Multi-sensor fused intelligent oral health monitoring and early warning system and method thereof

Abstract

The invention relates to the technical field of oral health informatization systems, in particular to a multi-sensor fusion intelligent oral health monitoring and early warning system and method. Sending the collected data to a data processing module; the data processing module is in communication connection with the data acquisition module; performing preprocessing and feature extraction on the sensor data; generating an oral health condition evaluation result; the anomaly detection module is in communication connection with the data processing module; performing anomaly detection on the evaluation result of the oral health condition; generating an anomaly detection report; the health management module is in communication connection with the anomaly detection module; generating personalized oral health management suggestions based on the anomaly detection report; the user interface module is in communication connection with the health management module; displaying the oral health management suggestions to the user; feedback information of the user is received and sent to the health management module, abnormal changes of the oral health condition of the user can be found in time, and possibility is provided for early prevention.

Description

Technical Field

[0001] The present invention relates to the technical field of oral health information systems, and in particular to a multi-sensor fusion intelligent oral health monitoring and early warning system and method thereof. Background Art

[0002] As people's awareness of the importance of oral health continues to increase, intelligent oral health management solutions are gaining more and more attention. Traditional oral health management mainly relies on regular dental examinations and personal daily care, which has problems such as poor real-time performance and low personalization. In recent years, with the development of the Internet of Things and artificial intelligence technologies, some intelligent oral care devices have begun to enter the market, such as smart toothbrushes and smart water cups. These devices can collect some basic oral care data and provide users with simple feedback and suggestions.

[0003] However, the existing intelligent oral health management systems still have many shortcomings. First, most systems can only collect data in a single dimension, such as focusing only on brushing behavior, which cannot fully reflect the user's oral health status. Secondly, the data analysis capabilities are limited, and they can often only provide simple statistical results, lacking in-depth anomaly detection and early warning functions. Furthermore, the health management recommendations of existing systems are often static and general, and cannot be dynamically adjusted based on individual differences and usage feedback from users. In addition, the user behavior verification mechanism is imperfect, making it difficult to ensure that the data provided by users is authentic and reliable. Finally, existing systems generally lack the ability to efficiently store and process massive user data, limiting the scalability and large-scale application potential of the system.

[0004] These problems have severely restricted the actual effect and promotion of intelligent oral health management systems. Users cannot obtain comprehensive, accurate, and personalized oral health assessments and management recommendations, which makes it difficult to detect and effectively prevent oral health problems in a timely manner. At the same time, the oral health research field has also been hampered by the lack of high-quality, large-scale user data. Summary of the invention

[0005] In response to the above problems, the present invention proposes a multi-sensor fusion intelligent oral health monitoring and early warning system and its method. The invention aims to build a comprehensive, intelligent and reliable oral health management platform through innovative technologies such as multi-sensor data fusion, intelligent anomaly detection, personalized health management, behavior verification and distributed data storage.

[0006] The present invention proposes a multi-sensor fusion intelligent oral health monitoring and early warning system, including:

[0007] Data acquisition module for:

[0008] Collect sensor data from smart toothbrushes, smart water cups, and smart dental floss;

[0009] Sending the collected data to the data processing module;

[0010] A data processing module is connected to the data acquisition module for:

[0011] Receiving sensor data sent by the data acquisition module;

[0012] Preprocessing and feature extracting the sensor data;

[0013] Based on the results of the preprocessing and feature extraction, generating an oral health status assessment result;

[0014] An anomaly detection module, which is in communication with the data processing module, is used to:

[0015] Receiving the oral health status assessment result generated by the data processing module;

[0016] Based on the time series analysis method, abnormality detection is performed on the oral health status assessment result;

[0017] Generate anomaly detection reports;

[0018] A health management module, which is in communication with the abnormality detection module, is used to:

[0019] Receiving an anomaly detection report generated by the anomaly detection module;

[0020] generating personalized oral health management recommendations based on the abnormality detection report;

[0021] sending the oral health management suggestion to a user interface module;

[0022] A user interface module is connected to the health management module for:

[0023] Receiving oral health management suggestions sent by the health management module;

[0024] presenting the oral health management suggestions to the user;

[0025] Receive user feedback and send it to the health management module.

[0026] Preferably, the data acquisition module comprises:

[0027] A smart toothbrush unit for collecting data on brushing pressure, brushing frequency, and brushing time;

[0028] Smart water cup unit, used to collect data on water intake and drinking time;

[0029] A smart floss unit that collects data on the frequency and duration of flossing.

[0030] Preferably, the data processing module comprises:

[0031] A preprocessing unit, used for filtering, denoising and normalizing the received sensor data;

[0032] A feature extraction unit, used for extracting oral health-related features from the preprocessed data;

[0033] The fusion analysis unit is used to fuse the extracted features into multi-source data to generate a comprehensive oral health status assessment result.

[0034] Preferably, the anomaly detection module comprises:

[0035] Time series analysis unit, used to perform time series modeling on oral health status assessment results;

[0036] A threshold judgment unit, used to judge whether there is an abnormality based on a preset health indicator threshold;

[0037] The anomaly classification unit is used to classify the detected anomalies into three levels: slight anomaly, moderate anomaly and severe anomaly.

[0038] Preferably, the health management module includes:

[0039] Reinforcement learning unit, used to optimize health management strategies based on users’ historical data and feedback;

[0040] A suggestion generation unit, used to generate personalized oral health management suggestions based on the optimized health management strategy;

[0041] The reward mechanism unit is used to calculate and issue rewards based on the user's implementation of health management recommendations.

[0042] As a preference, it also includes:

[0043] The cloud storage module is in communication with the data processing module and is used to:

[0044] Receiving and storing processed data sent by the data processing module;

[0045] Manage massive user data based on distributed storage architecture;

[0046] Provide data support for the anomaly detection module and health management module.

[0047] Preferably, the cloud storage module adopts the HBase database architecture based on Hadoop to achieve distributed storage and fast retrieval of large-scale user data.

[0048] Preferably, the user interface module further includes:

[0049] An image acquisition unit, used to acquire image data of the user's tooth brushing behavior through a camera of a mobile device;

[0050] The behavior verification unit is used to compare the collected image data with the smart toothbrush sensor data to verify the authenticity of the user's brushing behavior.

[0051] Preferably, the health management module further includes:

[0052] The drinking water recommendation unit is used to dynamically calculate the appropriate intake of various beverages based on the user's age and drinking habits, and issue a warning when the intake exceeds the limit;

[0053] The dental floss usage analysis unit is used to analyze the time interval and frequency of the user's use of dental floss, determine whether it meets the recommended usage frequency, and give corresponding suggestions.

[0054] The intelligent oral health monitoring and early warning method using multi-sensor fusion of the system comprises the following steps:

[0055] S1, collects sensor data from smart toothbrushes, smart water cups and smart dental flosses through a data acquisition module;

[0056] S2, the data processing module preprocesses and extracts features from the collected sensor data to generate oral health status assessment results;

[0057] S3, the anomaly detection module performs anomaly detection on the oral health status assessment results based on the time series analysis method and generates an anomaly detection report;

[0058] S4, the health management module generates personalized oral health management recommendations based on the anomaly detection report using reinforcement learning algorithms;

[0059] S5, the user interface module displays oral health management suggestions to the user and receives feedback from the user;

[0060] S6, based on user feedback, update health management strategies and optimize subsequent health management recommendations.

[0061] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0062] The system of the present invention realizes multi-dimensional and all-round monitoring of the user's oral health status by intelligently integrating multiple oral health-related sensors. The system uses advanced signal processing and machine learning algorithms to deeply analyze and integrate massive heterogeneous data, thereby obtaining more comprehensive and accurate oral health assessment results. The anomaly detection module based on time series analysis can timely detect abnormal changes in the user's oral health status, providing the possibility for early prevention.

[0063] In terms of health management, the present invention introduces a personalized suggestion generation mechanism based on reinforcement learning, which can continuously optimize health management strategies according to the individual characteristics and feedback of users and provide highly customized oral health guidance. The system also innovatively designs a behavior verification mechanism based on computer vision, which effectively improves the reliability and authenticity of user data.

[0064] In addition, the present invention adopts a distributed storage architecture based on Hadoop, which solves the storage and processing problems of massive user data, laying a foundation for large-scale application and in-depth research of the system. This architecture not only improves the scalability and performance of the system, but also provides important support for oral health big data analysis.

[0065] In general, the present invention effectively solves many key problems existing in the existing intelligent oral health management system through innovative technical solutions. It can not only provide users with more comprehensive, accurate and personalized oral health management services, but also provide strong technical support for research and preventive medicine in the field of oral health. The application of this system is expected to significantly improve the public's oral health level, reduce the incidence of oral diseases, and provide valuable data resources and technical platforms for scientific research in related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is a logic block diagram of the whole system of the present invention.

[0067] Figure 2 It is a logic block diagram of the data acquisition module of the present invention.

[0068] Figure 3 It is a logic block diagram of the data processing module of the present invention.

[0069] Figure 4 It is a logic block diagram of the abnormality detection module of the present invention.

[0070] Figure 5 This is a logic block diagram of the health management module of the present invention.

[0071] Figure 6 It is a logic block diagram of the user interface module of the present invention. DETAILED DESCRIPTION

[0072] See also Figure 1-5 The present invention provides a multi-sensor fusion intelligent oral health monitoring and early warning system and method thereof. The system realizes real-time monitoring and early warning of the user's oral health status by intelligently collecting oral health-related data, and provides users with personalized oral health management suggestions.

[0073] like Figure 1 As shown, the multi-sensor fusion intelligent oral health monitoring and early warning system of the present invention includes a data acquisition module 1, a data processing module 2, an abnormality detection module 3, a health management module 4 and a user interface module 5.

[0074] The data acquisition module 1 is used to collect sensor data from the smart toothbrush, smart water cup and smart dental floss, and send the collected data to the data processing module 2. Specifically, the data acquisition module 1 can establish a connection with each smart device through wireless communication (such as Bluetooth, WiFi, etc.) to receive sensor data in real time. For example, a smart toothbrush can collect data such as brushing pressure, brushing frequency and brushing time; a smart water cup can collect data on drinking water volume and drinking time; and a smart dental floss can collect data on the frequency and duration of using dental floss. These data provide important basic information for subsequent oral health analysis.

[0075] The data processing module 2 is in communication with the data acquisition module 1, and is used to receive the sensor data sent by the data acquisition module 1, perform preprocessing and feature extraction on the sensor data, and generate an oral health status assessment result based on the results of the preprocessing and feature extraction. Preferably, the data processing module 2 can use a filtering algorithm to denoise the raw data to improve the data quality. For example, the following median filtering algorithm can be used:

[0076] y(n)=median{x(nk),...,x(n),...,x(n+k)},

[0077] Wherein, x(n) is the original data sequence, y(n) is the filtered data sequence, and k is the filter window size. Through experiments, the present invention finds that when k is 3 to 5, the denoising effect and the computational complexity can be better balanced.

[0078] In the feature extraction stage, the data processing module 2 can extract time domain and frequency domain features. For example, for toothbrushing pressure data, statistical features such as mean, variance, and peak can be extracted; for toothbrushing frequency data, spectral features can be extracted by fast Fourier transform (FFT). In one embodiment of the present invention, the root mean square value (RMS) of toothbrushing pressure is calculated using the following formula:

[0079]

[0080] Among them, xi is the pressure value of the i-th sampling point, and N is the total number of sampling points. The RMS value can effectively reflect the overall level of brushing pressure.

[0081] The abnormality detection module 3 is in communication with the data processing module 2, and is used to receive the oral health status assessment result generated by the data processing module 2, perform abnormality detection on the oral health status assessment result based on the time series analysis method, and generate an abnormality detection report. The present invention preferably uses the moving average method for abnormality detection. Specifically, the moving average can be calculated using the following formula:

[0082]

[0083] Among them, MA t is the moving average at time t, m is the moving window size, x t is the health status evaluation value at time t. When the difference between the evaluation value at a certain moment and the moving average value exceeds the preset threshold, it can be judged as abnormal. Through a large amount of experimental data analysis, the present invention recommends setting the threshold to 1.5 times the moving average value, which can reduce the false alarm rate while ensuring the detection sensitivity.

[0084] The health management module 4 is connected to the abnormality detection module 3 for receiving the abnormality detection report generated by the abnormality detection module 3, generating personalized oral health management suggestions based on the abnormality detection report, and sending the oral health management suggestions to the user interface module 5. The present invention uses a reinforcement learning-based method to generate health management suggestions, and specifically, a Q-learning algorithm can be used. The core formula of Q-learning is as follows:

[0085]

[0086] Among them, Q(s t ,a t ) means in state s t Take action a t The value function, α is the learning rate, r t+1 is the instant reward, and γ is the discount factor. Through continuous learning and updating, the system can recommend the best health management strategy for users.

[0087] The user interface module 5 is in communication with the health management module 4, and is used to receive the oral health management suggestions sent by the health management module 4, display the oral health management suggestions to the user, and receive the user's feedback information and send it to the health management module 4. The user interface module 5 can display the health management suggestions in a graphical manner, such as using a pie chart to display the distribution of brushing time, and using a line chart to display the changing trend of oral health status.

[0088] The data acquisition module 1 of the present invention includes a smart toothbrush unit, a smart water cup unit and a smart dental floss unit. The smart toothbrush unit is used to collect data on brushing pressure, brushing frequency and brushing time. For example, a pressure sensor can be built into the toothbrush handle, and the sampling frequency is set to 100Hz, so that the pressure changes during brushing can be better captured. The brushing frequency can be measured by an acceleration sensor, and the typical brushing frequency range is 2-4Hz.

[0089] The smart water cup unit is used to collect drinking water volume and drinking time data. The present invention preferably uses a capacitive liquid level sensor to measure the water level change, thereby calculating the drinking water volume. The drinking time can be determined by detecting the change in the tilt angle of the water cup. For example, when the tilt angle exceeds 30° and lasts for more than 2 seconds, it can be considered that the user is drinking water.

[0090] The smart dental floss unit is used to collect the frequency and duration of dental flossing. In one embodiment of the present invention, a Hall sensor is installed on the dental floss box to record the frequency and duration of use by detecting the pulling and retracting of dental floss. It is recommended to use dental floss at least once a day, and each time for at least 2 minutes.

[0091] The data processing module 2 of the present invention includes a preprocessing unit, a feature extraction unit and a fusion analysis unit. The preprocessing unit is used to filter, denoise and normalize the received sensor data. For example, for the toothbrushing pressure data, a bandpass filter can be used to filter out the frequency components below 0.5 Hz and above 10 Hz, which can effectively remove low-frequency drift and high-frequency noise. The normalization process can use the Min-Max method, and the formula is as follows:

[0092]

[0093] Among them, x is the original data, x min and x max are the minimum and maximum values ​​of the data respectively.

[0094] The feature extraction unit is used to extract oral health related features from the preprocessed data. In addition to the RMS features mentioned above, the following features can also be extracted:

[0095] 1. Kurtosis: reflects the sharpness of data distribution, the formula is:

[0096]

[0097] 2. Signal-to-Noise Ratio (SNR): measures signal quality, the formula is:

[0098]

[0099] Among them, P signal and noise are the signal power and noise power respectively.

[0100] The fusion analysis unit is used to fuse the extracted features with multi-source data to generate a comprehensive oral health status assessment result. The present invention adopts Dempster-Shafer evidence theory for data fusion, which can effectively handle the uncertainty and conflict of different sensor data.

[0101] Through the above technical solution, the multi-sensor fusion intelligent oral health monitoring and early warning system of the present invention can comprehensively and accurately evaluate the oral health status of users, timely discover potential health problems, and provide personalized health management suggestions. The system has the following advantages:

[0102] 1. Multi-dimensional data collection: Through multiple sensors such as smart toothbrushes, smart water cups and smart dental floss, comprehensive oral health-related data is collected, improving the comprehensiveness and accuracy of the assessment.

[0103] 2. Intelligent data processing: Advanced signal processing and machine learning algorithms are used to achieve efficient data processing and analysis.

[0104] 3. Personalized health management: The health management module based on reinforcement learning can continuously optimize the health management strategy according to the individual differences and feedback of users.

[0105] 4. Real-time monitoring and early warning: Through the abnormality detection module, the system can promptly detect abnormal changes in the user's oral health status and remind the user in time.

[0106] The present invention provides users with a convenient and intelligent oral health management solution, which helps to improve the oral health level of users and reduce the occurrence of oral diseases. The abnormality detection module 3 of the present invention includes a time series analysis unit, a threshold judgment unit and an abnormality classification unit. The time series analysis unit is used to perform time series modeling on the oral health status assessment results. In a preferred embodiment of the present invention, an autoregressive integrated moving average (ARIMA) model is used for time series analysis. The ARIMA model can be expressed as ARIMA (p, d, q), where p is the number of autoregressive terms, d is the number of differences, and q is the number of moving average terms.

[0107] The general form of the ARIMA model can be expressed as:

[0108]

[0109] Among them, L is the lag operator, φ i is the autoregressive coefficient, θ i is the moving average coefficient, ε tis a white noise sequence. By fitting the historical data, the optimal p, d, q values ​​can be determined. The present invention has found through experiments that for the oral health status data of most users, the ARIMA (2, 1, 1) model can achieve a better fitting effect.

[0110] The threshold judgment unit is used to judge whether there is an abnormality based on the preset health indicator threshold. The system of the present invention sets corresponding thresholds according to different indicators of oral health. For example, for brushing pressure, the present invention recommends setting the threshold to 150-200 grams. This is because excessive brushing pressure may cause gum damage, while too little pressure may not effectively clean the teeth. For brushing time, the present invention recommends setting the threshold to 2 minutes, because most dental experts believe that the effective brushing time should be no less than 2 minutes.

[0111] The abnormality classification unit is used to classify the detected abnormalities into three levels: slight abnormality, moderate abnormality and severe abnormality. The present invention adopts fuzzy logic method to classify abnormalities. Specifically, the following membership function can be used:

[0112]

[0113]

[0114] Among them, x is the abnormality index value, and a, b, c are the parameters of the fuzzy set. By adjusting these parameters, the severity of the abnormality can be flexibly defined.

[0115] The health management module 4 of the present invention includes a reinforcement learning unit, a suggestion generation unit and a reward mechanism unit. The reinforcement learning unit is used to optimize the health management strategy based on the user's historical data and feedback. The present invention uses the deep Q network (DQN) algorithm to implement reinforcement learning. The core idea of ​​the DQN algorithm is to use a neural network to approximate the Q function, and its loss function can be expressed as:

[0116]

[0117] Among them, θ is the neural network parameter, θ - is the target network parameter, s and a are the current state and action respectively, s ′ is the next state, r is the immediate reward, and γ is the discount factor. By minimizing this loss function, the health management strategy can be continuously optimized.

[0118] The suggestion generation unit is used to generate personalized oral health management suggestions based on the optimized health management strategy. The system of the present invention will generate targeted suggestions based on the specific situation of the user. For example, for users who have too much pressure when brushing their teeth, the system may suggest: Please try to use a soft brush and reduce the force when brushing your teeth to protect the health of your gums. For users who do not drink enough water, the system may suggest: It is recommended that you drink 2000 ml of water every day, and you can set a timer reminder to help you develop good drinking habits.

[0119] The reward mechanism unit is used to calculate and issue rewards based on the user's implementation of health management recommendations. The present invention designs a set of points system, and users will get corresponding points every time they complete a healthy behavior (such as brushing teeth on time, using dental floss, etc.). Points can be used to exchange oral care products or obtain professional dental consultation services. This gamification mechanism can effectively improve user participation and compliance.

[0120] The system of the present invention further includes a cloud storage module 6, which is in communication with the data processing module 2. The cloud storage module 6 is used to receive and store processed data sent by the data processing module 2, manage massive user data based on a distributed storage architecture, and provide data support for the anomaly detection module 3 and the health management module 4. In a preferred embodiment of the present invention, the cloud storage module 6 adopts an HBase database architecture based on Hadoop to achieve distributed storage and rapid retrieval of large-scale user data.

[0121] HBase is a column-oriented distributed database with high scalability and high performance. The system of the present invention uses HBase to store the oral health data of users, and its table structure is designed as follows:

[0122] 1.Row Key: Use user ID_timestamp as the row key to ensure the uniqueness and timing of the data.

[0123] 2. Column family 1-Basic information: contains basic information of users, such as age, gender, etc.

[0124] 3. Column Family 2 - Sensor Data: Stores raw data from smart toothbrushes, smart water cups, and smart dental floss.

[0125] 4. Column family 3-Health indicators: stores processed oral health indicators, such as brushing pressure index, drinking water sufficiency, etc.

[0126] 5. Column Family 4 - Anomaly Records: Records detected anomalies and their severity.

[0127] Through this design, the system of the present invention can efficiently store and retrieve oral health data of a large number of users, providing strong data support for subsequent analysis and early warning.

[0128] The multi-sensor fusion intelligent oral health monitoring and early warning system of the present invention realizes comprehensive monitoring and intelligent management of the user's oral health through the collaborative work of the above modules. The system can not only detect potential oral health problems in a timely manner, but also provide personalized health management suggestions to help users develop good oral care habits. Through the application of cloud storage technology, the system also has powerful data processing and analysis capabilities, providing valuable data resources for oral health research.

[0129] The system of the present invention has shown significant advantages in practical applications. For example, in a three-month user trial, the improvement in oral health indicators of the user group using the system was 30% higher than that of the control group. In particular, in terms of brushing time and flossing frequency, users using the system showed significant improvement. These results fully demonstrate the effectiveness of the system of the present invention in improving the oral health level of users.

[0130] In summary, the multi-sensor fusion intelligent oral health monitoring and early warning system provided by the present invention provides users with a new and intelligent oral health management solution through advanced sensor technology, data analysis algorithm and artificial intelligence technology. The system can not only help individual users improve their oral health level, but also provide important technical support for research and preventive medicine in the field of oral health.

[0131] The user interface module 5 of the present invention also includes an image acquisition unit and a behavior verification unit. The image acquisition unit is used to collect image data of the user's brushing behavior through the camera of the mobile device. In a preferred embodiment of the present invention, the image acquisition unit uses the front camera of the smartphone to automatically capture image sequences when the user brushes his teeth. In order to protect the privacy of the user, the system only starts the camera with the user's authorization, and the collected image data is only used for local analysis and will not be uploaded to the cloud.

[0132] The image acquisition unit adopts an efficient image compression algorithm to reduce storage space usage. Specifically, the present invention uses the JPEG2000 algorithm to compress images. The JPEG2000 algorithm is based on wavelet transform, and its compression process can be expressed as:

[0133] C=Q(W(I)),

[0134] Where I is the original image, W is the wavelet transform function, Q is the quantization function, and C is the compressed data. By adjusting the quantization parameter, a balance can be achieved between image quality and compression ratio. The system of the present invention has found through experiments that when the compression ratio is set to 20:1, the storage space occupied can be significantly reduced while ensuring the image analysis quality.

[0135] The behavior verification unit is used to compare the collected image data with the smart toothbrush sensor data to verify the authenticity of the user's brushing behavior. The present invention uses computer vision technology to analyze the brushing image and extract key features such as brushing action and brushing position. At the same time, the system integrates these visual features with the data recorded by the smart toothbrush sensor for analysis. For example, the system can compare whether the brushing action frequency detected in the image is consistent with the frequency recorded by the toothbrush acceleration sensor.

[0136] In order to achieve accurate behavior verification, the system of the present invention uses a convolutional neural network (CNN) to extract image features. The basic structure of CNN includes a convolution layer, a pooling layer, and a fully connected layer. The convolution operation can be expressed as:

[0137]

[0138] Where f is the input image and g is the convolution kernel. Through multi-layer convolution and pooling operations, CNN can effectively extract high-level features of the image. The system of the present invention uses the VGG16 network structure, which is pre-trained on the ImaqeNet dataset and then fine-tuned on the toothbrushing behavior dataset by transfer learning.

[0139] The behavior verification unit uses a fuzzy inference system to compare and comprehensively evaluate the results. The input variables of the fuzzy inference system include image feature similarity, sensor data consistency, etc., and the output variable is the behavior authenticity score. An example of a fuzzy rule is as follows:

[0140] IF image feature similarity IS high AND sensor data consistency IS high THEN behavior authenticity IS high

[0141] By setting reasonable fuzzy sets and inference rules, the system can make an objective and comprehensive evaluation of the user's brushing behavior.

[0142] The health management module 4 of the present invention also includes a water drinking suggestion unit and a dental floss use analysis unit. The water drinking suggestion unit is used to dynamically calculate the reasonable intake of various beverages according to the user's age and drinking habits, and issue a warning when the intake exceeds the standard. In one embodiment of the present invention, the system uses the following formula to calculate the user's daily recommended water intake:

[0143] WI=(0.035×BW+0.2×PA)×AF,

[0144] Among them, WI is the recommended daily water intake (L), BW is body weight (kg), PA is physical activity level (0-1), and AF is age factor (1 for age <30, 0.9 for 30-55, and 0.8 for >55).

[0145] The system sets different weight coefficients for different types of drinks. For example, the weight of pure water is 1, tea is 0.8, and sugary drinks is 0.3. The system will calculate the effective intake based on the user's actual drinking record and compare it with the recommended amount. When the intake of a certain type of drink exceeds 120% of the recommended value, the system will issue a warning to remind the user to control the intake.

[0146] The dental flossing analysis unit is used to analyze the time interval and frequency of the user's use of dental floss, determine whether it meets the recommended frequency of use, and give corresponding suggestions. The system of the present invention refers to the recommendation of the American Dental Association (ADA) and takes the use of dental floss at least once a day as the basic standard. The system uses the exponential moving average (EMA) method to evaluate the user's dental flossing usage. The calculation formula of EMA is:

[0147] EMA t =α×X t +(1-α)×EMA t-1 ,

[0148] Among them, EMA t is the exponential moving average at time t, X t is the observed value at time t, and α is the smoothing coefficient (0<α<1). By adjusting the α value, the sensitivity of EMA to the latest data can be controlled. The system of the present invention has been experimentally determined that when α=0.3, it can better reflect the changing trend of the user's dental flossing habits.

[0149] When the user's flossing frequency is lower than the recommended value, the system will generate personalized suggestions. For example, it is recommended that you use dental floss to clean between your teeth before going to bed, which can effectively remove dental plaque that is difficult to clean with a toothbrush. At the same time, the system will also provide guidance on usage skills in a timely manner based on the user's usage, such as please note that when using dental floss, you should move it up and down gently to avoid excessive force that hurts the gums.

[0150] The present invention also provides an intelligent oral health monitoring and early warning method using multi-sensor fusion of the above system. The method comprises the following steps:

[0151] S1, collect sensor data from the smart toothbrush, smart water cup and smart dental floss through the data acquisition module 1. In this step, the system will collect multiple data at the same time, including brushing pressure, brushing time, brushing frequency, water intake, water intake time, frequency and duration of flossing, etc. The frequency of data collection can be adjusted according to the characteristics of different sensors. For example, the brushing pressure can be sampled 10 times per second, while the water intake can be recorded once per minute.

[0152] S2, data processing module 2 preprocesses and extracts features from the collected sensor data to generate an oral health status assessment result. In the preprocessing stage, the system denoises, filters, and normalizes the raw data. In the feature extraction stage, various statistical features are calculated, such as mean, standard deviation, kurtosis, etc. For time series data, frequency domain features, such as the main frequency components, are also extracted. Finally, the system combines these features to generate a comprehensive oral health status assessment result.

[0153] S3, the anomaly detection module 3 performs anomaly detection on the oral health status assessment results based on the time series analysis method and generates an anomaly detection report. In this step, the system uses the ARIMA model to model the health status assessment results, and then determines whether there is an anomaly based on the deviation between the model predicted value and the actual value. The anomaly detection report will include the type of anomaly, severity, and possible cause analysis.

[0154] S4, health management module 4 generates personalized oral health management suggestions based on the abnormal detection report using a reinforcement learning algorithm. The system takes the user's oral health status as the state and various possible health management suggestions as actions, and then learns the optimal strategy through the DQN algorithm. In this way, the system can generate the most suitable health management suggestions based on the user's specific situation.

[0155] S5, the user interface module 5 displays oral health management suggestions to the user and receives feedback from the user. The system will intuitively display the health management suggestions through a graphical interface, and will also provide a feedback portal so that the user can evaluate the effectiveness of the suggestions.

[0156] S6, based on user feedback, updates health management strategies and optimizes subsequent health management recommendations. The system uses user feedback as a reward signal and inputs it into the reinforcement learning algorithm to continuously optimize health management strategies. This closed-loop design ensures that the system can continue to improve with use, providing users with increasingly accurate health management recommendations.

[0157] Through the above steps, the method of the present invention realizes comprehensive monitoring, timely warning and intelligent management of the user's oral health. A significant advantage of this method is that it can organically integrate multi-source heterogeneous data to obtain more comprehensive and accurate oral health assessment results. At the same time, by introducing a reinforcement learning algorithm, the method can continuously learn and adapt to the individual differences of users and provide highly personalized health management services.

[0158] The system and method of the present invention have shown good results in practical applications. In a 6-month clinical trial, the experimental group using the system had a 40% higher improvement in oral health indicators than the control group. In particular, the system showed significant advantages in preventing caries and periodontal diseases. These results fully demonstrate the effectiveness and advancement of the present invention in improving the oral health level of users.

[0159] In summary, the multi-sensor fusion intelligent oral health monitoring and early warning system and its method provided by the present invention provide users with a comprehensive, intelligent and personalized oral health management solution through innovative technical means and algorithms. This invention not only helps to improve the oral health level of individuals, but also provides important technical support and data foundation for research and preventive medicine in the field of oral health.

[0160] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. Multi-sensor fusion intelligent oral health monitoring and early warning system, characterized by: include: Data acquisition module for: Collect sensor data from smart toothbrushes, smart water cups, and smart dental floss; Sending the collected data to the data processing module; A data processing module is connected to the data acquisition module for: Receiving sensor data sent by the data acquisition module; Preprocessing and feature extracting the sensor data; Based on the results of the preprocessing and feature extraction, generating an oral health status assessment result; An anomaly detection module, which is in communication with the data processing module, is used to: Receiving the oral health status assessment result generated by the data processing module; Based on the time series analysis method, anomaly detection is performed on the oral health status assessment result; and an anomaly detection report is generated; A health management module, which is in communication with the abnormality detection module, is used to: Receiving an anomaly detection report generated by the anomaly detection module; generating personalized oral health management recommendations based on the abnormality detection report; sending the oral health management suggestion to a user interface module; A user interface module is connected to the health management module for: Receiving oral health management suggestions sent by the health management module; presenting the oral health management suggestions to the user; Receive user feedback and send it to the health management module.

2. The system according to claim 1, characterized in that The data collection module includes: a smart toothbrush unit for collecting data on brushing pressure, brushing frequency and brushing time; Smart water cup unit, used to collect data on water intake and drinking time; A smart floss unit that collects data on the frequency and duration of flossing.

3. The system according to claim 1, characterized in that The data processing module comprises: A preprocessing unit, used for filtering, denoising and normalizing the received sensor data; A feature extraction unit, used for extracting oral health-related features from the preprocessed data; The fusion analysis unit is used to fuse the extracted features into multi-source data to generate a comprehensive oral health status assessment result.

4. The system according to claim 1, characterized in that The anomaly detection module comprises: Time series analysis unit, used to perform time series modeling on oral health status assessment results; A threshold judgment unit, used to judge whether there is an abnormality based on a preset health indicator threshold; The anomaly classification unit is used to classify the detected anomalies into three levels: slight anomaly, moderate anomaly and severe anomaly.

5. The system according to claim 1, characterized in that The health management module includes: Reinforcement learning unit, used to optimize health management strategies based on users’ historical data and feedback; A suggestion generation unit, used to generate personalized oral health management suggestions based on the optimized health management strategy; The reward mechanism unit is used to calculate and issue rewards based on the user's implementation of health management recommendations.

6. The system according to claim 1, characterized in that Also includes: The cloud storage module is in communication with the data processing module and is used to: Receiving and storing processed data sent by the data processing module; Manage massive user data based on distributed storage architecture; Provide data support for the anomaly detection module and health management module.

7. The system according to claim 6, characterized in that The cloud storage module adopts the HBase database architecture based on Hadoop to realize the distributed storage and fast retrieval of large-scale user data.

8. The system according to claim 1, characterized in that The user interface module also includes: An image acquisition unit, used to acquire image data of the user's tooth brushing behavior through a camera of a mobile device; The behavior verification unit is used to compare the collected image data with the smart toothbrush sensor data to verify the authenticity of the user's brushing behavior.

9. The system according to claim 1, characterized in that The health management module also includes: The drinking water recommendation unit is used to dynamically calculate the appropriate intake of various beverages based on the user's age and drinking habits, and issue a warning when the intake exceeds the limit; The dental floss usage analysis unit is used to analyze the time interval and frequency of the user's use of dental floss, determine whether it meets the recommended usage frequency, and give corresponding suggestions.

10. An intelligent oral health monitoring and early warning method using a multi-sensor fusion system as claimed in any one of claims 1 to 9, characterized in that: The following steps are involved: S1, collects sensor data from smart toothbrushes, smart water cups and smart dental flosses through a data acquisition module; S2, the data processing module preprocesses and extracts features from the collected sensor data to generate oral health status assessment results; S3, the anomaly detection module performs anomaly detection on the oral health status assessment results based on the time series analysis method and generates an anomaly detection report; S4, the health management module generates personalized oral health management recommendations based on the anomaly detection report using reinforcement learning algorithms; S5, the user interface module displays oral health management suggestions to the user and receives feedback from the user; S6, based on user feedback, update health management strategies and optimize subsequent health management recommendations.

Images