, when it is found that the structural change in the repair section will cause the local stress to exceed 150 MPa of the dentin, search for and plan a secondary stress conduction path in the adjacent healthy tooth tissue, and guide the main load to the secondary path by adjusting the geometric shape of the restoration. The embedded structure repair uses a bionic dental pulp post printed by PEEK material 3D; Sp5. Multimodal long-term stability monitoring and self-consistent verification: After the repair is completed, implant a passive micro-sensor array on the tooth crown surface or in the restoration to long-term monitor multi-source physiological parameters including bioelectrical impedance spectroscopy, temperature gradient, and acoustic emission signal; establish a personalized health baseline associated with the patient's digital twin model, and continuously analyze the data transmitted back by the sensor through cloud algorithms.

2. The oral pulp cleaning and restoration method according to claim 1, characterized in that For the four-dimensional risk assessment tensor field in Sp1, its time dimension is constructed through a pathological evolution model based on cellular automata. The pathological evolution model based on cellular automata uses the initial blood perfusion map and lesion boundary measured by optical coherence tomography as initial conditions to simulate the prediction path and rate of pulp tissue necrosis and inflammation spread.

3. A method for oral pulp cleaning and restoration according to claim 1, characterized in that, The multi-objective optimization algorithm in Sp2 is an improved ant colony algorithm, where the evaporation and update of pheromone depend on the path length, as well as the risk assessment tensor value of the area passed by the path, the predicted debris generation amount, and the proximity to the neurovascular bundle.

4. A method for oral pulp cleaning and restoration according to claim 1, characterized in that, In the Sp3, there is also a feedback sub-module based on the acoustic fingerprint of cavitation bubbles. By analyzing the acoustic signal spectrum when the cavitation bubbles generated during the ultrasonic debridement process rupture, it can identify the states of effective cleaning and ineffective or dangerous cleaning, and adjust the ultrasonic power and frequency in real time to maximize the cleaning efficiency and inhibit excessive erosion of the root canal wall.

5. A method for oral pulp cleaning and restoration according to claim 1, characterized in that, In the Sp3, when the Raman spectroscopy sensor detects the appearance of specific protein markers related to nerve tissue in the irrigation fluid, the system immediately pauses the mechanical cutting in this area and switches to a low-pressure and energy-free irrigation mode to protect the nerve bundle.

6. A method for oral pulp cleaning and restoration according to claim 1, characterized in that, The distribution strategy of the restoration in the Sp4 is generated by a topology optimization algorithm, which takes maximizing the overall stiffness of the restored tooth and minimizing stress concentration as the optimization objectives and the physiological performance parameters of dentin as the constraint conditions.

7. A method for oral pulp cleaning and restoration according to claim 1, characterized in that, The bioelectrical impedance spectroscopy analysis in the Sp5 is processed using a deep convolutional neural network. The deep convolutional neural network is trained with data to automatically identify weak but characteristic patterns related to microleakage, secondary caries, and root fractures from complex impedance spectra.

8. An endodontic diagnosis and treatment system for implementing the endodontic cleaning and restoration method according to any one of claims 1 to 7, characterized in that, It includes: A data acquisition module, including a cone beam computed tomography interface and an optical coherence tomography probe interface; A computing and modeling server, pre-installed with digital twin construction software, a pathological evolution model, a path planning algorithm, and restoration design software; An intelligent debridement execution unit, including a multi-degree-of-freedom robotic arm, a PEEK end effector integrated with an ultrasonic transducer and a fluid channel, and a control system and a fluid pump connected to the end effector; A real-time sensing and feedback module, including a Raman spectrometer for analyzing the reflux fluid, a micro pressure sensor for monitoring the apical pressure, and a hydrophone for collecting the acoustic signals of air bubbles; A restoration manufacturing and implantation module, including a multi-material DLP 3D printer and a navigation system for guiding the implantation of the restoration; A long-term monitoring and warning terminal, including a signal reading device for receiving and processing passive sensor data, and a user interface for displaying the analysis results and warning information.

9. The endodontic diagnosis and treatment system according to claim 8, characterized in that, A hard real-time control closed-loop with a delay less than 10 milliseconds is formed among the computing and modeling server, the intelligent debridement execution unit, and the real-time sensing and feedback module to ensure the immediacy and safety of the system response during high-speed debridement.

Images