Oral health monitoring system based on laser irradiation and data processing method thereof

Abstract

The present application relates to the technical field of spectral data processing, in particular to a kind of oral health monitoring system based on laser irradiation and its data processing method, comprising: obtaining the first spectral data of tooth in oral cavity area and the second spectral data of saliva;According to the feature distribution of each oral cavity area in each wave band spectral data each dimension and the data deviation under healthy state, obtain spectral residual contrast;According to the difference distribution of spectral residual contrast between the first spectral data and the second spectral data, in combination with the feature correlation of the first spectral data and the second spectral data, obtain collaborative change index;According to the fluctuation of collaborative change index, in combination with the spectral residual contrast of the first spectral data, obtain abnormal attention degree;Based on abnormal attention degree, the tooth health condition of each oral cavity area is monitored.The present application improves the recognition accuracy of oral health status, and provides auxiliary diagnosis result with better diagnosis effect for medical staff.

Description

Technical Field

[0001] This invention relates to the field of spectral data processing technology, specifically to an oral health monitoring system based on laser irradiation and its data processing method. Background Technology

[0002] Oral health not only affects an individual's chewing, pronunciation, and aesthetic appearance, but is also closely related to overall health. Traditional oral examination methods mainly rely on clinical observation and imaging techniques such as X-rays, which are invasive, highly susceptible to the subjective influence of the operator, or carry radiation risks. Laser and related spectral technologies have rapidly developed in medical testing in recent years, offering advantages such as high sensitivity, real-time performance, non-invasiveness, and portability. Utilizing laser and spectral data analysis technologies, rapid, objective, and non-invasive detection of pathological information in oral hard tissues (such as teeth) and soft tissue environments (such as saliva) can be achieved, thereby enabling early diagnosis, risk assessment, and personalized health management.

[0003] In existing technologies, laser-based oral health monitoring primarily utilizes laser-induced fluorescence and reflectance spectroscopy to acquire real-time spectral data of teeth and saliva, and analyzes pathological characteristics to achieve early, non-invasive diagnosis and health assessment. However, within the oral cavity, the contributions of different components (such as water, proteins, enzymes, immune molecules, metabolites, and trace elements) to the spectral response vary significantly. Furthermore, the concentrations and proportions of these biochemical components typically change very little under different health conditions, resulting in insignificant changes in corresponding spectral characteristics. This leads to insufficient accuracy in identifying oral health status using current feature extraction algorithms, consequently resulting in poor effectiveness in assisting oral health monitoring. Summary of the Invention

[0004] To address the shortcomings of existing methods in terms of accuracy in identifying oral health status due to their failure to consider the synergistic changes in oral components, and consequently their poor effectiveness in assisting oral health monitoring, this invention aims to provide an oral health monitoring system based on laser irradiation and its data processing method. The specific technical solution adopted is as follows:

[0005] In a first aspect, the present invention provides a data processing method for an oral health monitoring system based on laser irradiation, comprising:

[0006] Acquire spectral data of each oral cavity region at different wavelengths, wherein the spectral data includes first spectral data of teeth and second spectral data of saliva within the same oral cavity region;

[0007] Based on the characteristic distribution of spectral data of each oral region in each band and the deviation of data under healthy conditions, the spectral residual contrast of spectral data of each oral region in each band is obtained.

[0008] Based on the difference distribution of spectral residual contrast between the first and second spectral data in different bands within each oral cavity region, and combined with the correlation of the characteristic distributions of the corresponding first and second spectral data in various dimensions, the co-variation index of the first and second spectral data in each band within each oral cavity region is obtained.

[0009] Based on the fluctuation of the co-variance index of the first and second spectral data in each oral region under each band, and combined with the spectral residual contrast of the first spectral data under each band, the abnormal attention of the first spectral data in each oral region under each band is obtained.

[0010] The dental health status of each oral region is monitored based on the aforementioned abnormal attention levels.

[0011] Preferably, the step of obtaining the spectral residual contrast of the spectral data of each oral region in each band based on the characteristic distribution of each dimension of the spectral data of each oral region and the data deviation under healthy conditions specifically includes:

[0012] Based on the differences in the spectral data of each oral region in each band and the spectral data of the same oral region in the same band under healthy conditions, the spectral deviation value corresponding to the spectral data of each oral region in each band is obtained.

[0013] Based on the balanced distribution of the spectral deviation values ​​of the same spectral data in all oral regions under each band, the balanced characteristic value under each band is determined.

[0014] Based on the proportion of the difference between the spectral deviation value corresponding to the spectral data of each oral region in each band and the equalization characteristic value of the corresponding spectral data in the same band, the spectral residual contrast of the spectral data of each oral region in each band is determined.

[0015] Preferably, the step of obtaining the spectral deviation value corresponding to the spectral data of each oral region in each band based on the differences in the spectral feature distribution of each oral region in each band with the spectral data of the same oral region in the same band under healthy conditions specifically includes:

[0016] For any oral cavity region and any spectral data in any band, obtain the peak data, wavelength of the peak, peak width and peak area of ​​the spectral data, and construct the characteristic distribution sequence of the spectral data.

[0017] The difference distance between the characteristic distribution sequence of each type of spectral data for each oral region in each band and the characteristic distribution sequence of the same type of spectral data for the same oral region in the same band under healthy conditions is used as the spectral deviation value corresponding to the spectral data of each oral region in each band.

[0018] Preferably, the step of obtaining the coordinated change index of the first and second spectral data in each band within each oral cavity region based on the difference distribution of spectral residual contrast between the first and second spectral data in different bands, combined with the correlation of the characteristic distributions of the corresponding first and second spectral data in various dimensions, specifically includes:

[0019] Based on the difference in spectral residual contrast between the first and second spectral data in different bands within each oral cavity region, the degree of drift consistency between different bands for each oral cavity region is obtained.

[0020] Based on the similarity between the characteristic distribution sequence of the first spectral data of each band and the characteristic distribution sequence of the second spectral data of each band within the same oral cavity region, the degree of spectral correlation between different bands of each oral cavity region is obtained.

[0021] The product of the drift consistency and the spectral correlation is determined as a co-variance index of the first and second spectral data in each band within each oral cavity region.

[0022] Preferably, the step of determining the degree of drift consistency of each oral cavity region across different bands based on the difference in spectral residual contrast between the first and second spectral data at different bands within each oral cavity region specifically includes:

[0023] For any oral cavity region, the first spectral data in any band is denoted as the first characteristic spectrum, and the second spectral data in any band is denoted as the second characteristic spectrum.

[0024] The degree of drift consistency between the first and second characteristic spectra is determined based on the negative correlation coefficient between the spectral residual contrast corresponding to the first characteristic spectrum and the spectral residual contrast corresponding to the second characteristic spectrum.

[0025] Preferably, the step of obtaining the spectral correlation degree between different bands for each oral region based on the similarity between the characteristic distribution sequence of the first spectral data of each band and the characteristic distribution sequence of the second spectral data of each band within the same oral region specifically includes:

[0026] The Pearson correlation coefficient between the characteristic distribution sequence corresponding to the first characteristic spectrum and the characteristic distribution sequence corresponding to the second characteristic spectrum is taken as the degree of spectral correlation between the first characteristic spectrum and the second characteristic spectrum.

[0027] Preferably, the step of obtaining the abnormal attention level of the first spectral data of each oral region in each band based on the fluctuation of the coordinated change index of the first and second spectral data in each band, combined with the spectral residual contrast of the first spectral data in each band, specifically includes:

[0028] Based on the degree of deviation between the co-variance index between the first spectral data and the second spectral data of each band in each oral region and the overall distribution of all oral regions, the anomaly confidence level between the first spectral data and the second spectral data of each band in each oral region is obtained.

[0029] Based on the anomaly confidence level and the co-variance index between the first spectral data of each band and the second spectral data of all bands in each oral region, the degree of response deviation of the first spectral data of each oral region in each band is obtained.

[0030] The product of the response deviation degree and the spectral residual contrast of the first spectral data of the corresponding oral region in the same band is used as the abnormal attention level of the first spectral data of each oral region in each band.

[0031] Preferably, the step of obtaining the anomaly confidence level between the first spectral data and the second spectral data of each band within each oral cavity region based on the degree of deviation between the co-variance index between the first spectral data and the second spectral data of each band within each oral cavity region and the overall distribution of all oral cavity regions specifically includes:

[0032] Select any one oral cavity region as the selected oral cavity region, and select any two bands as the first band and the second band respectively;

[0033] The mean of the co-variance index between the first spectral data of the first band and the second spectral data of the second band in all oral regions is calculated to obtain the co-variance characteristic value;

[0034] The difference between the co-variance index and the co-variance feature value between the first spectral data of the first band and the second spectral data of the second band within the selected oral cavity region is used as the anomaly confidence level between the first spectral data of the first band and the second spectral data of the second band within the selected oral cavity region.

[0035] Preferably, the step of obtaining the response deviation of the first spectral data of each oral region in each band based on the anomaly confidence level and the co-variance index between the first spectral data of each band and the second spectral data of all bands within each oral region specifically includes:

[0036] For a selected oral cavity region, the co-variance index between the first spectral data of the first band and the second spectral data of each band is normalized to obtain the feature weight of the second spectral data of each band.

[0037] Using the aforementioned feature weights, the anomaly confidence levels between the first spectral data of the first band and the second spectral data of each band are weighted and averaged to obtain the response deviation degree of the first spectral data of the selected oral region in the first band.

[0038] Secondly, the present invention provides a laser-based oral health monitoring system, which implements the steps of a data processing method for a laser-based oral health monitoring system, the laser-based oral health monitoring system comprising:

[0039] The data acquisition module is used to acquire spectral data of each oral cavity region under different wavelengths, wherein the spectral data includes first spectral data of teeth and second spectral data of saliva within the same oral cavity region;

[0040] The residual comparison module is used to obtain the spectral residual contrast of the spectral data of each oral region in each band based on the characteristic distribution of each dimension of the spectral data of each oral region and the deviation of the data under healthy conditions.

[0041] The collaborative analysis module is used to obtain the collaborative change index of the first and second spectral data in each band within each oral cavity region based on the difference distribution of the spectral residual contrast between the first and second spectral data in different bands and the correlation of the characteristic distribution of the corresponding first and second spectral data in each dimension.

[0042] The anomaly analysis module is used to obtain the degree of attention to anomalies in the first spectral data of each oral region in each band by combining the fluctuation of the co-variation index of the first and second spectral data in each band with the spectral residual contrast of the first spectral data in each band.

[0043] The health monitoring module is used to monitor the dental health status of each oral region based on the abnormal attention level.

[0044] The embodiments of the present invention have at least the following beneficial effects:

[0045] This invention first collects localized data from the oral cavity region, providing a data foundation for subsequent analysis of the synergistic relationship between teeth and saliva in different localized areas and at different wavelengths. Then, firstly, it analyzes the deviation between the actually collected spectral data and spectral data under healthy conditions, preliminarily quantifying anomalies and deviations in the collected spectral data and obtaining spectral residual contrast. Secondly, it analyzes the differences and correlations in the deviations of spectral data of teeth and saliva at various wavelengths, assessing the synergistic changes in spectral signals between teeth and saliva at different wavelengths within a localized oral cavity region. Matching analysis allows for cross-validation of single data sources, reducing errors caused by noise from single data sources or individual differences. Finally, by comprehensively analyzing the results of the synergistic changes and comparing the deviations, it assesses corresponding abnormalities in teeth, obtains anomaly attention levels, improves the accuracy of oral health status identification, and provides medical personnel with more effective auxiliary diagnostic results. Attached Figure Description

[0046] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart of the steps in a data processing method for an oral health monitoring system based on laser irradiation provided by the present invention;

[0048] Figure 2 This is a partial comparison diagram of the first spectral data and the healthy baseline provided by the present invention;

[0049] Figure 3 This is a flowchart of the steps of the method for obtaining spectral residual contrast provided by the present invention;

[0050] Figure 4 This is a flowchart of the steps for obtaining the coordinated change index provided by the present invention;

[0051] Figure 5 This is a flowchart of the steps of the method for obtaining abnormal attention provided by the present invention;

[0052] Figure 6 This is a schematic diagram of a module of an oral health monitoring system based on laser irradiation provided by the present invention. Detailed Implementation

[0053] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following detailed description, in conjunction with the accompanying drawings and preferred embodiments, describes the specific implementation, structure, features, and effects of an oral health monitoring system and its data processing method based on laser irradiation proposed according to the present invention.

[0054] Before introducing the specific solutions provided in the embodiments of this application, some terms used in this application will be explained to facilitate understanding by those skilled in the art, and will not be used to limit the scope of this application.

[0055] The data acquisition process for oral laser irradiation is as follows:

[0056] I. Instrument Preparation:

[0057] Laser excitation source and spectrometer, output mode: pulse, adjustable power 1-10mW. Spectrometer: resolution ≤1nm, detection range 400–900nm; Probe: two-in-one fiber optic probe with switchable focusing lens for aiming at tooth surface or salivary membrane; Positioning device: adjustable XYZ three-axis displacement stage or intraoral positioning fixture to ensure consistent position for repeated measurements; Controller: triggers synchronous acquisition of laser and spectrometer data.

[0058] II. Acquisition of Dental Spectral Data:

[0059] Patient preparation: Rinse with water, dry the tooth surface (or blow dry gently) to remove saliva residue, fix the soft tissue with a mouth mirror or retractor, and expose the scanning area.

[0060] Probe alignment: Keep the fiber optic probe perpendicular to the tooth surface at a distance of about 2nm to 3mm, and cover the entire tooth surface in a grid or line scan manner (e.g., move it once every 1mm).

[0061] Laser irradiation and signal acquisition: Continuous irradiation with a laser at 655nm, pulse width of 10ms, interval of 100ms, and the spectrometer integration time set from 50ms to 100ms. Acquisition was performed 3 times and the average was taken to reduce random errors. The original spectrum (wavelength, intensity) and the scanning position were recorded by taking pictures.

[0062] Data annotation: Associate each set of spectra with tooth number and scanning coordinates.

[0063] Saliva sample preparation: Obtain a small amount of naturally secreted saliva (approximately 10–20 μL) from the mouth, or collect it immediately after gently touching the gums in different target areas. The sample can be directly formed into a thin film on a glass slide or placed in the detection slot of a microfluidic chip.

[0064] Probe positioning and laser irradiation: Use the same or switch to a 405nm laser source. Lightly touch the surface of the saliva film with the probe, keeping it in contact with the sample but not pressing it. Reduce the laser power to 1mW to 5mW to avoid sample flickering and thermal effects.

[0065] In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments may be combined in any suitable form.

[0066] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0067] The following description, in conjunction with the accompanying drawings, details a specific scheme for an oral health monitoring system based on laser irradiation and its data processing method provided by the present invention.

[0068] Please see Figure 1 The diagram illustrates a flowchart of a data processing method for a laser-based oral health monitoring system according to an embodiment of the present invention. The method includes the following steps:

[0069] Step S100: Obtain spectral data of each oral cavity region under different wavelengths, wherein the spectral data includes first spectral data of teeth and second spectral data of saliva within the same oral cavity region.

[0070] First, based on the structure of the human oral cavity, the oral cavity is divided into several functional areas. As a specific example, each tooth can correspond to an oral cavity area, providing a data foundation for the subsequent feature analysis process for data anomalies of different teeth.

[0071] This embodiment describes the distribution of spectral data for teeth and saliva within the same oral cavity region. As a specific example, this embodiment acquires first spectral data for teeth and second spectral data for saliva within the 400nm to 900nm wavelength range. This range is uniformly divided into multiple different bands, each with the same wavelength range. For example, a 100nm wavelength range can be used, meaning each wavelength range corresponds to one band. In this embodiment, a total of five different bands are included. It should be understood that this embodiment uses laser-induced fluorescence and reflectance spectroscopy to acquire spectral data for teeth and saliva, a technique well-known to those skilled in the art and will not be described in detail here.

[0072] Thus, two types of spectral data can be obtained in each oral cavity region under each band: one is the spectral data of the teeth, which is referred to as the first spectral data in this embodiment, and the other is the spectral data of saliva, which is referred to as the second spectral data.

[0073] Step S200: Based on the characteristic distribution of spectral data of each oral region in each band and the deviation of data under healthy conditions, obtain the spectral residual contrast of spectral data of each oral region in each band.

[0074] When oral health problems exist, the spectral responses of various components within the oral cavity exhibit abnormalities, thus aiding in the diagnosis of oral health. During the extraction of abnormal components from spectral data, the most significant deviations from the healthy baseline often correspond to early pathological changes or potential discrepancies. For example, a local comparison image of the first spectral data of a tooth and its corresponding healthy baseline spectral data might be shown. Figure 2 As shown, the degree of deviation of each component is obtained by comparing the actual acquired spectra with the healthy baseline. Among them, Figure 2 The horizontal axis represents wavelength, and the vertical axis represents signal intensity of spectral data. The dashed line represents the spectral curve of the healthy baseline, and the solid line represents the spectral data of patients with oral problems, which can reflect the differences in spectral data between corresponding bands.

[0075] As a concrete example, such as Figure 3 As shown, the method for obtaining the spectral residual contrast can be implemented by steps S201 to S203.

[0076] Step S201: Based on the differences in the spectral data of each oral region in each band and the spectral data of the same oral region in the same band under healthy conditions, obtain the spectral deviation value corresponding to the spectral data of each oral region in each band.

[0077] First, it's necessary to obtain healthy oral cavity data as a foundation for comparative analysis. In practice, different normal oral cavity databases can be constructed for different age groups. It should be understood that dental departments or hospitals store patient information in their internal systems. By acquiring and statistically analyzing the spectral data of patients of the same age group in terms of dental health, and for each identical oral region, the mean of the dental spectral data of all healthy samples from the same age group is calculated to constitute the first health data for the teeth in that region. Similarly, the mean of the saliva spectral data of all healthy samples from the same age group is calculated to constitute the second health data for the saliva in that region. It should be noted that age groups can be categorized by medical personnel based on the specific implementation scenario.

[0078] Thus, each oral cavity region corresponds to a set of spectral data under a healthy state in each wavelength band, including healthy spectral data for teeth (first healthy data) and healthy spectral data for saliva (second healthy data). For ease of description, this embodiment assumes that feature analysis is performed using the monitoring data of a single patient, whose age group has been determined. The corresponding first and second healthy data are retrieved from the database for analysis.

[0079] The first step is to obtain the peak data, peak wavelength, peak width, and peak area of ​​any spectral data in any band for any oral cavity region, and construct the characteristic distribution sequence of the spectral data.

[0080] Here, peak data refers to the maximum signal intensity value on the selected spectral data, and peak width refers to the half-maximum width of the peak on the selected spectrum, that is, the width of the peak at half the position of the peak data. Specifically, it can be obtained by obtaining a signal intensity equal to the peak data. The distance between the two wavelengths corresponding to the signal strength is taken as the peak width. The peak area can be calculated by integration, which is a well-known technique and will not be discussed further here.

[0081] It should be understood that feature extraction is performed on each spectral data in each band of each oral cavity region using the same method. At the same time, feature extraction is also performed on the spectral data in a healthy state to obtain the feature distribution sequence of each spectral data in each band of each oral cavity region.

[0082] The second step is to use the difference distance between the characteristic distribution sequence of each type of spectral data for each oral region in each band and the characteristic distribution sequence of the same type of spectral data for the same oral region in the same band under healthy conditions as the spectral deviation value corresponding to the spectral data of each oral region in each band.

[0083] In this embodiment, the analysis process is described using oral feature data of teeth and saliva corresponding to any oral cavity region as an example. Specifically, the analysis is illustrated using the first spectral data of the i-th band and the second spectral data of the i-th band contained in any oral cavity region as an example.

[0084] For the first spectral data corresponding to the teeth, calculate the DTW distance of the characteristic distribution sequence between the first spectral data of the i-th band and the first health data of the same oral region in the i-th band, and use it as the spectral deviation value corresponding to the first spectral data of the i-th band.

[0085] For the second spectral data corresponding to saliva, the DTW distance of the characteristic distribution sequence between the second spectral data of the i-th band and the second health data of the same oral region in the i-th band is calculated, and used as the spectral deviation value corresponding to the second spectral data of the i-th band.

[0086] It should be noted that the calculation method for the spectral deviation value is the same for each type of spectral data. The spectral deviation value characterizes the degree of characteristic difference and deviation between each type of spectral data and the spectral data under the corresponding healthy state.

[0087] Step S202: Based on the balanced distribution of the spectral deviation values ​​of the same spectral data in all oral regions under each band, determine the balanced characteristic value under each band.

[0088] For the first spectral data of teeth in the oral cavity, the mean of the spectral deviation values ​​corresponding to the first spectral data of all oral cavity regions under the i-th band is used as the equilibrium feature value corresponding to the first spectral data under the i-th band, which represents the overall distribution of the deviation of the first spectral data of teeth from the healthy state in all oral cavity regions under the same band.

[0089] It should be understood that the second spectral data of saliva is also calculated and analyzed in the same way. The mean value of the spectral deviation value corresponding to the second spectral data of all oral regions under the i-th band is used as the equilibrium characteristic value corresponding to the second spectral data under the i-th band, which represents the overall distribution of the deviation of the second spectral data of saliva from the healthy state in all oral regions of the entire oral cavity under the same band.

[0090] Step S203: Based on the difference ratio between the spectral deviation value corresponding to the spectral data of each oral region in each band and the equalization characteristic value of the corresponding spectral data in the same band, determine the spectral residual contrast of the spectral data of each oral region in each band.

[0091] By comparing the laser spectra of multiple regions in the oral cavity of the same individual, and comparing each band of the saliva spectrum obtained from each region with the saliva spectrum of other teeth and other regions, the local pathological signals are amplified and the residual contrast of each band is extracted.

[0092] For the first spectral data of teeth within the oral cavity region, calculate the absolute value of the difference between the spectral deviation value corresponding to the first spectral data in the i-th band and the equalization characteristic value corresponding to the first spectral data in the i-th band. The ratio of the absolute value of the difference to the equalization characteristic value is taken as the difference proportion, denoted as the spectral residual contrast of the first spectral data in the i-th band, which can be expressed by the formula:

[0093] ;

[0094] in, This represents the spectral residual contrast of the first spectral data in the i-th band of the oral cavity region. This represents the spectral deviation value of the first spectral data for the oral cavity region in the i-th band. This represents the equilibrium characteristic value of the first spectral data of the oral cavity region in the i-th band.

[0095] It should be understood that the calculation method for the spectral residual contrast of the second spectral data of saliva in each band of the oral cavity is the same as the calculation method for the spectral residual contrast of the first spectral data of teeth in each band of the oral cavity. It is only necessary to replace the data involved in the calculation with the relevant data in the same dimension of the corresponding type of spectral data.

[0096] The spectral residual contrast of each spectral data in each band in the oral cavity region characterizes the residual distribution of the difference between each spectral data and the overall characteristic distribution of the oral cavity, and can preliminarily reflect the degree of manifestation of local pathological signals.

[0097] Step S300: Based on the difference distribution of spectral residual contrast between the first spectral data and the second spectral data in different bands within each oral cavity region, and combined with the correlation of the characteristic distributions of the corresponding first spectral data and the second spectral data in each dimension, the collaborative change index of the first spectral data and the second spectral data in each band within each oral cavity region is obtained.

[0098] In oral spectroscopy, signals from major components such as water and proteins often occupy most of the dynamic range. Pathological changes in trace metabolites or minerals are often submerged in this large background signal. Therefore, synergistic analysis of tooth and saliva spectra is necessary to determine the existence of pathological abnormalities, which is more reliable than observing a single spectral signal. Thus, it is essential to combine data from teeth and saliva within the same oral region to analyze the synergistic relationships between different spectral bands in both.

[0099] Based on this, such as Figure 4 As shown, the method for analyzing the co-variation index of teeth and saliva in each oral cavity region under each band can be implemented by steps S301 to S303.

[0100] Step S301: Based on the difference in spectral residual contrast between the first spectral data and the second spectral data in different bands within each oral cavity region, the degree of drift consistency between different bands for each oral cavity region is obtained.

[0101] When oral problems are present, deviations occur in the spectral responses of multiple bands. The residual contrast of each band in the oral spectral data should change accordingly. If there are mutual influences between bands, the changes in their spectral residual contrast are also correlated. By analyzing the spectral residual contrast between bands, the consistency of spectral drift in tooth and saliva spectral data across bands can be obtained.

[0102] Specifically, for any oral cavity region, the first spectral data in any band is denoted as the first characteristic spectrum, and the second spectral data in any band is denoted as the second characteristic spectrum. As a specific example, the first spectral data of the teeth in the x-th oral cavity region in the n-th band is taken as the first characteristic spectrum, and the second spectral data of the saliva in the x-th oral cavity region in the m-th band is taken as the second characteristic spectrum. It should be understood that n and m are both band numbers, and their values ​​can be the same or different, used to represent the distribution of data differences between different types of spectral data corresponding to the same or different bands within the same oral cavity region.

[0103] Furthermore, based on the negative correlation coefficient between the spectral residual contrast corresponding to the first characteristic spectrum and the spectral residual contrast corresponding to the second characteristic spectrum, the degree of drift consistency between the first and second characteristic spectra is determined. As a concrete example, the degree of drift consistency between the first spectral data of the tooth in the x-th oral region at the n-th band and the second spectral data of saliva in the m-th band can be expressed by the formula:

[0104] ;

[0105] in, This indicates the degree of drift consistency between the first spectral data of the tooth in the x-th oral region under the n-th band and the second spectral data of the saliva in the m-th band, which is also the degree of drift consistency between the first characteristic spectrum and the second characteristic spectrum in the x-th oral region. This represents the spectral residual contrast of the first spectral data of the x-th oral region in the n-th band, which is also the spectral residual contrast of the first characteristic spectrum of the x-th oral region. The spectral residual contrast of the second spectral data of the x-th oral region in the m-th band represents the spectral residual contrast of the second characteristic spectrum of the x-th oral region. Represents an exponential function with base e, using The differences are negatively correlated in the form of [the method described].

[0106] This represents the difference in spectral residual contrast between the first characteristic spectrum and the second characteristic spectrum, when the ratio... The smaller the difference from 1, the closer the spectral residual contrast between the two is. In other words, the smaller the difference, the more consistent the degree of signal deviation change, and the larger the corresponding drift consistency value. Drift consistency reflects the degree of consistency in the signal deviation changes of spectral data of teeth and saliva in different bands within the same oral cavity region.

[0107] Step S302: Based on the similarity between the characteristic distribution sequence of the first spectral data of each band and the characteristic distribution sequence of the second spectral data of each band within the same oral cavity region, the spectral correlation degree between different bands of each oral cavity region is obtained.

[0108] When oral health problems exist, they manifest as changes in the spectral characteristics of relevant bands. By analyzing the correlation between related bands in different types of spectral data, the reliability of abnormal bands can be improved. Based on the spectral performance of each band in the oral spectral dataset, the correlation between bands in the spectral data of teeth and saliva can be obtained.

[0109] Specifically, the Pearson correlation coefficient between the characteristic distribution sequences corresponding to the first and second characteristic spectra is used as the spectral correlation between the first and second characteristic spectra. The spectral correlation reflects the similarity of the characteristic distributions of the spectral data of teeth and saliva in different bands within the same oral cavity region.

[0110] Step S303: The product of the drift consistency and the spectral correlation is determined as a co-variance index of the first and second spectral data in each band within each oral cavity region.

[0111] By comprehensively evaluating the synergistic changes in spectral data of teeth and saliva across different bands within the same oral cavity region, as well as the consistency of signal deviation changes, we can effectively assess these relationships.

[0112] The greater the spectral correlation between the first and second characteristic spectra within the oral cavity region, the more similar the data feature distributions of the first and second characteristic spectra are. This indicates that the spectral data distributions of teeth and saliva in the corresponding bands are similar. At the same time, the greater the drift consistency between the first and second characteristic spectra within the oral cavity region, the higher the consistency of the signal deviation changes between the first and second characteristic spectra are. This indicates that the deviation changes of the spectral data of teeth and saliva in the corresponding bands are similar. In this case, the synergistic relationship between the first and second characteristic spectra within the oral cavity region is strong.

[0113] Step S400: Based on the fluctuation of the coordinated change index of the first spectral data and the second spectral data in each band within each oral cavity region, and combined with the spectral residual contrast of the first spectral data in each band, the abnormal attention level of the first spectral data in each oral cavity region in each band is obtained.

[0114] The physiological functions and component distribution of teeth and saliva can reflect a certain data synergy in terms of spectral response. By simultaneously collecting spectral data of teeth and surrounding saliva and analyzing their matching, the reliability of abnormality identification can be significantly improved when both show matching anomalies or when specific bands deviate from the healthy baseline. This amplifies weak but real pathological signals, making the extracted features more valuable for auxiliary reference, thereby significantly improving the sensitivity and specificity of oral health monitoring.

[0115] Based on this, such as Figure 5 As shown, the method for obtaining the abnormal attention of the first spectral data of each oral cavity region in each band can be implemented by steps S401 to S403.

[0116] Step S401: Based on the degree of deviation between the co-variance index between the first spectral data and the second spectral data of each band in each oral region and the overall distribution of all oral regions, the anomaly confidence level between the first spectral data and the second spectral data of each band in each oral region is obtained.

[0117] By comparing the co-variation relationships between bands in the spectral data of different local regions of the oral cavity, the greater the difference in the co-variation relationship in the problematic oral region, the higher the reliability of the spectral signal anomaly in the corresponding band, that is, the higher the degree of characteristic anomaly exhibited. The reliability of the spectral signal anomaly in a band is obtained based on the differences in the co-variation relationships in different regions.

[0118] The first step is to select any oral cavity region as the selected oral cavity region, and any two bands as the first band and the second band, respectively. As a specific example, in this embodiment, the y-th oral cavity region is selected, the i-th band is the first band, and the j-th band is the second band. Based on the reason that the n-th band and the m-th band are the same, i and j are both the band numbers, which can be the same or different, and will be used later to represent the spectral data of the corresponding bands of teeth and saliva, respectively.

[0119] The second step is to calculate the mean of the co-variance index between the first spectral data of the first band and the second spectral data of the second band in all oral regions, and obtain the co-variance characteristic value.

[0120] The third step is to use the difference between the co-variance index and the co-variance feature value between the first spectral data of the first band and the second spectral data of the second band in the selected oral cavity region as the anomaly confidence level between the first spectral data of the first band and the second spectral data of the second band in the selected oral cavity region.

[0121] As a concrete example, the formula for calculating the credibility of anomalies can be expressed as: ,in, This represents the anomaly confidence level between the first spectral data in the i-th band and the second spectral data in the j-th band within the y-th oral region. This represents the co-variance index between the first spectral data in the i-th band and the second spectral data in the j-th band within the y-th oral region. This represents the mean of the co-variance index between the first spectral data in the i-th band and the second spectral data in the j-th band across all oral regions, which is also known as the co-variance characteristic value.

[0122] It reflects the difference in the cooperative variation relationship between the first spectral data in the i-th band and the second spectral data in the j-th band in the y-th oral cavity region and the whole oral cavity region. The larger the value, the more serious the deviation of the cooperative variation relationship in the y-th region from the overall distribution, the greater the degree of anomaly, and the greater the corresponding anomaly confidence value.

[0123] Step S402: Based on the anomaly confidence level and the co-variance index between the first spectral data of each band and the second spectral data of all bands within each oral cavity region, obtain the response deviation degree of the first spectral data of each oral cavity region in each band.

[0124] When oral problems exist, the degree of abnormality in the spectral signal of teeth in the oral cavity area is of great reference value. Due to various interferences, the spectral features may be relatively weak. By studying the synergistic changes of teeth and saliva in each band, the spectral abnormalities in each band can be amplified, and a weighted comprehensive assessment of dental abnormalities in each oral cavity area can be performed.

[0125] The first step, for a selected oral cavity region, is to normalize the co-variance index between the first spectral data of the first band and the second spectral data of each band to obtain the feature weights of the second spectral data of each band. The normalization method is a well-known technique and will not be described in detail here.

[0126] The second step involves using the aforementioned feature weights to perform a weighted average of the anomaly confidence levels between the first spectral data of the first band and the second spectral data of each band, thereby obtaining the degree of response deviation of the first spectral data of the selected oral region in the first band.

[0127] By analyzing the co-variation relationships between spectral data corresponding to teeth and saliva, a weighted average of the anomalies in the spectral signals is applied. The closer the co-variation relationship between two bands, the more valuable the band deviation of the corresponding spectral data is as a supplementary reference. As a concrete example, the degree of response deviation can be expressed by the formula:

[0128] ;

[0129] in, This indicates the degree of response deviation of the first spectral data in the i-th band within the y-th oral region. This represents the anomaly confidence level between the first spectral data in the i-th band and the second spectral data in the j-th band within the y-th oral region. This represents the co-variance index between the first spectral data in the i-th band and the second spectral data in the j-th band within the y-th oral region. This indicates the number of bands, and Norm is the normalization function.

[0130] The larger the value of the synergistic change index, the stronger the synergistic relationship between the spectral signal changes of the first spectral data in the first band and the second spectral data in the second band. Consequently, the reference value of the data anomalies shown by the first spectral data in the first band and the second spectral data in the second band is higher.

[0131] Step S403: The product of the response deviation degree and the spectral residual contrast of the first spectral data of the corresponding oral region in the same band is used as the abnormal attention level of the first spectral data of each oral region in each band.

[0132] The degree of response deviation of the first spectral data in the i-th band within the y-th oral region reflects the degree of abnormality in the spectral data of the teeth in that band within the corresponding oral region. The spectral residual contrast of the first spectral data in the i-th band within the y-th oral region reflects the degree of abnormality in the spectral data of the teeth in that band compared to a healthy state within the corresponding oral region. Combining these two characteristics allows for a relatively accurate characterization of the degree of abnormality exhibited by the teeth in the i-th band within the y-th oral region in terms of spectral signal. The greater the degree of abnormality, the greater the attention required, meaning that the teeth in that oral region need to be closely observed to aid in oral diagnosis.

[0133] Step S500: Monitor the dental health status of each oral region based on the abnormal attention level.

[0134] The higher the value of the abnormal attention score of the first spectral data corresponding to the teeth in each band within each oral cavity region, the more abnormal the spectral signal of the teeth in the corresponding band within that oral cavity region is, and the more attention needs to be paid to the teeth in that oral cavity region.

[0135] In some embodiments, a threshold can be used to alert medical staff to pay close attention to teeth within a specific oral region, thereby enabling auxiliary diagnosis of teeth using spectral data. As a concrete example, abnormal attention levels are normalized. When the normalized abnormal attention level is greater than or equal to a preset abnormality threshold, medical staff are alerted to pay close attention to the teeth within that oral region. When the normalized abnormal attention level is less than the preset abnormality threshold, it indicates that the likelihood of abnormalities in the teeth within that oral region is low, and therefore no further attention is needed, i.e., no alert is required. The abnormality threshold can be set to 0.7, and implementers can adjust this setting according to specific implementation scenarios. It should be noted that the normalization method is a well-known technique and will not be described in detail here.

[0136] In some embodiments, neural networks can also be used to output results indicating whether teeth are abnormal or normal, providing medical personnel with data for reference and assisting in the formal diagnostic process. That is, the higher the abnormal attention value of the first spectral data corresponding to the teeth in each band within each oral region, the higher the attention weight value in the corresponding dimension, and the neural network can directly output the health status or abnormal status of the teeth in the corresponding oral region.

[0137] As a concrete example, a hybrid architecture of convolutional neural network and Transformer encoder is adopted: first, convolutional layers are used to extract microscopic features from oral spectral sequences, and then a multi-head self-attention Transformer module is used to adaptively assign attention weights to each band position, amplify the band response most relevant to the lesion, and finally, after global pooling, it is connected to the classification head to realize the determination of oral health or abnormality.

[0138] After denoising, baseline correction, and normalization, the first spectral data of the teeth in all bands are input into the convolutional layer. The convolutional kernel slides between adjacent wavelength points, enabling the model to sensitively capture microscopic features and screen local chemical composition or structure in each small band, thereby enhancing the initial response to subtle pathological signals.

[0139] The convolutional output is reshaped along the wavelength axis and positionally encoded before being fed into a multi-layer Transformer encoder. Its multi-head self-attention mechanism automatically assigns weights to each band position. In the self-attention calculation, the abnormal attention levels of the first spectral data for each band obtained above are superimposed as a bias term for attention scoring. This causes the attention weights to automatically favor wavelengths with higher diagnostic value during forward propagation. That is, self-attention only assigns high weights when certain wavelengths show abnormalities in the combined analysis of dental and salivary spectra, thereby eliminating incidental background and individual differences.

[0140] The final output of the Transformer is a feature map of each band after attention weighting. It is then converged into a fixed-dimensional semantic vector through global average pooling, eliminating positional differences. The nonlinear decision boundary between health and abnormality is learned through a multilayer perceptron (MLP), and a healthy or abnormal result is output to realize the monitoring of oral health risks.

[0141] In summary, the analysis of oral health status using laser spectral data is hampered by the subtle changes in the concentration and proportion of oral components, resulting in insignificant changes in spectral characteristics. Therefore, this invention, through the synergistic analysis of tooth and saliva spectral data, extracts key features reflecting subtle biochemical and structural changes, thereby enabling early, accurate, and non-invasive monitoring of oral health status. This improves detection sensitivity and accuracy, reduces the risk of misdiagnosis, and provides a reliable basis for personalized oral health management.

[0142] By comparing the standard spectrum under healthy conditions, minute changes in component proportions can be detected, providing early warning signals for pathological conditions and initially screening out spectral regions that differ significantly from normal data. Detailed analysis of different areas within the oral cavity (such as local areas of teeth and saliva) can accurately locate the distribution areas of abnormal components. By calculating the spectral contrast between abnormal components and the normal background, the degree of abnormality can be further quantified, improving the clarity and reliability of diagnostic signals. Under healthy oral conditions, the biochemical components in the hard tissues of teeth and saliva should exhibit a certain degree of synergistic consistency. Matching analysis can cross-validate single data sources, reducing errors caused by noise from single data points or individual differences. This effectively distinguishes between normal and abnormal states, reduces the misdiagnosis rate, and improves the accuracy and robustness of oral health monitoring.

[0143] like Figure 6 As shown, the present invention also provides a laser-based oral health monitoring system, which implements the steps of a data processing method for a laser-based oral health monitoring system. The laser-based oral health monitoring system includes:

[0144] The data acquisition module is used to acquire spectral data of each oral cavity region under different wavelengths, wherein the spectral data includes first spectral data of teeth and second spectral data of saliva within the same oral cavity region;

[0145] The residual comparison module is used to obtain the spectral residual contrast of the spectral data of each oral region in each band based on the characteristic distribution of each dimension of the spectral data of each oral region and the deviation of the data under healthy conditions.

[0146] The collaborative analysis module is used to obtain the collaborative change index of the first and second spectral data in each band within each oral cavity region based on the difference distribution of the spectral residual contrast between the first and second spectral data in different bands and the correlation of the characteristic distribution of the corresponding first and second spectral data in each dimension.

[0147] The anomaly analysis module is used to obtain the degree of attention to anomalies in the first spectral data of each oral region in each band by combining the fluctuation of the co-variation index of the first and second spectral data in each band with the spectral residual contrast of the first spectral data in each band.

[0148] The health monitoring module is used to monitor the dental health status of each oral region based on the abnormal attention level.

[0149] Since the specific implementation process of a data processing method for a laser-based oral health monitoring system has already been described in detail, it will not be repeated here.

[0150] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A data processing method for an oral health monitoring system based on laser irradiation, characterized in that, The method includes the following steps: Acquire spectral data of each oral cavity region at different wavelengths, wherein the spectral data includes first spectral data of teeth and second spectral data of saliva within the same oral cavity region; Based on the characteristic distribution of spectral data of each oral region in each band and the deviation of data under healthy conditions, the spectral residual contrast of spectral data of each oral region in each band is obtained. Based on the difference distribution of spectral residual contrast between the first and second spectral data in different bands within each oral cavity region, and combined with the correlation of the characteristic distributions of the corresponding first and second spectral data in various dimensions, the co-variation index of the first and second spectral data in each band within each oral cavity region is obtained. Based on the fluctuation of the co-variance index of the first and second spectral data in each oral region under each band, and combined with the spectral residual contrast of the first spectral data under each band, the abnormal attention of the first spectral data in each oral region under each band is obtained. The dental health status of each oral region is monitored based on the aforementioned abnormal attention levels.

2. The data processing method for an oral health monitoring system based on laser irradiation according to claim 1, characterized in that, The method involves obtaining the spectral residual contrast of the spectral data for each oral region in each band based on the characteristic distribution of each dimension of the spectral data for each oral region and the data deviation under healthy conditions. Specifically, this includes: Based on the differences in the spectral data of each oral region in each band and the spectral data of the same oral region in the same band under healthy conditions, the spectral deviation value corresponding to the spectral data of each oral region in each band is obtained. Based on the balanced distribution of the spectral deviation values ​​of the same spectral data in all oral regions under each band, the balanced characteristic value under each band is determined. Based on the proportion of the difference between the spectral deviation value corresponding to the spectral data of each oral region in each band and the equalization characteristic value of the corresponding spectral data in the same band, the spectral residual contrast of the spectral data of each oral region in each band is determined.

3. The data processing method for a laser-based oral health monitoring system according to claim 2, characterized in that, The method involves obtaining the spectral deviation value corresponding to the spectral data of each oral region in each band based on the differences in the spectral feature distribution of each oral region in each band compared to the spectral data of the same oral region in the same band under healthy conditions. Specifically, this includes: For any oral cavity region and any spectral data in any band, obtain the peak data, wavelength of the peak, peak width and peak area of ​​the spectral data, and construct the characteristic distribution sequence of the spectral data. The difference distance between the characteristic distribution sequence of each type of spectral data for each oral region in each band and the characteristic distribution sequence of the same type of spectral data for the same oral region in the same band under healthy conditions is used as the spectral deviation value corresponding to the spectral data of each oral region in each band.

4. The data processing method for a laser-based oral health monitoring system according to claim 3, characterized in that, Based on the difference distribution of spectral residual contrast between the first and second spectral data in different bands within each oral cavity region, and combined with the correlation of the characteristic distributions of the corresponding first and second spectral data in various dimensions, a collaborative change index of the first and second spectral data in each band within each oral cavity region is obtained, specifically including: Based on the difference in spectral residual contrast between the first and second spectral data in different bands within each oral cavity region, the degree of drift consistency between different bands for each oral cavity region is obtained. Based on the similarity between the characteristic distribution sequence of the first spectral data of each band and the characteristic distribution sequence of the second spectral data of each band within the same oral cavity region, the degree of spectral correlation between different bands of each oral cavity region is obtained. The product of the drift consistency and the spectral correlation is determined as a co-variance index of the first and second spectral data in each band within each oral cavity region.

5. The data processing method for a laser-based oral health monitoring system according to claim 4, characterized in that, The method of determining the degree of drift consistency of each oral cavity region across different bands based on the difference in spectral residual contrast between the first and second spectral data in different bands within each oral cavity region specifically includes: For any oral cavity region, the first spectral data in any band is denoted as the first characteristic spectrum, and the second spectral data in any band is denoted as the second characteristic spectrum. The degree of drift consistency between the first and second characteristic spectra is determined based on the negative correlation coefficient between the spectral residual contrast corresponding to the first characteristic spectrum and the spectral residual contrast corresponding to the second characteristic spectrum.

6. The data processing method for a laser-based oral health monitoring system according to claim 5, characterized in that, The method of obtaining the spectral correlation degree between different bands for each oral region based on the similarity between the characteristic distribution sequence of the first spectral data of each band and the characteristic distribution sequence of the second spectral data of each band within the same oral region specifically includes: The Pearson correlation coefficient between the characteristic distribution sequence corresponding to the first characteristic spectrum and the characteristic distribution sequence corresponding to the second characteristic spectrum is taken as the degree of spectral correlation between the first characteristic spectrum and the second characteristic spectrum.

7. The data processing method for a laser-based oral health monitoring system according to claim 1, characterized in that, The method involves analyzing the fluctuations of the coordinated change index of the first and second spectral data within each oral cavity region in each band, and combining this with the spectral residual contrast of the first spectral data in each band to obtain the abnormal attention level of the first spectral data for each oral cavity region in each band. Specifically, this includes: Based on the degree of deviation between the co-variance index between the first spectral data and the second spectral data of each band in each oral region and the overall distribution of all oral regions, the anomaly confidence level between the first spectral data and the second spectral data of each band in each oral region is obtained. Based on the aforementioned anomaly confidence level and the co-variance index between the first spectral data of each band and the second spectral data of all bands within each oral cavity region, the degree of response deviation of the first spectral data of each oral cavity region in each band is obtained. The product of the response deviation degree and the spectral residual contrast of the first spectral data of the corresponding oral region in the same band is used as the abnormal attention level of the first spectral data of each oral region in each band.

8. The data processing method for a laser-based oral health monitoring system according to claim 7, characterized in that, The degree of discrepancy between the co-variance index between the first spectral data and the second spectral data of each band within each oral region and the overall distribution of all oral regions is used to determine the anomaly confidence level between the first spectral data and the second spectral data of each band within each oral region. Specifically, this includes: Select any one oral cavity region as the selected oral cavity region, and select any two bands as the first band and the second band respectively; The mean of the co-variance index between the first spectral data of the first band and the second spectral data of the second band in all oral regions is calculated to obtain the co-variance characteristic value; The difference between the co-variance index and the co-variance feature value between the first spectral data of the first band and the second spectral data of the second band within the selected oral cavity region is used as the anomaly confidence level between the first spectral data of the first band and the second spectral data of the second band within the selected oral cavity region.

9. The data processing method for a laser-based oral health monitoring system according to claim 8, characterized in that, The method of obtaining the response deviation of the first spectral data of each oral region in each band based on the anomaly confidence level and the co-variance index between the first spectral data of each band and the second spectral data of all bands within each oral region specifically includes: For a selected oral cavity region, the co-variance index between the first spectral data of the first band and the second spectral data of each band is normalized to obtain the feature weight of the second spectral data of each band. Using the aforementioned feature weights, the anomaly confidence levels between the first spectral data of the first band and the second spectral data of each band are weighted and averaged to obtain the response deviation degree of the first spectral data of the selected oral region in the first band.

10. An oral health monitoring system based on laser irradiation, characterized in that, This system is used to implement the data processing method of a laser-based oral health monitoring system as described in any one of claims 1-9, wherein the laser-based oral health monitoring system comprises: The data acquisition module is used to acquire spectral data of each oral cavity region under different wavelengths, wherein the spectral data includes first spectral data of teeth and second spectral data of saliva within the same oral cavity region; The residual comparison module is used to obtain the spectral residual contrast of the spectral data of each oral region in each band based on the characteristic distribution of each dimension of the spectral data of each oral region and the deviation of the data under healthy conditions. The collaborative analysis module is used to obtain the collaborative change index of the first and second spectral data in each band within each oral cavity region based on the difference distribution of the spectral residual contrast between the first and second spectral data in different bands and the correlation of the characteristic distribution of the corresponding first and second spectral data in each dimension. The anomaly analysis module is used to obtain the degree of attention to anomalies in the first spectral data of each oral region in each band by combining the fluctuation of the co-variation index of the first and second spectral data in each band with the spectral residual contrast of the first spectral data in each band. The health monitoring module is used to monitor the dental health status of each oral region based on the abnormal attention level.

Images