Evaluation method and system based on intelligent electric toothbrush

Abstract

The invention provides an evaluation method and system based on an intelligent electric toothbrush. According to the technical scheme, the evaluation method comprises the steps that S1, a fluorescence excitation device, an image obtaining device and a sensing device of the intelligent electric toothbrush are arranged; s2, performing tooth fluorescence activation according to a fluorescence excitation device, acquiring a tooth fluorescence image of the teeth according to an image acquisition device, and acquiring a brush head movement track of the intelligent electric toothbrush according to a sensing device; s3, obtaining a fluorescence image data set according to the tooth fluorescence image, and dividing the fluorescence image data set into a training set, a verification set and a test set; s4, establishing a tooth evaluation model of the intelligent electric toothbrush; s41, performing model training evaluation on the tooth evaluation model according to the training set, the verification set and the test set to obtain target model parameters of the tooth evaluation model and a target tooth evaluation model; and S5, performing dental plaque detection and evaluation on the teeth according to the intelligent electric toothbrush to obtain a dental plaque evaluation report of the teeth. According to the invention, the dental plaque detection precision of teeth is obviously improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent toothbrushes, and particularly relates to an evaluation method and system based on an intelligent electric toothbrush. Background Art

[0002] At present, oral health problems are widespread, and dental plaque detection mostly relies on professional equipment and personnel, with high costs and difficulty in popularization. Moreover, many dental plaque detection methods rely on traditional laser fluorescence technology, quantitative light-induced fluorescence (QLF) technology, and digital imaging technology. These technologies irradiate the tooth surface with specific light sources and use fluorescence reactions to evaluate the accumulation of dental plaque. However, these methods are insufficient in accuracy and sensitivity when dealing with complex oral images or early dental plaque. For example, laser fluorescence devices are very sensitive to foreign objects (such as dental materials) in the oral environment, which may lead to misdiagnosis.

[0003] This application provides an evaluation method and system based on an intelligent electric toothbrush. The intelligent electric toothbrush combines intelligent image processing technology and a real-time feedback system, can accurately identify the morphology and distribution range of dental plaque, and significantly improve the accuracy of dental plaque detection. This method enables users to more frequently monitor their oral health status, promptly discover problems of dental plaque accumulation, and thus has important significance for promoting early prevention and intervention of oral diseases, as well as enhancing the awareness and management level of national oral health. Summary of the Invention

[0004] In the first aspect of this application, an evaluation method based on an intelligent electric toothbrush is provided, including the following steps:

[0005] S1: Obtain the fluorescence wavelength data of dental plaque, and set the fluorescence excitation device of the intelligent electric toothbrush according to the fluorescence wavelength data; set the image acquisition device and sensing device of the intelligent electric toothbrush.

[0006] S2: Activate tooth fluorescence according to the fluorescence excitation device, obtain the tooth fluorescence image of the teeth according to the image acquisition device, and obtain the movement trajectory of the brush head of the intelligent electric toothbrush according to the sensing device.

[0007] S3: Obtain a fluorescence image dataset from the tooth fluorescence image; perform image processing and image region segmentation on the tooth fluorescence image, perform region segmentation on the tooth fluorescence image to obtain a region segmentation result; perform region annotation on the tooth fluorescence images in the fluorescence image dataset according to the region segmentation result, and divide the fluorescence image dataset into a training set, a validation set, and a test set.

[0008] S4: Establish a tooth evaluation model for the intelligent electric toothbrush.

[0009] S41: Train the tooth evaluation model based on the training set and the validation set, and evaluate the tooth evaluation model based on the test set to obtain the target model parameters of the tooth evaluation model; obtain the target tooth evaluation model based on the target model parameters;

[0010] S5: Detect and evaluate dental plaque on teeth based on the intelligent electric toothbrush to obtain a dental plaque evaluation report for the teeth.

[0011] Preferably, S1 is specifically:

[0012] The fluorescence excitation device includes a fluorescence excitation light source and a filter system, where the filter system includes an excitation filter and an emission filter;

[0013] Activate the fluorescence of dental plaque on teeth based on the fluorescence excitation light source to obtain a tooth fluorescence signal;

[0014] And filter the tooth fluorescence signal based on the excitation filter and the emission filter.

[0015] Preferably, S2 is specifically:

[0016] Perform image correction and enhancement on the tooth fluorescence image to obtain a target fluorescence region;

[0017] Segment the target fluorescence region based on the brush head movement trajectory, and divide the target fluorescence region into a dental plaque region and a normal region;

[0018] Quantify the area and concentration of the dental plaque region to obtain dental plaque region parameters, where the dental plaque region parameters include the dental plaque area and dental plaque concentration corresponding to the dental plaque region.

[0019] Preferably, S3 is specifically:

[0020] Label the normal region of the tooth fluorescence image in the fluorescence image dataset as the normal region;

[0021] Label the dental plaque region of the tooth fluorescence image in the fluorescence image dataset as the caries region, and perform region parameter labeling on the caries region based on the dental plaque region parameters corresponding to the dental plaque region;

[0022] Perform data enhancement processing on all the tooth fluorescence images in the fluorescence image dataset to obtain the fluorescence images data to be divided;

[0023] Divide the fluorescence images data to be divided into a training set, a validation set, and a test set.

[0024] Preferably, S4 is specifically:

[0025] A tooth evaluation model is established based on a convolutional neural network architecture. The tooth evaluation model includes an input end, an analysis end, and an output end, and the tooth evaluation model is provided with a channel attention mechanism and a spatial attention mechanism;

[0026] The input end is used to input tooth fluorescence images;

[0027] The analysis end is provided with a semantic segmentation network, which is used to perform image detection on the tooth fluorescence image according to the model parameters to obtain the spatial distribution characteristics of dental plaque in the tooth, and perform semantic recognition on the spatial distribution characteristics according to the semantic segmentation network to obtain an image recognition result;

[0028] The image recognition result includes the number of dental plaques, the area of dental plaques, and the position of dental plaques on the tooth;

[0029] The output end is used to output the image recognition result of the analysis end.

[0030] Preferably, S41 is specifically:

[0031] Update the model parameters of the tooth evaluation model according to the training set;

[0032] Verify the model parameters of the tooth evaluation model according to the validation set;

[0033] Evaluate the model parameters of the tooth evaluation model according to the test set to obtain the target model parameters of the tooth evaluation model.

[0034] Preferably, evaluating the model parameters of the convolutional neural network model according to the test set to obtain the target model parameters of the convolutional neural network is specifically:

[0035] Test the corresponding model parameters of the tooth evaluation model according to the test set to obtain the performance index evaluation result, the visual analysis result, and the model tuning result of the corresponding model parameters of the tooth evaluation model;

[0036] Obtain the target model parameters of the tooth evaluation model according to the performance index evaluation result, the visual analysis result, and the model tuning result.

[0037] Preferably, S5 is specifically:

[0038] Constitute a dental plaque evaluation report based on the number of dental plaques, the area of dental plaques, and the position of dental plaques on the tooth.

[0039] In the second aspect of the present application, an intelligent electric toothbrush evaluation system is provided. This system is applied to an intelligent electric toothbrush evaluation method. The system includes:

[0040] Toothbrush configuration module: Set the fluorescence wavelength data of dental plaque, and obtain the fluorescence excitation device of the intelligent electric toothbrush according to the fluorescence wavelength data; Set the image acquisition device and sensing device of the intelligent electric toothbrush;

[0041] Tooth data acquisition module: Activate tooth fluorescence according to the fluorescence excitation device, obtain the tooth fluorescence image of the teeth according to the image acquisition device, and obtain the brush head movement trajectory of the intelligent electric toothbrush according to the sensing device;

[0042] Data analysis module: Obtain a fluorescence image dataset according to the tooth fluorescence image; Perform image processing and image region segmentation on the tooth fluorescence image, and perform region segmentation on the tooth fluorescence image to obtain a region segmentation result; Perform region annotation on the tooth fluorescence images in the fluorescence image dataset according to the region segmentation result, and divide the fluorescence image dataset into a training set, a validation set, and a test set;

[0043] Model setting module: Establish a tooth evaluation model for the intelligent electric toothbrush;

[0044] Train the tooth evaluation model according to the training set and the validation set, and evaluate the tooth evaluation model according to the test set to obtain the target model parameters of the tooth evaluation model; Obtain the target tooth evaluation model according to the target model parameters;

[0045] Tooth evaluation module: Detect and evaluate dental plaque on teeth with the intelligent electric toothbrush to obtain a dental plaque evaluation report for the teeth.

[0046] In summary, the beneficial effects of this application are:

[0047] 1. In this application, by setting the fluorescence wavelength data of dental plaque, the fluorescence excitation device of the intelligent electric toothbrush is obtained according to the fluorescence wavelength data, and the image acquisition device and sensing device of the intelligent electric toothbrush are set; Activate tooth fluorescence according to the fluorescence excitation device, obtain the tooth fluorescence image of the teeth according to the image acquisition device, and obtain the brush head movement trajectory of the intelligent electric toothbrush according to the sensing device; The fluorescence excitation device includes a fluorescence excitation light source and a filter system. Through the fluorescence excitation light source, the fluorescence signals of dental plaque and healthy tooth tissue can be significantly distinguished, ensuring the sensitivity and specificity of detection; Filter unnecessary wavelengths through the excitation filter in the filter system, block the excitation light through the emission filter in the filter system, and at the same time allow the fluorescence wavelength data to pass through; In addition, the image acquisition device uses a 12 million pixel CMOS image sensor, and the sensing device uses a piezoelectric film sensor. By combining the use of the piezoelectric film sensor and the piezoelectric film sensor, the biting force and movement trajectory of the teeth are monitored in real time to further optimize the acquisition and analysis of tooth fluorescence images.

[0048] 2. This application obtains a fluorescence image dataset based on dental fluorescence images; performs image processing and image region segmentation on the dental fluorescence images, and performs region segmentation on the dental fluorescence images to obtain a region segmentation result; performs region annotation on the dental fluorescence images in the fluorescence image dataset according to the region segmentation result, and divides the fluorescence image dataset into a training set, a validation set, and a test set; by combining a deep learning segmentation algorithm and traditional image processing techniques, and combining the motion trajectory data of the sensor device, realizes fine segmentation of the dental plaque region, and at the same time uses the segmented data for area and concentration quantification to ensure the accuracy and reliability of the analysis.

[0049] 3. This application establishes a dental evaluation model for an intelligent electric toothbrush, and the dental evaluation model is constructed by a convolutional neural network. The dental evaluation model is trained according to the training set and the validation set, and the dental evaluation model is verified according to the test set to obtain the target model parameters of the dental evaluation model. According to the target model parameters, a target dental evaluation model is obtained. Finally, the target dental evaluation model is used to detect and evaluate dental plaque on teeth to obtain a dental plaque evaluation report for the teeth. This application trains by performing large-scale annotation of fluorescence images on the dental evaluation model, automatically extracts the spatial distribution characteristics of the plaque, and combines a semantic segmentation network to accurately identify the morphology and distribution range of the plaque, with high precision and automation characteristics, significantly improving the detection efficiency and reliability, and at the same time having good scalability to meet the clinical rapid detection needs. Brief Description of the Drawings

[0050] In order to more clearly illustrate the technical solutions of the embodiments of this application, the following will briefly describe some of the drawings in the embodiments of this application. It should be understood that the following drawings only show some embodiments of this application, and therefore should not be considered as limiting the scope of this application.

[0051] Figure 1 It is a schematic flowchart of an evaluation method based on an intelligent electric toothbrush provided by the present invention;

[0052] Figure 2 It is a schematic structural diagram of an evaluation system based on an intelligent electric toothbrush provided by the present invention. Detailed Embodiments

[0053] The following combines the embodiments and Figure 1 and Figure 2 to further elaborate on this application in detail, but the implementation manners of this application are not limited thereto.

[0054] Refer to Figure 1As shown in the figure, it is a schematic flowchart of a method for evaluating an intelligent electric toothbrush provided by an embodiment of the present application. In the embodiment of the present application, the applied scenario is that the present application designs a portable and easy-to-operate intelligent electric toothbrush, and through the intelligent electric toothbrush, the user can detect dental plaque on the teeth at home, enabling the user to monitor the oral health status more frequently, promptly discover the problem of dental plaque accumulation, thereby promoting the early prevention and early intervention of oral diseases, enhancing the national awareness of oral health and management level. Moreover, the intelligent electric toothbrush in the present application combines intelligent image processing technology and a real-time feedback system, which can accurately identify the morphology and distribution range of dental plaque, significantly improving the detection accuracy of dental plaque on teeth.

[0055] A method for evaluating an intelligent electric toothbrush includes the following steps:

[0056] S1: Obtain the fluorescence wavelength data of dental plaque, and set the fluorescence excitation device of the intelligent electric toothbrush according to the fluorescence wavelength data; set the image acquisition device and the sensing device of the intelligent electric toothbrush.

[0057] S2: Activate the tooth fluorescence according to the fluorescence excitation device, obtain the tooth fluorescence image of the teeth according to the image acquisition device, and obtain the brush head movement trajectory of the intelligent electric toothbrush according to the sensing device.

[0058] S3: Obtain a fluorescence image dataset from the tooth fluorescence image; perform image processing and image region segmentation on the tooth fluorescence image, and perform region segmentation on the tooth fluorescence image to obtain a region segmentation result; perform region annotation on the tooth fluorescence images in the fluorescence image dataset according to the region segmentation result, and divide the fluorescence image dataset into a training set, a validation set, and a test set.

[0059] S4: Establish a tooth evaluation model for the intelligent electric toothbrush.

[0060] S41: Train the tooth evaluation model according to the training set and the validation set, and evaluate the tooth evaluation model according to the test set to obtain the target model parameters of the tooth evaluation model; obtain the target tooth evaluation model according to the target model parameters.

[0061] S5: Detect and evaluate the dental plaque on the teeth according to the intelligent electric toothbrush to obtain a dental plaque evaluation report for the teeth.

[0062] In some embodiments, during the use of the intelligent electric toothbrush in the present application, the dental plaque on the user's teeth is detected according to the fluorescence excitation device, the image acquisition device, the sensing device, and the tooth evaluation model, and a dental plaque evaluation report for the user's teeth is obtained according to the dental plaque detection result.

[0063] S1 is specifically:

[0064] The fluorescence excitation device includes a fluorescence excitation light source and a filter system, where the filter system includes an excitation filter and an emission filter;

[0065] The dental plaque on the tooth is fluorescence-activated by the fluorescence excitation light source to obtain a tooth fluorescence signal;

[0066] And the tooth fluorescence signal is filtered according to the excitation filter and the emission filter.

[0067] In some embodiments, the fluorescence wavelength data refers to the fluorescence wavelength data range of dental plaque. According to the characteristics that dental plaque may produce green or red fluorescence under ultraviolet or blue light excitation, through literature review and experimental measurement, the fluorescence wavelength data range is determined to ensure that the specificity and intensity of the fluorescence signal can meet the detection requirements;

[0068] According to the fluorescence wavelength data range, a suitable fluorescence excitation light source is selected. The light source needs to have high intensity, narrow bandwidth and stability, and can accurately excite the target fluorescence; the excitation filter is used to filter unnecessary wavelengths to ensure that the fluorescence light source only outputs the dental plaque excitation wavelength; the emission filter is used to block the excitation light and allow the fluorescence wavelength data to pass through at the same time; in practical applications, a 405nm violet LED is selected as the fluorescence excitation light source. This wavelength can significantly distinguish the fluorescence signals of dental plaque and healthy tooth tissue, while reducing background light interference. Combined with an efficient filter system, it optimizes the capture efficiency of the fluorescence signal to ensure the sensitivity and specificity of the detection;

[0069] In addition, in practical applications, the image acquisition device uses a 12-megapixel CMOS image sensor, combined with a high-resolution optical lens, to ensure the clarity of the fluorescence image and the ability to capture details; the sensing device uses a piezoelectric film sensor. By combining the use of a piezoelectric film sensor and a MEMS sensor, the biting force and movement trajectory of the tooth are monitored in real time to further optimize the acquisition and analysis of the fluorescence signal, ensuring that all tooth surfaces can be accurately covered during the imaging process.

[0070] Specifically, S2 is as follows:

[0071] The tooth fluorescence image is corrected and enhanced to obtain the target fluorescence region;

[0072] The target fluorescence region is segmented according to the movement trajectory of the toothbrush head, and the target fluorescence region is divided into a dental plaque region and a normal region;

[0073] The area and concentration of the dental plaque region are quantified to obtain dental plaque region parameters, where the dental plaque region parameters include the dental plaque area and dental plaque concentration corresponding to the dental plaque region.

[0074] In some embodiments, the image correction and enhancement refer to image correction and image enhancement of the tooth fluorescence image;

[0075] In the process of image correction for dental fluorescence images, first, background noise is removed from the original dental fluorescence images, and the fluorescence intensity is normalized to balance the signal strength differences between different samples, thereby improving the accuracy and stability of subsequent analysis. Combining with sensor data, the inconsistencies in the images are further corrected to ensure the accuracy and consistency of the images;

[0076] In the process of image enhancement for dental fluorescence images, Gaussian filtering is used to remove random noise, while enhancing the image edges and contrast to make the fluorescence regions clearer. Combining with the motion trajectory data, the image enhancement process is further optimized to ensure the clear separation of the target fluorescence regions from the background regions and improve the image quality.

[0077] In some embodiments, by extracting the dental plaque regions, the target fluorescence regions are divided into dental plaque regions and normal regions. After segmentation, the region boundaries of the dental plaque regions are smoothed and isolated small regions are removed to ensure the accuracy of the segmentation results and compliance with biological significance; in addition, by quantifying the area and concentration of the dental plaque, the accuracy and reliability of the analysis results are ensured; among them, region division and quantification can be realized through region segmentation algorithms and quantification algorithms respectively, and the region segmentation algorithms and quantification algorithms are preset according to the actual detection requirements and actual application scenarios of dental plaque detection.

[0078] Specifically, S3 is as follows:

[0079] Label the normal regions of the dental fluorescence images in the fluorescence image dataset as normal regions;

[0080] Label the dental plaque regions of the dental fluorescence images in the fluorescence image dataset as carious regions, and perform region parameter labeling on the carious regions according to the dental plaque region parameters corresponding to the dental plaque regions;

[0081] Perform data augmentation processing on all the dental fluorescence images in the fluorescence image dataset to obtain the fluorescence images to be divided;

[0082] Divide the fluorescence images to be divided into a training set, a validation set, and a test set.

[0083] In some embodiments, performing data augmentation processing on all the dental fluorescence images in the fluorescence image dataset means performing data augmentation operations such as random rotation, scaling, flipping, and illumination adjustment on the dental fluorescence images; in the process of dividing the fluorescence image dataset, the data distributions of the training set, the validation set, and the test set are balanced and cover diversity. For example, the model division rule can be set as the stratified sampling rule or the division rule based on clustering, etc. to achieve this.

[0084] Specifically, S4 is as follows:

[0085] A tooth evaluation model is established based on a convolutional neural network architecture. The tooth evaluation model includes an input end, an analysis end, and an output end, and the tooth evaluation model is provided with a channel attention mechanism and a spatial attention mechanism;

[0086] The input end is used to input tooth fluorescence images;

[0087] The analysis end is provided with a semantic segmentation network, which is used to perform image detection on the tooth fluorescence image according to the model parameters to obtain the spatial distribution characteristics of dental plaque in the tooth, and perform semantic recognition on the spatial distribution characteristics according to the semantic segmentation network to obtain an image recognition result;

[0088] The image recognition result includes the number of dental plaque, the area of dental plaque, and the location of dental plaque on the tooth;

[0089] The output end is used to output the image recognition result of the analysis end.

[0090] In some embodiments, the tooth evaluation model adopts a convolutional neural network architecture and is trained with a large amount of labeled data, which can automatically extract the morphological, distribution, and location characteristics of dental plaque, and combine with a semantic segmentation network to achieve high-precision recognition and quantitative analysis of dental plaque; in addition, the tooth evaluation model also introduces a channel attention mechanism and a spatial attention mechanism to further improve the adaptability to complex backgrounds and the detection accuracy.

[0091] Specifically, S41 is as follows:

[0092] Update the model parameters of the tooth evaluation model according to the training set;

[0093] Verify the model parameters of the tooth evaluation model according to the validation set;

[0094] Evaluate the model parameters of the tooth evaluation model according to the test set to obtain the target model parameters of the tooth evaluation model.

[0095] In some embodiments, the tooth evaluation model training extracts features and generates prediction results through forward propagation. After calculating the error using the loss function, the weights are adjusted through backpropagation to optimize the model parameters. Among them, the training set is used to update the parameters, and the validation set is used to evaluate the performance to ensure that the model has good performance on both the training data and new data; in addition, after the model training is completed, an independent test set is used to verify the performance of the model, obtain the prediction results and compare them with the true labels to statistically analyze the performance of the model in actual applications.

[0096] Evaluate the model parameters of the convolutional neural network model according to the test set to obtain the target model parameters of the convolutional neural network. Specifically:

[0097] Test the model parameters corresponding to the tooth evaluation model based on the test set to obtain the performance index evaluation results, visual analysis results, and model tuning results of the model parameters corresponding to the tooth evaluation model;

[0098] Obtain the target model parameters of the tooth evaluation model based on the performance index evaluation results, visual analysis results, and model tuning results.

[0099] In some embodiments, the performance index evaluation results are to comprehensively evaluate the performance of the model using multiple performance indexes, including the correct prediction ratio of the overall model, i.e., accuracy, the ability of the model to correctly detect caries regions, i.e., sensitivity, and the ability of the model to correctly identify normal regions and reduce the false alarm rate, i.e., specificity; the visual analysis results are to visualize the model evaluation results, intuitively display its performance, compare the predicted segmentation map with the true label map, analyze the specific positions of the error detections, and intuitively understand the performance of the model in the segmentation task; the model tuning results are to identify the deficiencies of the model based on the evaluation results and optimize them by increasing the data samples of relevant scenarios, adopting transfer learning or hierarchical training strategies, and reducing the model parameter quantity and improving the operation efficiency through pruning or quantization techniques to better meet the actual application requirements.

[0100] Specifically, S5 is as follows:

[0101] Constitute a dental plaque evaluation report based on the quantity, area, and position of the dental plaque on the teeth.

[0102] Refer to Figure 2 As shown, it is a schematic structural diagram of an intelligent electric toothbrush evaluation system provided by an embodiment of the present application. The system includes:

[0103] Toothbrush configuration module: Set the fluorescence wavelength data of the dental plaque, and obtain the fluorescence excitation device of the intelligent electric toothbrush according to the fluorescence wavelength data; set the image acquisition device and the sensing device of the intelligent electric toothbrush;

[0104] Tooth data acquisition module: Activate the tooth fluorescence according to the fluorescence excitation device, obtain the tooth fluorescence image of the teeth according to the image acquisition device, and obtain the brush head movement trajectory of the intelligent electric toothbrush according to the sensing device;

[0105] Data analysis module: Obtain a fluorescence image dataset based on the tooth fluorescence image; perform image processing and image region segmentation on the tooth fluorescence image, and perform region segmentation on the tooth fluorescence image to obtain a region segmentation result; perform region annotation on the tooth fluorescence images in the fluorescence image dataset according to the region segmentation result, and divide the fluorescence image dataset into a training set, a validation set, and a test set;

[0106] Model setting module: Establish a tooth evaluation model for the intelligent electric toothbrush;

[0107] The tooth evaluation model is trained based on the training set and the validation set, and the tooth evaluation model is evaluated based on the test set to obtain the target model parameters of the tooth evaluation model; the target tooth evaluation model is obtained based on the target model parameters.

[0108] Tooth evaluation module: Based on the intelligent electric toothbrush, plaque detection and evaluation of the teeth are carried out to obtain a plaque evaluation report of the teeth.

[0109] The above are only the preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art, several improvements and refinements made without departing from the principle of the present invention should also be regarded as within the protection scope of the present invention.

Claims

1. An intelligent electric toothbrush-based evaluation method, characterized in that Including: S1: Obtain the fluorescence wavelength data of dental plaque, and set the fluorescence excitation device of the intelligent electric toothbrush according to the fluorescence wavelength data; Set the image acquisition device and the sensing device of the intelligent electric toothbrush; S2: Activate the tooth fluorescence according to the fluorescence excitation device, obtain the tooth fluorescence image of the teeth according to the image acquisition device, and obtain the brush head movement trajectory of the intelligent electric toothbrush according to the sensing device; S3: Obtain a fluorescence image dataset from the tooth fluorescence image; perform image processing and image region segmentation on the tooth fluorescence image, and perform region segmentation on the tooth fluorescence image to obtain a region segmentation result; perform region annotation on the tooth fluorescence images in the fluorescence image dataset according to the region segmentation result, and divide the fluorescence image dataset into a training set, a validation set, and a test set; S4: Establish a tooth evaluation model for the intelligent electric toothbrush; S41: Train the tooth evaluation model according to the training set and the validation set, and evaluate the tooth evaluation model according to the test set to obtain the target model parameters of the tooth evaluation model; obtain the target tooth evaluation model according to the target model parameters; S5: Detect and evaluate dental plaque on the teeth according to the intelligent electric toothbrush to obtain a dental plaque evaluation report of the teeth.

2. The evaluation method based on an intelligent electric toothbrush according to claim 1, wherein Specifically, S1 is: The fluorescence excitation device includes a fluorescence excitation light source and a filter system, where the filter system includes an excitation filter and an emission filter; Fluorescently activate the dental plaque on the teeth according to the fluorescence excitation light source to obtain a tooth fluorescence signal; And filter the tooth fluorescence signal according to the excitation filter and the emission filter.

3. The method for evaluating an intelligent electric toothbrush according to claim 2, wherein, Specifically, S2 is: Perform image correction and enhancement on the tooth fluorescence image to obtain a target fluorescence region; Perform region segmentation on the target fluorescence region according to the brush head movement trajectory, and divide the target fluorescence region into a dental plaque region and a normal region; Quantify the area and concentration of the dental plaque region to obtain dental plaque region parameters, where the dental plaque region parameters include the dental plaque area and the dental plaque concentration corresponding to the dental plaque region.

4. The evaluation method based on an intelligent electric toothbrush according to claim 3, wherein Specifically, S3 is: Label the normal region of the tooth fluorescence image in the fluorescence image dataset as the normal region; Label the dental plaque region of the tooth fluorescence image in the fluorescence image dataset as the dental caries region, and perform region parameter annotation on the dental caries region according to the dental plaque region parameters corresponding to the dental plaque region; Perform data enhancement processing on all the tooth fluorescence images in the fluorescence image dataset to obtain the fluorescence image data to be divided; Divide the fluorescence image data to be divided into a training set, a validation set, and a test set.

5. The method for evaluating an intelligent electric toothbrush according to claim 4, wherein Specifically, S4 is: Establish a tooth evaluation model based on the convolutional neural network architecture. The tooth evaluation model includes an input end, an analysis end, and an output end, and the tooth evaluation model is provided with a channel attention mechanism and a spatial attention mechanism; The input end is used to input the tooth fluorescence image; The analysis terminal is provided with a semantic segmentation network, which is used to perform image detection on the tooth fluorescence image according to the model parameters to obtain the spatial distribution characteristics of dental plaque in the tooth, and perform semantic recognition on the spatial distribution characteristics according to the semantic segmentation network to obtain an image recognition result; The image recognition result includes the number of dental plaques, the area of dental plaques, and the positions of dental plaques on the tooth; The output terminal is used to output the image recognition result of the analysis terminal.

6. The method for evaluating an intelligent electric toothbrush according to claim 5, wherein, S41 specifically is: Update the model parameters of the tooth evaluation model according to the training set; Verify the model parameters of the tooth evaluation model according to the validation set; Evaluate the model parameters of the tooth evaluation model according to the test set to obtain the target model parameters of the tooth evaluation model.

7. The method for evaluating an intelligent electric toothbrush according to claim 6, wherein Evaluating the model parameters of the convolutional neural network model according to the test set to obtain the target model parameters of the convolutional neural network specifically is: Testing the corresponding model parameters of the tooth evaluation model according to the test set to obtain the performance index evaluation result, the visual analysis result, and the model tuning result of the corresponding model parameters of the tooth evaluation model; Obtain the target model parameters of the tooth evaluation model according to the performance index evaluation result, the visual analysis result, and the model tuning result.

8. The method for evaluating an intelligent electric toothbrush according to claim 7, wherein S5 specifically is: Constitute a dental plaque evaluation report based on the number of dental plaques, the area of dental plaques, and the positions of dental plaques on the tooth.

9. An intelligent electric toothbrush-based evaluation system for implementing an intelligent electric toothbrush-based evaluation method according to any one of claims 1-8, characterized in that, The system includes: Toothbrush configuration module: Set the fluorescence wavelength data of dental plaque, and obtain the fluorescence excitation device of the intelligent electric toothbrush according to the fluorescence wavelength data; Set the image acquisition device and the sensing device of the intelligent electric toothbrush; Tooth data acquisition module: Activate the tooth fluorescence according to the fluorescence excitation device, obtain the tooth fluorescence image of the tooth according to the image acquisition device, and obtain the brush head movement trajectory of the intelligent electric toothbrush according to the sensing device; Data analysis module: Obtain a fluorescence image data set according to the tooth fluorescence image; Perform image processing and image region segmentation on the tooth fluorescence image, and perform region segmentation on the tooth fluorescence image to obtain a region segmentation result; Perform region annotation on the tooth fluorescence images in the fluorescence image data set according to the region segmentation result, and divide the fluorescence image data set into a training set, a validation set, and a test set; Model setting module: Establish the tooth evaluation model of the intelligent electric toothbrush; Train the tooth evaluation model according to the training set and the validation set, and evaluate the tooth evaluation model according to the test set to obtain the target model parameters of the tooth evaluation model; Obtain the target tooth evaluation model according to the target model parameters; Tooth evaluation module: Detect and evaluate dental plaque on the tooth according to the intelligent electric toothbrush to obtain a dental plaque evaluation report of the tooth.

Images