Multimodal intelligent dental diagnosis and treatment method and system

Abstract

The application provides a multi-modal intelligent dental diagnosis and treatment method and system, comprising: an intelligent saliva suction device pre-processes, bubble detects and eliminates saliva samples with a collection standard, and obtains processed saliva samples; the processed saliva samples are distributed to multiple channels; an exhalation sampling headrest collects breathing samples, and a flexible wearable unit collects physiological data; the processed saliva samples, the breathing samples and the physiological data are used as multi-dimensional heterogeneous data; the multi-dimensional heterogeneous data is input into corresponding equipment to obtain electrochemical data and spectral data; the electrochemical data and the spectral data are pre-processed and feature-extracted to obtain feature data; the feature data is divided into multiple stages by using a sliding time window mechanism, high-frequency data is statistically down-sampled, the feature data is subjected to a multi-modal fusion engine, and the current state of a patient is compared with historical personal baseline to output a deviation score and a radar chart.

Description

Technical Field

[0001] This invention relates to the fields of high-end medical devices and smart healthcare technology, and in particular to multimodal intelligent dental diagnosis and treatment methods and systems. Background Technology

[0002] Currently, detection technologies for saliva, breath, and physiological parameters are relatively mature, such as microfluidic dry biochemical analyzers, Raman spectroscopy for bacterial identification, and wearable flexible electrochemical sensors. However, these devices currently exist mostly as standalone laboratory instruments or handheld devices, existing in a data silo state.

[0003] Existing equipment only outputs a single indicator and lacks an AI-based comprehensive analysis model that combines oral microbial data, physiological parameters, and biochemical indicators, making it difficult to achieve accurate risk assessment of systemic diseases. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a multimodal intelligent dental treatment method and system that combines electrochemical sensing, Raman spectroscopy and microfluidic technology to achieve comprehensive coverage from local oral lesions to systemic metabolic status; establish a health model based on "personal baseline" and identify subtle abnormalities through AI algorithms, transforming passive treatment into proactive health management and risk warning.

[0005] In a first aspect, embodiments of the present invention provide a multimodal intelligent dental treatment method, the method comprising: The intelligent saliva aspirator preprocesses, detects and eliminates air bubbles in saliva samples that meet the required volume, and then obtains the processed saliva sample. The processed saliva sample is distributed into multiple channels; Breath sampling headrest collects respiratory samples, while flexible wearable unit collects physiological data; The processed saliva sample, the respiratory sample, and the physiological data are used as multidimensional heterogeneous data. The multidimensional heterogeneous data is input into the corresponding device to obtain electrochemical data and spectral data; The electrochemical data and the spectral data are preprocessed and feature extracted to obtain feature data; The feature data is used with a sliding time window mechanism to divide a single diagnosis and treatment process into multiple stages, and high-frequency data is statistically downsampled. The feature data is processed by a multimodal fusion engine, and the patient's current status is compared with their historical personal baseline to output a deviation score and a radar chart.

[0006] Furthermore, the plurality of channels includes a first channel, a second channel, a third channel, a fourth channel, a fifth channel, and a sixth channel; the first channel is an electrochemical sensor, the second channel is an immunochromatographic assay, the third channel is a microfluidic biochemical chip, the fourth channel is a nucleic acid extraction device, the fifth channel is a sample retention chamber, and the sixth channel is a Raman sample pool.

[0007] Furthermore, the multidimensional heterogeneous data is input into the corresponding device to obtain electrochemical data and spectroscopic data, including: The processed saliva sample was input into a microfluidic electrochemical sensor to obtain peak current, peak potential, transfer resistance, and reaction kinetic constants. The processed saliva sample was input into a SERS-Raman spectrometer to obtain the characteristic peak positions, peak intensity ratios, and full width at half maximum (FWHM). The processed saliva sample was input into a microfluidic dry biochemical analyzer to obtain liver and kidney function indicators and blood lipid indicators. The respiratory sample is input into the expiratory sensor to obtain the response time, maximum response amplitude, and recovery rate. The physiological data is input into the flexible wristband to obtain HRV, blood oxygen saturation, and blood pressure.

[0008] Furthermore, the electrochemical data and the spectral data are preprocessed and feature extracted to obtain feature data, including: Based on the baseline drift caused by large fluctuations in saliva viscosity, an adaptive filter is used for noise processing to obtain the processed baseline drift. The baseline subtraction algorithm was applied to the CV curve to extract the Faraday current value of the redox peak.

[0009] Furthermore, the electrochemical data and the spectral data are preprocessed and feature extracted to obtain feature data, including: Based on the fluorescence background interference in Raman spectroscopy, an asymmetric least squares smoothing algorithm is used for baseline correction. Principal component analysis was used for dimensionality reduction to extract the fingerprint features of bacterial cell walls.

[0010] Furthermore, the feature data is processed by a multimodal fusion engine, including: Local features were extracted by processing Raman spectra and electrochemical curves using 1D-CNN; By processing PPG and respiratory time-series data using LSTM, physiological change trends can be captured. The system automatically calculates the weights between different modalities, concatenates the feature vectors from each mode, and inputs them into a fully connected layer to output the patient's current physiological state vector.

[0011] Furthermore, the method also includes: Based on the deviation score, a warning traffic light system is set up, and multiple warning levels are divided.

[0012] Secondly, embodiments of the present invention provide a multimodal intelligent dental treatment system, the system comprising: A smart saliva aspirator is used to preprocess, detect and eliminate air bubbles in saliva samples that have reached the required volume, and then distribute the processed saliva samples into multiple channels. Breath sampling headrest, used for collecting breath samples; Flexible wearable units for collecting physiological data; The AI ​​data processing module is used to treat the processed saliva sample, the respiratory sample, and the physiological data as multidimensional heterogeneous data; input the multidimensional heterogeneous data into the corresponding devices to obtain electrochemical data and spectral data; preprocess and extract features from the electrochemical data and the spectral data to obtain feature data; use a sliding time window mechanism to divide a single diagnosis and treatment process into multiple stages, and perform statistical downsampling on high-frequency data; use a multimodal fusion engine to process the feature data, and compare the patient's current status with historical personal baselines to output deviation scores and radar charts.

[0013] Thirdly, embodiments of the present invention provide an electronic device, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the method described above.

[0014] Fourthly, embodiments of the present invention provide a computer-readable medium having processor-executable non-volatile program code that causes the processor to perform the method described above.

[0015] This invention provides a multimodal intelligent dental treatment method and system, comprising: a smart saliva aspirator preprocessing a saliva sample of sufficient volume, detecting and eliminating air bubbles to obtain a processed saliva sample; distributing the processed saliva sample to multiple channels; a breath sampling headrest collecting respiratory samples, and a flexible wearable unit collecting physiological data; using the processed saliva sample, breath sample, and physiological data as multidimensional heterogeneous data; inputting the multidimensional heterogeneous data into corresponding devices to obtain electrochemical data and spectral data; and preprocessing and extracting features from the electrochemical data and spectral data. The system obtains feature data; it uses a sliding time window mechanism to divide a single treatment process into multiple stages and performs statistical downsampling on high-frequency data; it then uses a multimodal fusion engine to process the feature data and compares the patient's current status with their historical personal baseline to output deviation scores and radar charts; it combines electrochemical sensing, Raman spectroscopy, and microfluidic technology to achieve comprehensive coverage from local oral lesions to systemic metabolic status; and it establishes a health model based on the "personal baseline," using AI algorithms to identify subtle abnormalities, transforming passive treatment into proactive health management and risk warning.

[0016] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0017] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0018] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart of the multimodal intelligent dental treatment method provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the sampling end, processing end, and cloud processing process provided in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the multimodal intelligent dental treatment process provided in Embodiment 1 of the present invention; Figure 4 This is a schematic diagram of another multimodal intelligent dental treatment process provided in Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the multimodal intelligent dental treatment system provided in Embodiment 2 of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] To facilitate understanding of this embodiment, the embodiments of the present invention will be described in detail below.

[0022] Example 1: Figure 1 This is a flowchart of the multimodal intelligent dental treatment method provided in Embodiment 1 of the present invention.

[0023] Reference Figure 1 The method includes the following steps: Step S101: The intelligent saliva aspirator preprocesses the saliva sample with the required amount, detects and eliminates air bubbles, and then obtains the processed saliva sample. Step S102: Distribute the processed saliva sample into multiple channels; Step S103: Exhalation sampling headrest collects respiratory samples, and flexible wearable unit collects physiological data; Here, the exhalation sampling headrest integrates a gas collection port on the side wing of the dental chair headrest, with a built-in sensor and gas chromatography column to capture volatile markers such as hydrogen sulfide, acetone, and nitric oxide.

[0024] Flexible wearable unit: medical silicone base, embedded flexible PCB, PPG sensor on the back of the wrist, electrochemical sensor on the palm of the wrist, and skin conductivity electrode.

[0025] Reference Figure 2 The base integrates the equipment compartment, which consists of a top layer, a middle layer, and a bottom layer.

[0026] Top layer: Sample preprocessing and allocation 1) Microfluidic intelligent distribution system; 2) Viscosity adjustment.

[0027] Middle layer: Core analysis module 1) Microfluidic dry biochemical analyzer: detects biochemical indicators such as liver and kidney function and blood lipids; 2) Fluorescence immunoassay analyzer: Utilizes antigen-antibody reactions to detect markers such as high-sensitivity troponin and IL-6; 3) Microbial Raman Analyzer: Utilizes surface-enhanced Raman scattering technology to distinguish between dead and live bacteria and drug resistance.

[0028] Bottom layer: Support modules 1) Intelligent power management and backup power; 2) Thermal management system; 3) Dehumidification module and waste liquid recovery.

[0029] Step S104: The processed saliva sample, breath sample, and physiological data are used as multidimensional heterogeneous data. Step S105: Input the multidimensional heterogeneous data into the corresponding device to obtain electrochemical data and spectral data; Here, at the data input layer of the AI ​​data processing center, multidimensional heterogeneous data is standardized and input into the corresponding devices to obtain electrochemical data and spectral data.

[0030] Step S106: Preprocess and extract features from the electrochemical and spectral data to obtain feature data; Step S107: The feature data is divided into multiple stages by using a sliding time window mechanism, and high-frequency data is statistically downsampled. Step S108: Perform multimodal fusion on the feature data; Step S109: Compare the patient's current condition with their historical personal baseline and output the deviation score and radar chart.

[0031] Furthermore, the multiple channels include a first channel, a second channel, a third channel, a fourth channel, a fifth channel, and a sixth channel; the first channel is an electrochemical sensor, the second channel is an immunochromatographic assay, the third channel is a microfluidic biochemical chip, the fourth channel is a nucleic acid extraction, the fifth channel is a sample retention chamber, and the sixth channel is a Raman sample pool.

[0032] Specifically, the intelligent saliva aspirator: in normal mode, it removes waste liquid; in sampling mode, it uses a solenoid valve to switch the flow path. A built-in buffer chamber removes large food particles, and an ultrasonic defoaming module removes saliva bubbles.

[0033] Control process: Doctor starts sampling sequence → switches to sampling mode → flow rate sensor detects that the sampling volume meets the standard → automatically stops sampling and backflushing.

[0034] The flow control intelligent distribution system solves the problem of large fluctuations in saliva consistency, which can easily lead to air bubble blockage, and effectively distributes a limited sample volume to multiple testing items.

[0035] 1) Sample pretreatment: Separate the supernatant → monitor the pH in real time with an online pH electrode and add buffer solution → the viscosity sensor provides feedback on the dilution factor.

[0036] 2) Bubble detection and elimination: Ultrasonic detection and pulse pressure to remove bubbles.

[0037] 3) Sample allocation: Channel 1 - Electrochemical sensor → Channel 2 - Immunochromatography → Channel 3 - Microfluidic biochemical chip → Channel 4 - Nucleic acid extraction → Channel 5 - Sample retention chamber → Channel 6 - Raman sample pool.

[0038] Furthermore, refer to Figure 3 Step S105 includes the following steps: Step S201: Input the processed saliva sample into the microfluidic electrochemical sensor to obtain peak current, peak potential, transfer resistance and reaction kinetic constant; Step S202: Input the processed saliva sample into the SERS-Raman spectrometer to obtain the characteristic peak positions, peak intensity ratios, and full width at half maximum (FWHM). Step S203: Input the processed saliva sample into a microfluidic dry biochemical analyzer to obtain liver and kidney function indicators and blood lipid indicators. Step S204: Input the breath sample into the expiratory sensor to obtain the response time, maximum response amplitude, and recovery rate; Step S205: Input physiological data into the flexible wristband to obtain HRV, blood oxygen saturation, and blood pressure.

[0039] Furthermore, step S106 includes the following steps: Step S301: Based on the baseline drift caused by large fluctuations in saliva viscosity, an adaptive filter is used for noise processing to obtain the processed baseline drift. Step S302: Perform baseline subtraction algorithm processing on the CV curve to extract the Faraday current value of the redox peak.

[0040] Furthermore, step S106 includes the following steps: Step S401: Based on the fluorescence background interference in the Raman spectrum, an asymmetric least squares smoothing algorithm is used for baseline correction. Step S402: Principal component analysis is used to reduce dimensionality and extract the fingerprint features of bacterial cell walls.

[0041] Specifically, in electrochemical data processing: an adaptive filter is used to remove noise to address baseline drift caused by large fluctuations in saliva viscosity; a baseline subtraction algorithm is applied to the CV curve to accurately extract the Faraday current value of the redox peak.

[0042] Spectral data processing: Asymmetric least squares smoothing algorithm was used for baseline correction to address fluorescence background interference in Raman spectra; principal component analysis was used for dimensionality reduction to extract bacterial cell wall fingerprint features.

[0043] Time-series data alignment: Since the sampling frequencies of physiological signals (milliseconds) and biochemical tests (minutes) are different, the system adopts a sliding time window mechanism to divide a single diagnosis and treatment process into multiple stages and perform statistical downsampling on high-frequency data.

[0044] Furthermore, step S108 includes the following steps: Step S501: Process Raman spectra and electrochemical curves using 1D-CNN to extract local features; Step S502: Process PPG and respiratory time series data using LSTM to capture physiological change trends; Step S503: Automatically calculate the weights between different modalities, concatenate the feature vectors of each path, input them into the fully connected layer, and output the patient's current physiological state vector.

[0045] Specifically, a single-mode encoder: 1D-CNN was used to process Raman spectroscopy and electrochemical curves to extract local features; LSTM was used to process PPG and respiratory time-series data to capture physiological change trends.

[0046] Cross-modal attention layer: The system automatically calculates the weights between different modalities. When assessing different disease risks, the weights are automatically adjusted. The feature vectors from each pathway are concatenated and input into a fully connected layer to form a high-dimensional "patient's current physiological state vector."

[0047] Furthermore, the method also includes: Step S601: Set up warning traffic lights based on deviation scores and divide them into multiple warning levels.

[0048] Specifically, refer to Figure 4 Status correlation and clinical decision-making: The patient's current status is compared with the historical personal baseline, the deviation score is output, and the diagnostic confidence is improved through multi-source evidence chain.

[0049] Output: Outputs a radar chart, dynamically displaying health scores from different dimensions. It also establishes a warning traffic light system, dividing the system into three warning levels.

[0050] This application addresses the problems of complex diagnostic and treatment processes and long waiting times; it solves the problem of high-density integration of multiple precision analytical instruments within the limited space of a dental chair; and it solves the problem of synchronous fusion and clinical interpretation of multi-source heterogeneous data.

[0051] This application achieves integrated diagnosis and treatment with non-invasive detection: it highly integrates the functions of "sampling + detection + analysis" into the dental chair, and takes advantage of the window period of dental treatment to complete the collection and analysis of saliva, exhalation and physiological parameters without interfering with normal diagnosis and treatment. Constructing a multimodal biosensor network: combining electrochemical sensing, Raman spectroscopy, and microfluidics to achieve comprehensive coverage from local oral lesions to systemic metabolic status; Proactive AI-based early warning decision-making: Establish a health model based on "personal baseline" and use AI algorithms to identify subtle abnormalities, shifting from passive treatment to proactive health management and risk warning.

[0052] This application mainly aims to achieve: System Integration Architecture: A design scheme that physically and data-wise integrates microfluidic biochemical analysis, Raman spectroscopy detection, and dental integrated treatment unit.

[0053] Intelligent sampling path: In particular, a body fluid collection pretreatment system with dual modes of "saliva suction / sampling", online viscosity adjustment and bubble elimination functions.

[0054] Multimodal fusion algorithm: A cross-validation algorithm based on Raman spectral fingerprint (microbial information) and electrochemical sensing data (metabolic information) to construct a personal oral-systemic health association model.

[0055] In situ microbial drug susceptibility testing method: a method for rapidly assessing oral bacterial drug resistance at the treatment site by using a SERS-Raman module integrated in the dental chair in combination with heavy water labeling.

[0056] In addition, the sampling end in this application can use either MEMS sensors or cavity ring-down spectroscopy (CRDS) to detect trace gases in the breath test; the detection module can use either fluorescence immunoassay or electrochemiluminescence (ECL) or magnetic particle chemiluminescence modules to meet different sensitivity requirements; the detection device compartment can be designed as an independent trolley on the side of the dental chair, interconnected with the main control of the dental chair via wireless communication, rather than being completely embedded in the base, to adapt to the renovation of old clinics.

[0057] The beneficial effects achieved by this application are as follows: Non-invasive panoramic monitoring: No blood sampling is required. Comprehensive data covering biochemical, immune, microbiological and physiological signs can be obtained through saliva, breath and skin contact, which greatly improves the compliance of elderly patients.

[0058] High throughput and real-time performance: Compared to sending tests out (24-48 hours), this system can complete multiple tests simultaneously within 30-45 minutes of dental treatment, achieving "on-demand diagnosis, testing and treatment".

[0059] Systemic risk warning: By monitoring cardiovascular markers in saliva and combining them with real-time voltage fluctuations, it is possible to effectively prevent cardiovascular and cerebrovascular diseases that are common during dental surgery, thus ensuring the safety of medical treatment for the elderly.

[0060] Example 2: Figure 5 This is a schematic diagram of the multimodal intelligent dental treatment system provided in Embodiment 2 of the present invention.

[0061] Reference Figure 5 The system includes: The intelligent saliva aspirator is used to pre-process saliva samples that meet the required volume, detect and eliminate air bubbles, and then distribute the processed saliva samples to multiple channels. Breath sampling headrest, used for collecting breath samples; Flexible wearable units for collecting physiological data; The AI ​​data processing module is used to process saliva samples, breath samples, and physiological data as multidimensional heterogeneous data; input the multidimensional heterogeneous data into the corresponding devices to obtain electrochemical data and spectral data; preprocess and extract features from the electrochemical and spectral data to obtain feature data; use a sliding time window mechanism to divide a single diagnosis and treatment process into multiple stages, and perform statistical downsampling on high-frequency data; use a multimodal fusion engine to process the feature data, and compare the patient's current status with historical personal baselines to output deviation scores and radar charts.

[0062] This invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the multimodal intelligent dental treatment method provided in the above embodiments.

[0063] This invention also provides a computer-readable medium having processor-executable non-volatile program code, on which a computer program is stored, and which, when run by a processor, performs the steps of the multimodal intelligent dental treatment method described above.

[0064] The computer program product provided in this embodiment of the invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0065] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and apparatus described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0066] Furthermore, in the description of the embodiments of the present invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.

[0067] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0068] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0069] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A multimodal intelligent dental diagnosis and treatment method, characterized in that, The method includes: The intelligent saliva aspirator preprocesses, detects and eliminates air bubbles in saliva samples that meet the required volume, and then obtains the processed saliva sample. The processed saliva sample is distributed into multiple channels; Breath sampling headrest collects respiratory samples, while flexible wearable unit collects physiological data; The processed saliva sample, the respiratory sample, and the physiological data are used as multidimensional heterogeneous data. The multidimensional heterogeneous data is input into the corresponding device to obtain electrochemical data and spectral data; The electrochemical data and the spectral data are preprocessed and feature extracted to obtain feature data; The feature data is used with a sliding time window mechanism to divide a single diagnosis and treatment process into multiple stages, and high-frequency data is statistically downsampled. The feature data is processed by a multimodal fusion engine, and the patient's current status is compared with their historical personal baseline to output a deviation score and a radar chart.

2. The multimodal intelligent dental treatment method according to claim 1, characterized in that, The multiple channels include a first channel, a second channel, a third channel, a fourth channel, a fifth channel, and a sixth channel; the first channel is an electrochemical sensor, the second channel is an immunochromatographic assay, the third channel is a microfluidic biochemical chip, the fourth channel is a nucleic acid extraction device, the fifth channel is a sample retention chamber, and the sixth channel is a Raman sample pool.

3. The multimodal intelligent dental treatment method according to claim 1, characterized in that, The multidimensional heterogeneous data is input into the corresponding device to obtain electrochemical data and spectroscopic data, including: The processed saliva sample was input into a microfluidic electrochemical sensor to obtain peak current, peak potential, transfer resistance, and reaction kinetic constants. The processed saliva sample was input into a SERS-Raman spectrometer to obtain the characteristic peak positions, peak intensity ratios, and full width at half maximum (FWHM). The processed saliva sample was input into a microfluidic dry biochemical analyzer to obtain liver and kidney function indicators and blood lipid indicators. The breath sample is input into the expiratory sensor to obtain the response time, maximum response amplitude, and recovery rate. The physiological data is input into the flexible wristband to obtain HRV, blood oxygen saturation, and blood pressure.

4. The multimodal intelligent dental treatment method according to claim 1, characterized in that, The electrochemical data and the spectral data are preprocessed and feature extracted to obtain feature data, including: Based on the baseline drift caused by large fluctuations in saliva viscosity, an adaptive filter is used for noise processing to obtain the processed baseline drift. The baseline subtraction algorithm was applied to the CV curve to extract the Faraday current value of the redox peak.

5. The multimodal intelligent dental treatment method according to claim 1, characterized in that, The electrochemical data and the spectral data are preprocessed and feature extracted to obtain feature data, including: Based on the fluorescence background interference in Raman spectroscopy, an asymmetric least squares smoothing algorithm is used for baseline correction. Principal component analysis was used for dimensionality reduction to extract the fingerprint features of bacterial cell walls.

6. The multimodal intelligent dental treatment method according to claim 1, characterized in that, The feature data is processed by a multimodal fusion engine, including: Local features were extracted by processing Raman spectra and electrochemical curves using 1D-CNN; By processing PPG and respiratory time-series data using LSTM, physiological change trends can be captured. The system automatically calculates the weights between different modalities, concatenates the feature vectors from each mode, and inputs them into a fully connected layer to output the patient's current physiological state vector.

7. The multimodal intelligent dental treatment method according to claim 1, characterized in that, The method further includes: Based on the deviation score, a warning traffic light system is set up, and multiple warning levels are divided.

8. A multimodal intelligent dental treatment system, characterized in that, The system includes: A smart saliva aspirator is used to preprocess, detect and eliminate air bubbles in saliva samples that have reached the required volume, and then distribute the processed saliva samples into multiple channels. Breath sampling headrest, used for collecting breath samples; Flexible wearable units for collecting physiological data; The AI ​​data processing module is used to treat the processed saliva sample, the respiratory sample, and the physiological data as multidimensional heterogeneous data; input the multidimensional heterogeneous data into the corresponding devices to obtain electrochemical data and spectral data; preprocess and extract features from the electrochemical data and the spectral data to obtain feature data; use a sliding time window mechanism to divide a single diagnosis and treatment process into multiple stages, and perform statistical downsampling on high-frequency data; use a multimodal fusion engine to process the feature data, and compare the patient's current status with historical personal baselines to output deviation scores and radar charts.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the method described in any one of claims 1 to 7.

10. A computer-readable medium having processor-executable non-volatile program code, characterized in that, The program code causes the processor to execute the method described in any one of claims 1 to 7.

Images