Anti-caries early warning method and system based on oral health detection

Abstract

The invention discloses an anti-caries early warning method and system based on oral health detection, and belongs to the technical field of data processing. The method comprises the following steps: acquiring and processing target detection data to generate a clear enamel demineralization feature image and a buffer capability score; analyzing pixel-level fluorescence intensity distribution by adopting a convolutional neural network model, determining a quantized value of the enamel demineralization degree, and further determining a comprehensive dental caries risk level category; judging a risk change trend according to the historical target detection data sequence to obtain a dynamic evaluation update value; a data fusion algorithm is adopted to obtain a final decayed tooth risk assessment index, and risk early warning information is generated according to the early warning template; by integrating fluorescence imaging, an image enhancement algorithm, a convolutional neural network, spectral analysis, k-means clustering and time sequence analysis, a complete process from enamel demineralization feature extraction to comprehensive risk assessment is constructed, the accuracy and individuation level of decayed tooth risk assessment are remarkably improved, and a scientific basis is provided for early prevention and intervention.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method and system for early warning of dental caries based on oral health detection. Background Technology

[0002] Oral health is a crucial component of overall health, directly impacting people's quality of life and overall well-being. Dental caries, the most common oral disease, not only causes tooth pain and functional loss but can also trigger systemic health problems such as cardiovascular disease. Therefore, developing a precise and efficient oral health detection system is essential for the early detection and prevention of dental caries. Current detection technologies often face limitations in identifying caries risk, requiring breakthroughs to improve preventative effectiveness. Existing methods typically rely on traditional dental examinations, such as X-ray imaging or visual examination, which are insufficient for detecting early caries. X-ray imaging has low sensitivity to early enamel demineralization, making it difficult to capture minute lesions, while visual examination heavily relies on the dentist's experience, easily leading to missed or misdiagnosed cases due to subjective judgment. Furthermore, existing methods often fail to comprehensively assess a patient's oral environment and lifestyle habits, resulting in a lack of personalized risk assessments and hindering effective guidance for preventative measures. Technically, the core challenge of detection systems lies in accurately quantifying key indicators within the oral cavity. The degree of enamel demineralization, as an early signal of caries formation, is difficult to accurately capture through conventional methods due to its subtle changes. While fluorescence detection technology shows promise in identifying demineralization lesions, its adaptability to the complex oral environment is insufficient. For example, interference from saliva flow or uneven tooth surface can affect detection accuracy. Furthermore, the results of enamel demineralization assessment need to be analyzed in conjunction with dynamic factors such as the patient's salivary buffering capacity and dietary habits. Integrating this multi-source data into a unified assessment system has become another key technical bottleneck. There is a close relationship between the two: accurate demineralization detection is fundamental, but only by combining it with personalized oral environment data can a comprehensive risk assessment be achieved.

[0003] Therefore, how to accurately detect the degree of enamel demineralization in a complex oral environment and effectively integrate the test results with patients' personalized oral physiology and lifestyle data to form a dynamic and accurate caries risk assessment has become a key issue that this study urgently needs to address. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for early warning of dental caries based on oral health detection, so as to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] A caries prevention and early warning method based on oral health testing, comprising the following steps:

[0007] Step S100: Collect and process target detection data to generate clear enamel demineralization feature images and buffering capacity scores. The target detection data includes tooth surface image data and saliva sample data.

[0008] Step S200: Based on the clear image of enamel demineralization features, a convolutional neural network model is used to analyze the pixel-level fluorescence intensity distribution and determine the quantitative value of the degree of enamel demineralization.

[0009] Step S300: Determine the overall caries risk level category based on the quantitative value of enamel demineralization and the buffering capacity score;

[0010] Step S400: Based on the historical target detection data sequence and the comprehensive caries risk level category, determine the risk change trend to obtain a dynamic assessment update value;

[0011] Step S500: Use a data fusion algorithm to integrate the quantitative value of enamel demineralization, buffer capacity score and dynamic assessment update value to obtain the final caries risk assessment index, and generate risk warning information based on the warning template.

[0012] As a preferred embodiment of the oral health detection-based caries prevention and early warning method and system of the present invention, the specific implementation process of collecting and processing target detection data to generate clear enamel demineralization feature images and buffering capacity scores includes:

[0013] Collect target detection data, which includes tooth surface image data and saliva sample data; collect target tooth surface image data using a fluorescence imaging device; and acquire target saliva sample data using a saliva collection sensor.

[0014] The target tooth surface image data is processed using an image enhancement algorithm to remove saliva interference and surface inhomogeneity, resulting in a clear image of enamel demineralization features.

[0015] The wavelength intensity values ​​in the real-time saliva sample data were extracted using spectral analysis methods.

[0016] The buffering capacity score is calculated based on the wavelength intensity value and the corresponding buffering capacity contribution coefficient.

[0017] As a preferred embodiment of the oral health detection-based caries prevention and early warning method and system of the present invention, the specific implementation process of determining the quantitative value of enamel demineralization degree by analyzing pixel-level fluorescence intensity distribution using a convolutional neural network model based on clear enamel demineralization feature images includes:

[0018] The clear enamel demineralization feature image is input into a pre-trained convolutional neural network model;

[0019] The fluorescence intensity value of each pixel in the clear enamel demineralization feature image is extracted using the convolutional neural network model.

[0020] Based on the fluorescence intensity value and the corresponding weighting coefficient, the pixel-level fluorescence intensity distribution characteristics of the clear enamel demineralization feature image are determined, and the quantification value of the degree of enamel demineralization is calculated. The calculation formula is as follows:

[0021] ;

[0022] in, This indicates a quantitative value representing the degree of demineralization of tooth enamel. and These represent the number of pixel rows and columns in a clear image of enamel demineralization features, respectively. This represents the weight coefficient of the convolutional neural network at position (i, j). This represents the fluorescence intensity value at position (i, j). This represents the importance weight factor at position (i, j).

[0023] As a preferred embodiment of the caries prevention and early warning method and system based on oral health testing of the present invention, the specific implementation process of determining the comprehensive caries risk level category based on the quantitative value of enamel demineralization and buffer capacity score includes:

[0024] The quantitative value of the degree of enamel demineralization and the buffering capacity score are input into a preset k-means clustering model;

[0025] The risk of dental caries is classified using a k-means clustering model, and the overall dental caries risk level category is output.

[0026] As a preferred embodiment of the oral health detection-based caries prevention and early warning method and system of the present invention, the specific implementation process of judging the risk change trend based on historical target detection data sequence and comprehensive caries risk level category to obtain dynamic assessment update value includes:

[0027] Obtain historical target detection data sequences and comprehensive caries risk level categories;

[0028] Risk change characteristics in the historical target detection data sequence are extracted using time series analysis methods;

[0029] Based on the aforementioned risk change characteristics and the aforementioned comprehensive caries risk level category, the formulas for calculating the time weighting coefficient and risk volatility coefficient are as follows:

[0030] ;

[0031] ;

[0032] in, Indicates the first Time weighting coefficients at each time point This represents the total length of the historical target detection data sequence. Indicates the time decay factor. Represents a point-in-time variable; Indicates the first Risk volatility coefficient at a given point in time, This represents the size of a sliding window. Indicates the first The numerical score corresponding to the comprehensive caries risk level category at each time point. This represents the average numerical score corresponding to the overall caries risk category within a sliding window;

[0033] The dynamic assessment update value is determined based on the time weighting coefficient and the risk volatility coefficient, using the following formula:

[0034] ;

[0035] in, This indicates a dynamically evaluated and updated value. Indicates the first Risk volatility coefficient at a given point in time, This indicates the amount of change in risk level. Indicates a time interval.

[0036] As a preferred embodiment of the caries prevention and early warning method and system based on oral health testing of the present invention, the specific implementation process of integrating the quantitative value of enamel demineralization degree, buffer capacity score and dynamic assessment update value using a data fusion algorithm to obtain the final caries risk assessment index, and generating risk warning information according to the early warning template includes:

[0037] The data fusion algorithm is used to weight and fuse the quantitative value of enamel demineralization, the buffering capacity score, and the dynamic assessment update value to obtain the final caries risk assessment value.

[0038] Risk warning information is generated based on a warning template, which includes warning generation rules for generating risk warning information based on the final caries risk assessment value.

[0039] A caries prevention and early warning system based on oral health detection, the system includes a data acquisition and preprocessing module, a tooth enamel demineralization quantification module, an identification module, an update module, and a risk assessment module;

[0040] The acquisition and preprocessing module is used to acquire and process target detection data to generate clear images of enamel demineralization features and buffering capacity scores. The target detection data includes tooth surface image data and saliva sample data.

[0041] The enamel demineralization quantification module is used to analyze the pixel-level fluorescence intensity distribution based on a clear enamel demineralization feature image and to determine the quantification value of the degree of enamel demineralization.

[0042] The identification module is used to determine the overall caries risk level category based on the quantitative value of enamel demineralization and the buffering capacity score, using the k-means clustering algorithm.

[0043] The update module is used to determine the risk change trend based on the historical target detection data sequence and the comprehensive caries risk level category to obtain a dynamic assessment update value;

[0044] The risk assessment module is used to obtain the final caries risk assessment index and generate risk warning information based on the warning template.

[0045] As a preferred embodiment of the oral health detection-based caries prevention and early warning method and system of the present invention, the acquisition and preprocessing module includes an acquisition unit and a preprocessing unit;

[0046] The acquisition unit is used to acquire target detection data, which includes tooth surface image data and saliva sample data; acquires target tooth surface image data through a fluorescence imaging device; and acquires target saliva sample data through a saliva acquisition sensor.

[0047] The preprocessing unit is used to process the target detection data and generate clear images of enamel demineralization features and a buffering capacity score.

[0048] The target tooth surface image data is processed using an image enhancement algorithm to remove saliva interference and surface inhomogeneity, resulting in a clear image of enamel demineralization features.

[0049] The spectral intensity values ​​of the real-time saliva sample data were extracted using spectral analysis methods.

[0050] Based on the wavelength intensity value and the corresponding buffering capacity contribution coefficient, the buffering capacity score is calculated using the following formula:

[0051] ;

[0052] in, Indicates the buffering capacity score. This indicates the number of wavelength sampling points in spectral analysis. Indicates the first Spectral intensity values ​​at each wavelength point Indicates the first The buffer capacity contribution coefficient corresponding to each wavelength point This represents the normalization factor for spectral analysis.

[0053] As a preferred embodiment of the oral health detection-based caries prevention and early warning method and system of the present invention, the update module includes an analysis unit and a dynamic update unit;

[0054] The analysis unit is used to obtain risk change characteristics;

[0055] Obtain historical target detection data sequences and comprehensive caries risk level categories;

[0056] Risk change characteristics in the historical target detection data sequence are extracted using time series analysis methods;

[0057] The dynamic update unit is used to determine the dynamic evaluation update value;

[0058] Based on the risk change characteristics and the comprehensive caries risk level category, the time weighting coefficient and risk volatility coefficient are calculated using the following formula:

[0059] ;

[0060] ;

[0061] in, Indicates the first Time weighting coefficients at each time point This represents the total length of the historical target detection data sequence. Indicates the time decay factor. Represents a point-in-time variable; Indicates the first Risk volatility coefficient at a given point in time, This represents the size of a sliding window. Indicates the first The numerical score corresponding to the comprehensive caries risk level category at each time point. This represents the average numerical score corresponding to the overall caries risk category within a sliding window;

[0062] The dynamic assessment update value is determined based on the time weighting coefficient and the risk volatility coefficient, using the following formula:

[0063] ;

[0064] in, This indicates a dynamically evaluated and updated value. Indicates the first Risk volatility coefficient at a given point in time, This indicates the amount of change in risk level. Indicates a time interval.

[0065] As a preferred embodiment of the oral health detection-based caries prevention and early warning method and system of the present invention, the risk assessment module includes a fusion unit and an early warning unit:

[0066] The fusion unit is used to perform weighted fusion of the quantitative value of enamel demineralization, the buffer capacity score and the dynamic assessment update value through the data fusion algorithm to obtain the final caries risk assessment value.

[0067] The early warning unit is used to generate risk warning information based on an early warning template, wherein the early warning template includes early warning generation rules for generating risk warning information based on the final caries risk assessment value.

[0068] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: This invention provides a caries prevention and early warning method and system based on oral health detection. It collects and processes target detection data to generate clear images of enamel demineralization features and a buffering capacity score; it uses a convolutional neural network model to analyze pixel-level fluorescence intensity distribution to determine the quantitative value of enamel demineralization; based on the quantitative value of enamel demineralization and the buffering capacity score, it determines the comprehensive caries risk level category; based on historical target detection data sequences and the comprehensive caries risk level category, it judges the risk change trend to obtain a dynamic assessment update value; it uses a data fusion algorithm to obtain the final caries risk assessment index and generates risk warning information based on the warning template; this invention integrates fluorescence imaging, image enhancement algorithms, convolutional neural networks, spectral analysis, k-means clustering, and time series analysis to construct a complete process from enamel demineralization feature extraction to comprehensive risk assessment, significantly improving the accuracy and personalization of caries risk assessment and providing a scientific basis for early prevention and intervention. Attached Figure Description

[0069] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0070] Figure 1 This is a schematic diagram of a caries prevention and early warning system based on oral health detection according to the present invention; Figure 2 This is a schematic diagram illustrating the steps of a caries prevention and early warning method based on oral health testing according to the present invention. Detailed Implementation

[0071] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0072] Please see Figure 1 In this first embodiment: a caries prevention and early warning system based on oral health detection is provided. The system includes a data acquisition and preprocessing module, a tooth enamel demineralization quantification module, an identification module, an update module, and a risk assessment module.

[0073] The acquisition and preprocessing module is used to acquire and process target detection data to generate clear images of enamel demineralization features and buffering capacity scores. The target detection data includes tooth surface image data and saliva sample data.

[0074] The enamel demineralization quantification module is used to analyze the pixel-level fluorescence intensity distribution based on a clear enamel demineralization feature image and to determine the quantification value of the degree of enamel demineralization.

[0075] The identification module is used to determine the overall caries risk level category based on the quantitative value of enamel demineralization and the buffering capacity score, using the k-means clustering algorithm.

[0076] The update module is used to determine the risk change trend based on the historical target detection data sequence and the comprehensive caries risk level category to obtain a dynamic assessment update value;

[0077] The risk assessment module is used to obtain the final caries risk assessment index and generate risk warning information based on the warning template.

[0078] Specifically, the acquisition and preprocessing module includes an acquisition unit and a preprocessing unit;

[0079] The acquisition unit is used to acquire target detection data, which includes tooth surface image data and saliva sample data; acquires target tooth surface image data through a fluorescence imaging device; and acquires target saliva sample data through a saliva acquisition sensor.

[0080] The preprocessing unit is used to process the target detection data and generate clear images of enamel demineralization features and a buffering capacity score.

[0081] The target tooth surface image data is processed using an image enhancement algorithm to remove saliva interference and surface inhomogeneity, resulting in a clear image of enamel demineralization features.

[0082] The spectral intensity values ​​of the real-time saliva sample data were extracted using spectral analysis methods.

[0083] Based on the wavelength intensity value and the corresponding buffering capacity contribution coefficient, the buffering capacity score is calculated using the following formula:

[0084] ;

[0085] in, Indicates the buffering capacity score. This indicates the number of wavelength sampling points in spectral analysis. Indicates the first Spectral intensity values ​​at each wavelength point Indicates the first The buffer capacity contribution coefficient corresponding to each wavelength point This represents the normalization factor for spectral analysis.

[0086] Specifically, the update module includes an analysis unit and a dynamic update unit;

[0087] The analysis unit is used to obtain risk change characteristics;

[0088] Obtain historical target detection data sequences and comprehensive caries risk level categories;

[0089] Risk change characteristics in the historical target detection data sequence are extracted using time series analysis methods;

[0090] The dynamic update unit is used to determine the dynamic evaluation update value;

[0091] Based on the risk change characteristics and the comprehensive caries risk level category, the time weighting coefficient and risk volatility coefficient are calculated using the following formula:

[0092] ;

[0093] ;

[0094] in, Indicates the first Time weighting coefficients at each time point This represents the total length of the historical target detection data sequence. Indicates the time decay factor. Represents a point-in-time variable; Indicates the first Risk volatility coefficient at a given point in time, This represents the size of a sliding window. Indicates the first The numerical score corresponding to the comprehensive caries risk level category at each time point. This represents the average numerical score corresponding to the overall caries risk category within a sliding window;

[0095] The dynamic assessment update value is determined based on the time weighting coefficient and the risk volatility coefficient, using the following formula:

[0096] ;

[0097] in, This indicates a dynamically evaluated and updated value. Indicates the first Risk volatility coefficient at a given point in time, This indicates the amount of change in risk level. Indicates a time interval.

[0098] Specifically, the risk assessment module includes a fusion unit and an early warning unit:

[0099] The fusion unit is used to perform weighted fusion of the quantitative value of enamel demineralization, the buffer capacity score and the dynamic assessment update value through the data fusion algorithm to obtain the final caries risk assessment value.

[0100] The early warning unit is used to generate risk warning information based on an early warning template, wherein the early warning template includes early warning generation rules for generating risk warning information based on the final caries risk assessment value.

[0101] Please see Figure 2 In this second embodiment: a caries prevention and early warning method based on oral health detection is provided, which includes the following steps:

[0102] Step S100: Collect and process target detection data to generate clear enamel demineralization feature images and buffering capacity scores. The target detection data includes tooth surface image data and saliva sample data.

[0103] Step S200: Based on the clear image of enamel demineralization features, a convolutional neural network model is used to analyze the pixel-level fluorescence intensity distribution and determine the quantitative value of the degree of enamel demineralization.

[0104] Step S300: Determine the overall caries risk level category based on the quantitative value of enamel demineralization and the buffering capacity score;

[0105] Step S400: Based on the historical target detection data sequence and the comprehensive caries risk level category, determine the risk change trend to obtain a dynamic assessment update value;

[0106] Step S500: Use a data fusion algorithm to integrate the quantitative value of enamel demineralization, buffer capacity score and dynamic assessment update value to obtain the final caries risk assessment index, and generate risk warning information based on the warning template.

[0107] Specifically, the specific implementation process of collecting and processing target detection data to generate clear enamel demineralization feature images and buffering capacity scores includes:

[0108] Collect target detection data, which includes tooth surface image data and saliva sample data; collect target tooth surface image data using a fluorescence imaging device; and acquire target saliva sample data using a saliva collection sensor.

[0109] The target tooth surface image data is processed using an image enhancement algorithm to remove saliva interference and surface inhomogeneity, resulting in a clear image of enamel demineralization features.

[0110] The wavelength intensity values ​​in the real-time saliva sample data were extracted using spectral analysis methods.

[0111] The buffering capacity score is calculated based on the wavelength intensity value and the corresponding buffering capacity contribution coefficient.

[0112] Specifically, the process of determining the quantitative value of enamel demineralization degree by analyzing pixel-level fluorescence intensity distribution using a convolutional neural network model based on clear enamel demineralization feature images includes:

[0113] The clear enamel demineralization feature image is input into a pre-trained convolutional neural network model;

[0114] The fluorescence intensity value of each pixel in the clear enamel demineralization feature image is extracted using the convolutional neural network model.

[0115] Based on the fluorescence intensity value and the corresponding weighting coefficient, the pixel-level fluorescence intensity distribution characteristics of the clear enamel demineralization feature image are determined, and the quantification value of the degree of enamel demineralization is calculated. The calculation formula is as follows:

[0116] ;

[0117] in, This indicates a quantitative value representing the degree of demineralization of tooth enamel. and These represent the number of pixel rows and columns in a clear image of enamel demineralization features, respectively. This represents the weight coefficient of the convolutional neural network at position (i, j). This represents the fluorescence intensity value at position (i, j). This represents the importance weight factor at position (i, j).

[0118] Specifically, the implementation process for determining the comprehensive caries risk level category based on the quantitative value of enamel demineralization and the buffering capacity score includes:

[0119] The quantitative value of the degree of enamel demineralization and the buffering capacity score are input into a preset k-means clustering model;

[0120] The risk of dental caries is classified using a k-means clustering model, and the overall dental caries risk level category is output.

[0121] Specifically, the implementation process of determining the risk change trend based on historical target detection data sequences and comprehensive caries risk level categories to obtain dynamic assessment update values ​​includes:

[0122] Obtain historical target detection data sequences and comprehensive caries risk level categories;

[0123] Risk change characteristics in the historical target detection data sequence are extracted using time series analysis methods;

[0124] Based on the risk change characteristics and the comprehensive caries risk level category, the time weighting coefficient and risk volatility coefficient are calculated using the following formula:

[0125] ;

[0126] ;

[0127] in, Indicates the first Time weighting coefficients at each time point This represents the total length of the historical target detection data sequence. Indicates the time decay factor. Represents a point-in-time variable; Indicates the first Risk volatility coefficient at a given point in time, This represents the size of a sliding window. Indicates the first The numerical score corresponding to the comprehensive caries risk level category at each time point. This represents the average numerical score corresponding to the overall caries risk category within a sliding window;

[0128] The dynamic assessment update value is determined based on the time weighting coefficient and the risk volatility coefficient, using the following formula:

[0129] ;

[0130] in, This indicates a dynamically evaluated and updated value. Indicates the first Risk volatility coefficient at a given point in time, This indicates the amount of change in risk level. Indicates a time interval.

[0131] Specifically, the implementation process of integrating quantitative values ​​of enamel demineralization, buffer capacity scores, and dynamically updated assessment values ​​using a data fusion algorithm to obtain a final caries risk assessment index, and generating risk warning information based on a warning template, includes:

[0132] The degree of enamel demineralization is quantified using the data fusion algorithm. The buffering capacity score and the dynamically evaluated update value The final caries risk assessment value was obtained by weighted fusion. The fusion formula is:

[0133] ;

[0134] in, , , Indicates weight, ;

[0135] Risk warning information is generated based on a warning template, which includes warning generation rules for generating risk warning information based on the final caries risk assessment value.

[0136] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0137] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for early warning of dental caries based on oral health testing, characterized in that, The method includes the following steps: Collect and process target detection data to generate clear images of enamel demineralization features and buffering capacity scores. The target detection data includes tooth surface image data and saliva sample data. Based on clear images of enamel demineralization features, a convolutional neural network model was used to analyze pixel-level fluorescence intensity distribution and determine the quantitative value of the degree of enamel demineralization. The overall caries risk level category is determined based on the quantitative value of enamel demineralization and the buffering capacity score; Based on historical target detection data sequences and comprehensive caries risk level categories, the risk change trend is determined to obtain dynamic assessment and update values; A data fusion algorithm is used to integrate the quantitative value of enamel demineralization, buffer capacity score and dynamic assessment update value to obtain the final caries risk assessment index, and risk warning information is generated based on the warning template.

2. The method for caries prevention and early warning based on oral health detection according to claim 1, characterized in that, The specific implementation process of collecting and processing target detection data to generate clear enamel demineralization feature images and buffering capacity scores includes: Collect target detection data, which includes tooth surface image data and saliva sample data; collect target tooth surface image data using a fluorescence imaging device; and acquire target saliva sample data using a saliva collection sensor. The target tooth surface image data is processed using an image enhancement algorithm to remove saliva interference and surface inhomogeneity, resulting in a clear image of enamel demineralization features. The wavelength intensity values ​​in the real-time saliva sample data were extracted using spectral analysis methods. The buffering capacity score is calculated based on the wavelength intensity value and the corresponding buffering capacity contribution coefficient.

3. The method for caries prevention and early warning based on oral health detection according to claim 1, characterized in that, The specific implementation process of determining the quantitative value of enamel demineralization degree by analyzing pixel-level fluorescence intensity distribution using a convolutional neural network model based on clear enamel demineralization feature images includes: The clear enamel demineralization feature image is input into a pre-trained convolutional neural network model; The fluorescence intensity value of each pixel in the clear enamel demineralization feature image is extracted using the convolutional neural network model. Based on the fluorescence intensity value and the corresponding weighting coefficient, the pixel-level fluorescence intensity distribution characteristics of the clear enamel demineralization feature image are determined, and the quantification value of the degree of enamel demineralization is calculated. The calculation formula is as follows: ； in, This indicates a quantitative value representing the degree of demineralization of tooth enamel. and These represent the number of pixel rows and columns in a clear image of enamel demineralization features, respectively. This represents the weight coefficient of the convolutional neural network at position (i, j). This represents the fluorescence intensity value at position (i, j). This represents the importance weight factor at position (i, j).

4. The method for caries prevention and early warning based on oral health detection according to claim 1, characterized in that, The specific implementation process for determining the comprehensive caries risk level category based on the quantitative value of enamel demineralization and the buffering capacity score includes: The quantitative value of the degree of enamel demineralization and the buffering capacity score are input into a preset k-means clustering model; The risk of dental caries is classified using a k-means clustering model, and the overall dental caries risk level category is output.

5. A method for early warning of dental caries based on oral health detection according to claim 1, characterized in that, The specific implementation process for determining the risk change trend and obtaining a dynamic assessment update value based on historical target detection data sequences and comprehensive caries risk level categories includes: Obtain historical target detection data sequences and comprehensive caries risk level categories; Risk change characteristics in the historical target detection data sequence are extracted using time series analysis methods; Calculate the time weighting coefficient based on the risk change characteristics and the comprehensive caries risk level category. And the risk volatility coefficient, the calculation formula is: ； ； in, Indicates the first Time weighting coefficients at each time point This represents the total length of the historical target detection data sequence. Indicates the time decay factor. Represents a point-in-time variable; Indicates the first Risk volatility coefficient at a given point in time, This represents the size of a sliding window. Indicates the first The numerical score corresponding to the comprehensive caries risk level category at each time point. This represents the average numerical score corresponding to the overall caries risk category within a sliding window; The dynamic assessment update value is determined based on the time weighting coefficient and the risk volatility coefficient, using the following formula: ； in, This indicates a dynamically evaluated and updated value. Indicates the first Risk volatility coefficient at a given point in time, This indicates the amount of change in risk level. Indicates a time interval.

6. The method for caries prevention and early warning based on oral health detection according to claim 1, characterized in that, The specific implementation process of integrating quantitative values ​​of enamel demineralization, buffer capacity scores, and dynamically updated assessment values ​​using a data fusion algorithm to obtain a final caries risk assessment index, and generating risk warning information based on a warning template, includes: The data fusion algorithm is used to weight and fuse the quantitative value of enamel demineralization, the buffering capacity score, and the dynamic assessment update value to obtain the final caries risk assessment value. Risk warning information is generated based on a warning template, which includes warning generation rules for generating risk warning information based on the final caries risk assessment value.

7. A caries prevention and early warning system based on oral health detection, characterized in that, The system includes a data acquisition and preprocessing module, a tooth enamel demineralization quantification module, an identification module, an update module, and a risk assessment module. The acquisition and preprocessing module is used to acquire and process target detection data to generate clear images of enamel demineralization features and buffering capacity scores. The target detection data includes tooth surface image data and saliva sample data. The enamel demineralization quantification module is used to analyze the pixel-level fluorescence intensity distribution based on a clear enamel demineralization feature image and to determine the quantification value of the degree of enamel demineralization. The identification module is used to determine the overall caries risk level category based on the quantitative value of enamel demineralization and the buffering capacity score, using the k-means clustering algorithm. The update module is used to determine the risk change trend based on the historical target detection data sequence and the comprehensive caries risk level category to obtain a dynamic assessment update value; The risk assessment module is used to obtain the final caries risk assessment index and generate risk warning information based on the warning template.

8. A caries prevention and early warning system based on oral health detection according to claim 7, characterized in that: The acquisition and preprocessing module includes an acquisition unit and a preprocessing unit; The acquisition unit is used to acquire target detection data, which includes tooth surface image data and saliva sample data; Image data of the target tooth surface is acquired using a fluorescence imaging device; saliva sample data of the target is acquired using a saliva collection sensor. The preprocessing unit is used to process the target detection data and generate clear images of enamel demineralization features and a buffering capacity score. The target tooth surface image data is processed using an image enhancement algorithm to remove saliva interference and surface inhomogeneity, resulting in a clear image of enamel demineralization features. The spectral intensity values ​​of the real-time saliva sample data were extracted using spectral analysis methods. Based on the wavelength intensity value and the corresponding buffering capacity contribution coefficient, the buffering capacity score is calculated using the following formula: ； in, Indicates the buffering capacity score. This indicates the number of wavelength sampling points in spectral analysis. Indicates the first Spectral intensity values ​​at each wavelength point Indicates the first The buffer capacity contribution coefficient corresponding to each wavelength point This represents the normalization factor for spectral analysis.

9. A caries prevention and early warning system based on oral health detection according to claim 7, characterized in that: The update module includes an analysis unit and a dynamic update unit; The analysis unit is used to obtain risk change characteristics; Obtain historical target detection data sequences and comprehensive caries risk level categories; Risk change characteristics in the historical target detection data sequence are extracted using time series analysis methods; The dynamic update unit is used to determine the dynamic evaluation update value; Based on the risk change characteristics and the comprehensive caries risk level category, the time weighting coefficient and risk volatility coefficient are calculated using the following formula: ； ； in, Indicates the first Time weighting coefficients at each time point This represents the total length of the historical target detection data sequence. Indicates the time decay factor. Represents a point-in-time variable; Indicates the first Risk volatility coefficient at a given point in time, This represents the size of a sliding window. Indicates the first The numerical score corresponding to the comprehensive caries risk level category at each time point. This represents the average numerical score corresponding to the overall caries risk category within a sliding window; The dynamic assessment update value is determined based on the time weighting coefficient and the risk volatility coefficient, using the following formula: ； in, This indicates a dynamically evaluated and updated value. Indicates the first Risk volatility coefficient at a given point in time, This indicates the amount of change in risk level. Indicates a time interval.

10. A caries prevention and early warning system based on oral health detection according to claim 7, characterized in that: The risk assessment module includes a fusion unit and an early warning unit: The fusion unit is used to perform weighted fusion of the quantitative value of enamel demineralization, the buffer capacity score and the dynamic assessment update value through the data fusion algorithm to obtain the final caries risk assessment value. The early warning unit is used to generate risk warning information based on an early warning template, wherein the early warning template includes early warning generation rules for generating risk warning information based on the final caries risk assessment value.

Images