Myopia classification method and device, equipment and storage medium

Abstract

The invention relates to the technical field of auxiliary diagnosis, and discloses a shortsightedness classification method, device and equipment and a storage medium, the shortsightedness classification method comprises the following steps: obtaining target data of a target object, the target data comprising traditional Chinese medicine examination data and medical examination data; the traditional Chinese medicine examination data comprises at least one of physique type data, tongue condition data and pulse condition data; the medical examination data comprises at least one of vision data, eye axis length data and diopter data; for the target data, performing classification prediction by using a myopia classification model; and determining a myopia classification result of the target object based on a classification prediction result output by the myopia classification model. The accuracy of myopia classification can be improved, and subsequent myopia prevention and treatment effects are ensured.

Description

Technical Field

[0001] This invention relates to the field of auxiliary diagnostic technology, specifically to a method, device, equipment, and storage medium for myopia classification. Background Technology

[0002] Myopia is a common vision problem, and its classification and diagnosis are crucial for treatment and prevention. Traditional ophthalmological diagnostic methods mainly rely on visual acuity tests and fundus examinations, but these methods often only assess myopia from a localized or single perspective. They cannot provide more accurate myopia classification results, which affects the final effectiveness of myopia prevention and treatment. Summary of the Invention

[0003] In view of this, the present invention provides a method, apparatus, device and storage medium for myopia classification, in order to solve the problem that inaccurate myopia classification results affect the prevention and treatment effects.

[0004] In a first aspect, the present invention provides a method for classifying myopia, the method comprising:

[0005] Acquire target data for the target object, wherein the target data includes traditional Chinese medicine diagnosis data and medical examination data; the traditional Chinese medicine diagnosis data includes at least one of constitution type data, tongue appearance data, and pulse appearance data; the medical examination data includes at least one of visual acuity data, axial length data, and refractive error data;

[0006] For the target data, a myopia classification model is used for classification prediction;

[0007] Based on the classification prediction results output by the myopia classification model, the myopia classification result of the target object is determined.

[0008] In one optional implementation, acquiring the target data of the target object includes:

[0009] Obtain the answer data of the Traditional Chinese Medicine constitution identification questionnaire filled out by the target object;

[0010] Semantic analysis was performed on the response data of the Traditional Chinese Medicine constitution identification questionnaire to obtain the semantic analysis results.

[0011] Based on the semantic analysis results, the physical type data of the target object is obtained.

[0012] In one optional implementation, acquiring the target data of the target object includes:

[0013] Obtain an image of the target object's tongue;

[0014] Based on the tongue image, tongue image data of the target object is obtained, including tongue color data, tongue shape data, and tongue coating data.

[0015] In one optional implementation, acquiring the tongue image data of the target object based on the tongue image includes:

[0016] The shape features of the tongue are extracted from the tongue image using the local binary pattern algorithm;

[0017] Based on the aforementioned shape features, the tongue shape data is determined.

[0018] In one optional implementation, acquiring the tongue image data of the target object based on the tongue image includes:

[0019] The tongue image is identified using a deep learning model;

[0020] The tongue coating data is determined based on the recognition results of the deep learning model.

[0021] In one optional implementation, acquiring the target data of the target object includes:

[0022] Acquire pulse-related data continuously collected by pressure sensors or photoelectric sensors;

[0023] The pulse data is obtained by performing time series feature analysis on the pulse-related data using a long short-term memory network and a gated recurrent unit.

[0024] In one optional implementation, the myopia classification model includes a data layer, a feature layer, and a decision layer;

[0025] The data layer is used to concatenate the target data into a long vector in sequence;

[0026] The feature layer is used to fuse the different target data in the long vector into a unified feature representation using weighted averaging and / or principal component analysis methods.

[0027] The decision layer is used to perform myopia classification prediction based on the fused feature representation output by the feature layer.

[0028] Secondly, the present invention provides a myopia classification device, the device comprising:

[0029] The data acquisition module is used to acquire target data of the target object, wherein the target data includes traditional Chinese medicine diagnosis data and medical examination data; the traditional Chinese medicine diagnosis data includes at least one of constitution type data, tongue appearance data and pulse appearance data; the medical examination data includes at least one of visual acuity data, axial length data and refractive error data.

[0030] The prediction module is used to perform classification prediction using a myopia classification model for the target data;

[0031] The classification determination module is used to determine the myopia classification result of the target object based on the classification prediction result output by the myopia classification model.

[0032] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the myopia classification method of the first aspect or any corresponding embodiment described above.

[0033] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the myopia classification method of the first aspect or any corresponding embodiment described above.

[0034] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the myopia classification method of the first aspect or any corresponding embodiment described above.

[0035] The myopia classification method, device, equipment, and storage medium provided in this invention offer a new auxiliary method for myopia classification based on the theory of traditional Chinese medicine constitution identification. This method combines traditional Chinese medicine and modern ophthalmological knowledge to classify myopia, providing a more comprehensive reference for myopia classification and thus improving the accuracy of myopia classification. This helps in the development of personalized prevention and treatment plans, thereby ensuring the effectiveness of subsequent myopia prevention and treatment. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the specific embodiments or related technologies of the present invention, the drawings used in the description of the specific embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0037] Figure 1 This is a flowchart illustrating a myopia classification method according to an embodiment of the present invention;

[0038] Figure 2 This is a structural block diagram of a myopia classification device according to an embodiment of the present invention;

[0039] Figure 3 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] According to an embodiment of the present invention, a method for classifying myopia is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of executable computer instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0042] This embodiment provides a myopia classification method that can be used in various computer devices, including terminals and servers. Figure 1 This is a flowchart of a myopia classification method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:

[0043] Step S101: Obtain target data of the target object (i.e., the patient). The target data includes TCM diagnosis data and medical examination data. The TCM diagnosis data includes at least one of constitution type data, tongue appearance data, and pulse appearance data. The medical examination data includes at least one of visual acuity data, axial length data, and refractive error data.

[0044] Specifically, there are nine constitution types: balanced, qi-deficient, yang-deficient, yin-deficient, phlegm-dampness, damp-heat, blood stasis, qi stagnation, and special constitution. Medical examination data refers to modern ophthalmological testing data, specifically using vision testing equipment with autofocus and intelligent measurement functions to quickly and accurately measure the target subject's visual acuity and refractive error. Optical coherence tomography (OCT) imaging equipment is used to measure the axial length of the eye and obtain three-dimensional structural information of the eye. This three-dimensional structural information can be used as part of the medical examination data; that is, medical examination data can include three-dimensional structural information of the eye. Data such as visual acuity, axial length, and refractive error can be normalized to form corresponding feature vectors, facilitating subsequent input into a myopia classification model for myopia classification prediction.

[0045] In some optional implementations, step S101, namely obtaining the target data of the target object, includes:

[0046] Step S101-01: Obtain the answer data of the Traditional Chinese Medicine (TCM) constitution identification questionnaire completed by the target subject. The answer data of the TCM constitution identification questionnaire is obtained by the target subject answering the questions in the TCM constitution identification questionnaire.

[0047] Specifically, an electronic TCM constitution identification questionnaire can be provided directly for the target group to fill out. Alternatively, a paper TCM constitution identification questionnaire can be provided for the target group to fill out, then the questionnaire can be photographed with a mobile phone or scanned to generate a PDF / image, and finally the text can be extracted using Optical Character Recognition (OCR) function.

[0048] When using electronic TCM constitution identification questionnaires, subsequent questions can be dynamically adjusted based on the target audience's answers, reducing interference from irrelevant questions and improving the questionnaire's relevance and efficiency.

[0049] Step S101-02: Perform semantic analysis on the response data of the TCM constitution identification questionnaire to obtain the semantic analysis results.

[0050] Specifically, natural language processing (NLP) technology can be used to perform semantic analysis on the responses to the TCM constitution identification questionnaire, accurately extracting relevant information. In this embodiment of the invention, NLP technology is used to process the responses to the TCM constitution identification questionnaire, converting the responses into feature vectors.

[0051] In addition, before performing semantic analysis on the response data of the TCM constitution identification questionnaire, the response data text can be preprocessed by word segmentation and stop word removal, and then word embedding technology (Bidirectional Encoder Representations from Transformers, BERT) can be used to transform it into numerical feature vectors.

[0052] Step S101-03: Based on the semantic analysis results, obtain the physical type data of the target object.

[0053] For example, the semantic analysis results can be input into a traditional Chinese medicine constitution classification model. Then, based on the output of the traditional Chinese medicine constitution classification model, the constitution type data of the target object can be obtained.

[0054] In this embodiment of the invention, a questionnaire can be used to conduct a consultation with the target subject, and then a deep learning-based TCM constitution classification model can be used to classify the data obtained from the consultation into a TCM constitution type. This eliminates the need for a professional doctor to determine the target subject's constitution type through face-to-face consultation. This reduces labor costs and avoids the influence of subjective factors on the judgment results.

[0055] In other embodiments, the target's physical condition type data can also be determined by a doctor.

[0056] In some optional implementations, step S101, namely obtaining the target data of the target object, includes:

[0057] Step S101-a: Acquire an image of the target object's tongue. Specifically, a high-resolution camera and standardized lighting equipment can be used to capture the image of the target object's tongue.

[0058] After acquiring the tongue image captured by the camera, it can be processed by grayscale conversion, normalization, etc., and then tongue image data can be obtained based on the processed tongue image.

[0059] Step S101-b: Based on the tongue image, obtain the tongue image data of the target object, including tongue color data, tongue shape data, and tongue coating data.

[0060] Specifically, tongue color can be categorized as follows: pale white, red, crimson (deep red), purple, and bluish. Tongue shape can be categorized as follows: swollen, thin, fissured, teeth-marked, and prickly. Tongue coating data includes both texture and color. Coating color can be white, yellow, or grayish-black. Coating texture can be categorized as thick / thin, moist / dry, or greasy / rotten. Thick / thin: thin coating, thick coating. Moist / dry: moist, dry. Greasy / rotten: rotten coating, greasy coating.

[0061] In some specific implementations, step S101-b, namely, acquiring the tongue image data of the target object based on the tongue image, includes:

[0062] Step S101-b1: Extract the shape features of the tongue from the tongue image using the Local Binary Pattern (LBP) algorithm;

[0063] Step S101-b2: Determine the tongue shape data based on the shape features.

[0064] Specifically, the local binary mode algorithm used in this embodiment of the invention can be an improved local binary mode (LBP) algorithm, which can generate continuous binary images for the tongue area, avoiding holes and discontinuous structures.

[0065] In other embodiments, edge detection algorithms, such as the Canny operator, can be used to detect the outline of the tongue in the tongue image, and then determine parameters such as the thickness of the tongue and the depth of cracks.

[0066] In some other specific embodiments, step S101-b, namely, obtaining the tongue image data of the target object based on the tongue image, includes:

[0067] Step S101-b3: Use a deep learning model to identify the tongue image;

[0068] Step S101-b4: Determine the tongue coating data based on the recognition results of the deep learning model.

[0069] Specifically, a deep learning model for tongue coating data recognition can be constructed using a Convolutional Neural Network (CNN). In this embodiment of the invention, using deep learning algorithms to analyze and classify the tongue coating in tongue images can improve the accuracy and efficiency of tongue coating data identification.

[0070] Specifically, deep learning models can be U-Net (U-shaped Convolutional Network), DeepLabV3+ (Deep Labelling Version 3 Plus), Swin Transformer (Shifted Window Transformer), or Mobile-UNet (Mobile U-shaped Network). DeepLabV3+ combines dilated convolutions with multi-scale features, making it suitable for complex edges. Swin Transformer, based on an attention mechanism, is suitable for high-precision requirements. Lightweight models (such as Mobile-Unet) are suitable for mobile deployment.

[0071] The training process of a deep learning model includes:

[0072] I. Data Preparation

[0073] 1. Dataset Construction

[0074] Data source:

[0075] Acquire standardized tongue images (with uniform light source, angle, and background, such as using a tongue imager).

[0076] Public datasets (such as TongueNet, TCM-GRAKE_19).

[0077] Labeling requirements:

[0078] The tongue coating area is marked at the pixel level by a traditional Chinese medicine doctor (distinguishing between tongue coating, tongue body, and background).

[0079] Annotation tools: LabelMe, CVAT, or professional medical annotation platforms (such as 3D Slicer).

[0080] 2. Data Preprocessing

[0081] Image standardization:

[0082] Adjust the resolution (512×512 or 256×256 are recommended).

[0083] Color correction (to resolve uneven lighting issues, such as histogram equalization).

[0084] Data augmentation:

[0085] Geometric transformations (rotation, flipping, cropping).

[0086] Color perturbation (brightness and contrast adjustments to simulate differences between different devices).

[0087] II. Model Training

[0088] Input: Preprocessed tongue image (RGB three channels).

[0089] Output: Binary mask (0 = non-tongue coating, 1 = tongue coating), coating texture and coating color.

[0090] Loss function:

[0091] Dice Loss: Resolves category imbalance (small proportion of tongue coating area).

[0092] Joint loss: Dice Loss + BCE Loss (balancing global and local features).

[0093] Training techniques:

[0094] Transfer learning: using pre-trained weights (such as ImageNet).

[0095] Progressive training: First train the segmentation of the tongue body, then fine-tune the segmentation of the tongue coating.

[0096] In addition, tongue color data can be obtained through color space analysis. Similarly, tongue coating color can also be obtained through color space analysis.

[0097] In some optional implementations, step S101, obtaining the target data of the target object, includes:

[0098] Step S101 (a): Acquire pulse-related data continuously collected by a pressure sensor or photoelectric sensor.

[0099] Specifically, pulse diagnosis in Traditional Chinese Medicine includes multidimensional information:

[0100] Pulse location: floating, middle, deep (pulse depth);

[0101] Pulse strength: strength (intensity of pulse beats);

[0102] Pulse rate: speed (heart rate and rhythm);

[0103] Pulse characteristics: wiry, slippery, and hesitant (waveform features);

[0104] Pulse state: its coming and going (slope of rising / falling).

[0105] In this embodiment of the invention, pressure changes can be detected using a pressure sensor, thereby detecting vascular pulsation and simulating the "lifting, pressing, and searching" techniques used in traditional Chinese medicine palpation. Specifically, the pressure sensor can be a piezoelectric sensor, a piezoresistive sensor, or a capacitive sensor. Alternatively, this embodiment of the invention can also use a photoelectric sensor (photoplethysmography (PPG) technology) to collect pulse data, specifically detecting subcutaneous blood flow volume fluctuations through changes in light absorption, indirectly reflecting pulse characteristics.

[0106] For pulse-related data acquired by sensors, preprocessing such as filtering and noise reduction can be performed before analysis to obtain pulse data.

[0107] Step S101 (II): Using a Long Short-Term Memory Network (LSTM) and a Gated Recurrent Unit (GRU), perform time series feature analysis on the pulse-related data to obtain the pulse data.

[0108] In this embodiment of the invention, a bidirectional gated recurrent unit (Bi-GRU) can be used to extract the time-series features of the pulse signal acquired by the sensor, obtaining a feature vector as the pulse-related data. Then, LSTM+GRU is used to predict the pulse pattern data. In this embodiment of the invention, Bi-GRU is used to feature the one-dimensional time-series pulse signal acquired by the sensor, i.e., convert it into a feature vector. Specifically, the one-dimensional time-series pulse signal acquired by the sensor is used as input, and only the output of the intermediate layer of the Bi-GRU is extracted as the feature vector.

[0109] Step S102: For the target data, perform classification prediction using a myopia classification model. The myopia classification model may be a machine learning model, a deep learning model, or a non-deep learning model.

[0110] Step S103: Based on the classification prediction results output by the myopia classification model, determine the myopia classification result of the target object.

[0111] Specific myopia classification results, such as simple myopia and physiological myopia.

[0112] In some optional implementations, the myopia classification model includes a data layer, a feature layer, and a decision layer;

[0113] The data layer is used to concatenate the target data into a long vector in sequence;

[0114] The feature layer is used to fuse the different target data in the long vector into a unified feature representation using weighted averaging and / or Principal Component Analysis (PCA). Regarding the use of weighted averaging to fuse the different target data in the long vector into a unified feature representation, for example, suppose there are n vectors v1, v2, ..., vn, and the weights of each vector are w1, w2, ..., wn. The formula for calculating the weighted average vector Vavg is:

[0115]

[0116] Where vi is the i-th vector, wi is the weight of the i-th vector, and the denominator is the sum of all weights, used for normalization.

[0117] The decision layer is used to perform myopia classification prediction based on the fused feature representation output by the feature layer.

[0118] Specifically, in this embodiment of the invention, a multimodal fusion model, i.e., a myopia classification model, can be constructed using a deep learning framework (PyTorch). For example, it can be a neural network model containing multiple fully connected layers, activation functions (such as ReLU), and an output layer. Furthermore, after obtaining the initially constructed myopia classification model, training samples need to be collected to train the model. During training, hyperparameters (such as learning rate, regularization parameters, etc.) can be dynamically adjusted, and methods such as cross-validation can be used to evaluate the model's performance. In this embodiment of the invention, clinical data can be used to construct training samples to train and validate the myopia classification model, thereby optimizing the model's parameters and improving the accuracy and reliability of the myopia classification.

[0119] In other embodiments, the myopia classification model can also be a support vector machine.

[0120] With the development of Traditional Chinese Medicine (TCM) constitution identification theory, it has demonstrated unique value in the diagnosis and treatment of various diseases. TCM constitution identification theory posits that an individual's constitution type is closely related to the occurrence and development of diseases. By identifying an individual's constitution characteristics, a more comprehensive basis can be provided for disease classification and treatment. This embodiment uses TCM constitution identification theory to provide a new auxiliary method for myopia classification, combining TCM and modern ophthalmological knowledge to provide a more comprehensive reference for myopia classification, thereby improving the accuracy of myopia classification. This facilitates the development of personalized prevention and treatment plans, ultimately ensuring the effectiveness of subsequent myopia prevention and treatment.

[0121] In summary, the myopia classification method provided by this invention combines traditional Chinese medicine constitution identification theory with modern medical examination, offering a more comprehensive perspective for myopia classification. Furthermore, this invention also provides new application examples for the integration of traditional Chinese medicine and modern medicine, promoting the development of integrated traditional Chinese and Western medicine.

[0122] This embodiment also provides a myopia classification device for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0123] This embodiment provides a myopia classification device, such as... Figure 2 As shown, it includes:

[0124] The data acquisition module 201 is used to acquire target data of the target object, wherein the target data includes traditional Chinese medicine diagnosis data and medical examination data; the traditional Chinese medicine diagnosis data includes at least one of constitution type data, tongue appearance data and pulse appearance data; and the medical examination data includes at least one of visual acuity data, axial length data and refractive error data.

[0125] Prediction module 202 is used to perform classification prediction on the target data using a myopia classification model;

[0126] The classification determination module 203 is used to determine the myopia classification result of the target object based on the classification prediction result output by the myopia classification model.

[0127] In some optional implementations, the data acquisition module 201 includes:

[0128] The questionnaire response data acquisition unit is used to acquire the response data of the Traditional Chinese Medicine constitution identification questionnaire filled out by the target object.

[0129] The semantic analysis unit is used to perform semantic analysis on the response data of the TCM constitution identification questionnaire to obtain semantic analysis results.

[0130] The constitution classification determination unit is used to obtain the constitution type data of the target object based on the semantic analysis results.

[0131] In some optional implementations, the data acquisition module 201 includes:

[0132] The image acquisition unit is used to acquire an image of the tongue of the target object;

[0133] The tongue image data acquisition unit is used to acquire the tongue image data of the target object based on the tongue image, wherein the tongue image data includes tongue color data, tongue shape data and tongue coating data.

[0134] In some optional implementations, the tongue image data acquisition unit includes:

[0135] The shape feature extraction subunit is used to extract the shape features of the tongue from the tongue image using the local binary pattern algorithm;

[0136] The tongue shape determination subunit is used to determine the tongue shape data based on the shape features.

[0137] In some optional implementations, the tongue image data acquisition unit includes:

[0138] The recognition subunit is used to recognize the tongue image using a deep learning model;

[0139] The tongue coating data determination subunit is used to determine the tongue coating data based on the recognition results of the deep learning model.

[0140] In some optional implementations, the data acquisition module 201 includes:

[0141] The pulse-related data acquisition unit is used to acquire pulse-related data continuously collected by the pressure sensor or photoelectric sensor.

[0142] The pulse data acquisition unit is used to perform time series feature analysis on the pulse-related data using a long short-term memory network and a gated recurrent unit to obtain the pulse data.

[0143] In some optional implementations, the myopia classification model includes a data layer, a feature layer, and a decision layer;

[0144] The data layer is used to concatenate the target data into a long vector in sequence;

[0145] The feature layer is used to fuse the different target data in the long vector into a unified feature representation using weighted averaging and / or principal component analysis methods.

[0146] The decision layer is used to perform myopia classification prediction based on the fused feature representation output by the feature layer.

[0147] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0148] In this embodiment, the myopia classification device is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0149] This invention also provides a computer device having the above-described features. Figure 2 The myopia classification device shown.

[0150] Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 3 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 3 Take a processor 10 as an example.

[0151] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0152] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0153] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0154] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0155] The computer device also includes an input device 30 and an output device 40. The processor 10, memory 20, input device 30, and output device 40 can be connected via a bus or other means. Figure 3 Taking the example of a connection between China and Israel via a bus.

[0156] Input device 30 can receive input numerical or character information, and generate key signal inputs related to user settings and function control of the computer device, such as a touchscreen, keypad, mouse, trackpad, touchpad, joystick, one or more mouse buttons, trackball, joystick, etc. Output device 40 may include display devices, auxiliary lighting devices (e.g., LEDs), and haptic feedback devices (e.g., vibration motors). The aforementioned display devices include, but are not limited to, liquid crystal displays, light-emitting diodes, displays, and plasma displays. In some alternative embodiments, the display device may be a touchscreen.

[0157] The computer device also includes a communication interface for communicating with other devices or communication networks.

[0158] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0159] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0160] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for classifying myopia, characterized in that, The method includes: Acquire target data for the target object, wherein the target data includes traditional Chinese medicine diagnosis data and medical examination data; the traditional Chinese medicine diagnosis data includes at least one of constitution type data, tongue appearance data, and pulse appearance data; the medical examination data includes at least one of visual acuity data, axial length data, and refractive error data; For the target data, a myopia classification model is used for classification prediction; Based on the classification prediction results output by the myopia classification model, the myopia classification result of the target object is determined.

2. The method according to claim 1, characterized in that, The acquisition of target data of the target object includes: Obtain the answer data of the Traditional Chinese Medicine constitution identification questionnaire filled out by the target object; Semantic analysis was performed on the response data of the Traditional Chinese Medicine constitution identification questionnaire to obtain the semantic analysis results. Based on the semantic analysis results, the physical type data of the target object is obtained.

3. The method according to claim 1, characterized in that, The acquisition of target data of the target object includes: Obtain an image of the target object's tongue; Based on the tongue image, tongue image data of the target object is obtained, including tongue color data, tongue shape data, and tongue coating data.

4. The method according to claim 3, characterized in that, The step of obtaining tongue image data of the target object based on the tongue image includes: The shape features of the tongue are extracted from the tongue image using the local binary pattern algorithm; Based on the aforementioned shape features, the tongue shape data is determined.

5. The method according to claim 3, characterized in that, The step of obtaining tongue image data of the target object based on the tongue image includes: The tongue image is identified using a deep learning model; The tongue coating data is determined based on the recognition results of the deep learning model.

6. The method according to claim 1, characterized in that, The acquisition of target data of the target object includes: Acquire pulse-related data continuously collected by pressure sensors or photoelectric sensors; The pulse data is obtained by performing time series feature analysis on the pulse-related data using a long short-term memory network and a gated recurrent unit.

7. The method according to claim 1, characterized in that, The myopia classification model includes a data layer, a feature layer, and a decision layer; The data layer is used to concatenate the target data into a long vector in sequence; The feature layer is used to fuse the different target data in the long vector into a unified feature representation using weighted averaging and / or principal component analysis methods. The decision layer is used to perform myopia classification prediction based on the fused feature representation output by the feature layer.

8. A myopia classification device, characterized in that, The device includes: The data acquisition module is used to acquire target data of the target object, wherein the target data includes traditional Chinese medicine diagnosis data and medical examination data; the traditional Chinese medicine diagnosis data includes at least one of constitution type data, tongue appearance data and pulse appearance data; the medical examination data includes at least one of visual acuity data, axial length data and refractive error data. The prediction module is used to perform classification prediction using a myopia classification model for the target data; The classification determination module is used to determine the myopia classification result of the target object based on the classification prediction result output by the myopia classification model.

9. A computer device, characterized in that, include: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the myopia classification method according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the myopia classification method according to any one of claims 1 to 7.

Images