Alzheimer's disease risk warning method, device and system based on handwriting recognition

Abstract

The present application provides a method, device and system for early warning of Alzheimer's disease risk based on handwriting recognition, which relates to the field of health information processing technology. The method includes: receiving handwriting recognition data from the Bluetooth pen in real time while the assessor is filling out a paper diagnostic scale with a handwriting pen; calculating the test subject's diagnostic scale score based on the handwriting recognition data and determining the test subject's first Alzheimer's disease risk level; using an Alzheimer's disease risk prediction model to determine the test subject's second Alzheimer's disease risk level, and obtaining the test subject's Alzheimer's disease risk warning result data. The present application can be applied to test subjects such as the elderly who are inconvenient to use electronic scales, and can ensure the efficiency and automation of paper scale data processing, can effectively improve the accuracy and reliability of Alzheimer's disease risk warning, and thus can better achieve auxiliary diagnosis of Alzheimer's disease for the test subject.

Description

Technical Field

[0001] The present application relates to the field of health information processing technology, and in particular to an Alzheimer's disease risk warning method, device, and system based on handwriting recognition. Background Art

[0002] Alzheimer's disease is a common neurodegenerative disorder affecting the elderly. The current trend is to shift the window for diagnosis and intervention earlier, focusing on the early stages of Alzheimer's disease and implementing timely and effective intervention to slow or even halt the progression to dementia. The key to Alzheimer's disease diagnosis and treatment lies in early screening and identification. Neuropsychological testing using diagnostic scales is currently the core basis for clinical diagnosis of Alzheimer's disease.

[0003] Paper-based diagnostic scales are commonly used in clinical practice. However, the traditional pen-and-paper recording method is difficult to store and trace persistently, and the cumbersome scale evaluation criteria make it difficult for users to understand their own condition, which can easily lead to difficulties in storing and sharing valuable medical information. Even if the scores of each scale are manually calculated and then input into the computer system, redundant data calculations are very likely to cause errors, while also resulting in additional human resource costs and information such as handwriting characteristics cannot be fully utilized. Therefore, in the current digital age, digital scales that are independent of paper media have emerged as effective tools for Alzheimer's disease screening. This means developing corresponding scale systems on mobile devices, and displaying and filling out scales on electronic screens.

[0004] However, Alzheimer's disease-related scales are typically designed for the elderly. Digital scales are expensive to learn, time-consuming, and have low success rates for the elderly, which can easily lead to problems such as low patient compliance and inaccurate assessment results. Therefore, there is an urgent need to design a new Alzheimer's disease risk warning method that can address the time-consuming, high labor costs, and low degree of automation of paper-based diagnostic scales, while also addressing the problem that digital scales are unsuitable for subjects such as the elderly who are unable to use electronic scales. Summary of the Invention

[0005] In view of this, the embodiments of the present application provide an Alzheimer's disease risk warning method, device and system based on handwriting recognition to eliminate or improve one or more defects in the prior art.

[0006] One aspect of the present application provides an Alzheimer's disease risk early warning method based on handwriting recognition, comprising:

[0007] When an assessor fills out a diagnostic scale with a preset dot matrix code for assisting in diagnosing Alzheimer's disease of a subject using a Bluetooth pen, the Bluetooth pen receives in real time various handwriting recognition data, wherein each set of the handwriting recognition data includes: a correspondence between the scale content data determined by the Bluetooth pen through real-time recognition of the dot matrix code and the handwriting data recorded in real time by the Bluetooth pen;

[0008] Calculating the test subject's diagnostic scale score based on each of the handwriting recognition data generated when the evaluator fills out all the diagnostic scales, and determining a first Alzheimer's disease risk level corresponding to the test subject based on the diagnostic scale score;

[0009] and, determining a second Alzheimer's disease risk level corresponding to the subject using a preset Alzheimer's disease risk prediction model based on the handwriting data in each of the handwriting recognition data;

[0010] Alzheimer's disease risk warning result data of the subject is obtained based on the first Alzheimer's disease risk level and the second Alzheimer's disease risk level.

[0011] In some embodiments of the present application, the scale content data includes the page, question number, and page area information of the diagnostic scale on which the dot matrix code currently touched by the Bluetooth pen is located;

[0012] The handwriting data includes: handwriting morphological feature data, writing pressure value and writing pause time; wherein, the handwriting morphological data includes: timestamp, coordinate point sequence of the current handwriting in the preset coordinate system corresponding to the diagnostic scale and stroke thickness value.

[0013] In some embodiments of the present application, calculating the diagnostic scale score of the test subject based on the handwriting recognition data generated when the evaluator fills out all the diagnostic scales, and determining the first Alzheimer's disease risk level corresponding to the test subject based on the diagnostic scale score, includes:

[0014] According to the preset test task types to which the respective diagnostic scales belong, the handwriting recognition data corresponding to the diagnostic scales whose test task type is a selective test task are divided into a selective test group, and the handwriting recognition data corresponding to the diagnostic scales whose test task type is a drawing test task are divided into a drawing test group;

[0015] Based on each of the handwriting recognition data in the selective test group, obtaining the total score corresponding to each of the diagnostic scales belonging to the selective test task, and summarizing the total score corresponding to each of the diagnostic scales belonging to the selective test task to obtain the selective test score corresponding to the subject;

[0016] and, based on each of the handwriting recognition data in the drawing-type test group, obtaining a total score corresponding to each of the diagnostic scales belonging to the drawing-type test task, and summarizing the total score corresponding to each of the diagnostic scales belonging to the drawing-type test task to obtain a drawing-type test score corresponding to the subject;

[0017] The selective test score and the drawing test score corresponding to the subject are determined as the diagnostic scale score of the subject, and the first Alzheimer's disease risk level corresponding to the subject is determined according to the diagnostic scale score of the subject from the diagnostic standard data used to record the correspondence between the diagnostic scale score and the Alzheimer's disease risk level.

[0018] In some embodiments of the present application, obtaining the total score corresponding to each diagnostic scale belonging to the selective test task based on each piece of handwriting recognition data in the selective test group includes:

[0019] Executing a preset candidate box checking judgment step for each of the handwriting recognition data corresponding to each of the diagnostic scales belonging to the selective test task, respectively, to determine the score of each checked candidate box in each of the diagnostic scales belonging to the selective test task, and summarizing the score of each checked candidate box in each of the diagnostic scales belonging to the selective test task to obtain the total score corresponding to each of the diagnostic scales belonging to the selective test task;

[0020] The step of checking the candidate box includes:

[0021] Determine, based on the scale content data in the current handwriting recognition data, a page identifier in the diagnostic scale corresponding to the handwriting recognition data, and obtain, from the candidate box information corresponding to each of the preset diagnostic scales belonging to the selective test task, position information of each candidate box in the target page corresponding to the page identifier;

[0022] Based on the handwriting data in the current handwriting recognition data and the position information of each candidate box in the target page, determining whether the distance between the handwriting data and any candidate box in the target page is less than a preset distance threshold, and if so, determining that the candidate box whose distance to the handwriting data is less than the distance threshold is selected;

[0023] The score of the selected candidate box is determined based on the scores corresponding to the respective candidate boxes stored in the candidate box information.

[0024] In some embodiments of the present application, obtaining the total score corresponding to each diagnostic scale belonging to the drawing-type test task based on each piece of handwriting recognition data in the drawing-type test group includes:

[0025] converting the handwriting data corresponding to each of the handwriting recognition data in the drawing type test group into handwriting image data respectively;

[0026] Each of the handwriting image data is respectively input into a preset handwriting image scoring model to obtain the score corresponding to each of the handwriting recognition data in the drawing type test group, and based on the score corresponding to each of the handwriting recognition data in the drawing type test group, the total score corresponding to each of the diagnostic scales belonging to the drawing type test task is obtained.

[0027] In some embodiments of the present application, determining the second Alzheimer's disease risk level corresponding to the subject using a preset Alzheimer's disease risk prediction model based on the handwriting data in each of the handwriting recognition data includes:

[0028] generating handwriting morphological feature image data corresponding to each of the handwriting recognition data according to the handwriting morphological feature data in the handwriting data in the handwriting recognition data;

[0029] Setting data samples corresponding to each of the handwriting recognition data, wherein each of the data samples includes the handwriting morphological feature image data, the writing pressure value, and the writing pause time corresponding to the handwriting recognition data;

[0030] Inputting the data sample into a preset Alzheimer's disease risk prediction model so that the Alzheimer's disease risk prediction model outputs an Alzheimer's disease risk score for the subject; the Alzheimer's disease risk prediction model is previously trained and generated by a random forest model;

[0031] A second Alzheimer's disease risk level corresponding to the subject is determined according to the Alzheimer's disease risk score.

[0032] In some embodiments of the present application, the obtaining of the Alzheimer's disease risk warning result data of the subject based on the first Alzheimer's disease risk level and the second Alzheimer's disease risk level includes:

[0033] The first Alzheimer's disease risk level and the second Alzheimer's disease risk level are input into a preset diagnostic analysis report generation model so that the diagnostic analysis report generation model outputs a corresponding diagnostic analysis report, and the diagnostic analysis report is output as the Alzheimer's disease risk warning result data of the subject.

[0034] In some embodiments of the present application, further comprising:

[0035] The handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects are classified and stored in a relational database, and the handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects are extracted from the relational database. The handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects are extracted from the relational database as training data to train the Alzheimer's disease risk prediction model or the handwriting image scoring model.

[0036] Another aspect of the present application further provides an Alzheimer's disease risk warning device based on handwriting recognition, comprising:

[0037] A data acquisition module is configured to receive, in real time, individual handwriting recognition data from a Bluetooth pen while an assessor is filling out a diagnostic scale with a preset dot matrix code for assisting in diagnosing Alzheimer's disease in a subject using the Bluetooth pen, wherein each set of handwriting recognition data includes a correspondence between scale content data determined by the Bluetooth pen through real-time recognition of the dot matrix code and handwriting data recorded in real time by the Bluetooth pen;

[0038] a scale score calculation module, configured to calculate the diagnostic scale score of the test subject based on the handwriting recognition data generated when the evaluator fills out all the diagnostic scales, and determine the first Alzheimer's disease risk level corresponding to the test subject based on the diagnostic scale score;

[0039] and a handwriting risk calculation module, configured to determine a second Alzheimer's disease risk level corresponding to the subject using a preset Alzheimer's disease risk prediction model based on the handwriting data in each of the handwriting recognition data;

[0040] The risk warning result acquisition module is used to obtain the Alzheimer's disease risk warning result data of the subject based on the first Alzheimer's disease risk level and the second Alzheimer's disease risk level.

[0041] The third aspect of the present application further provides an Alzheimer's disease risk warning system based on handwriting recognition, comprising: a diagnostic scale for assisting in diagnosing Alzheimer's disease in a subject and provided with a preset dot matrix code, a Bluetooth pen, and a mobile client device; the Bluetooth pen and the mobile client device are in communication connection;

[0042] The mobile client device is used to execute the Alzheimer's disease risk warning method based on handwriting recognition;

[0043] The Bluetooth pen is provided with a sensor for collecting the handwriting data in real time.

[0044] The fourth aspect of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the Alzheimer's disease risk warning method based on handwriting recognition when executing the computer program.

[0045] A fifth aspect of the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the Alzheimer's disease risk warning method based on handwriting recognition.

[0046] The sixth aspect of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the Alzheimer's disease risk warning method based on handwriting recognition.

[0047] The present application provides an Alzheimer's disease risk warning method based on handwriting recognition. In the process of an assessor holding a Bluetooth pen to fill out a diagnostic scale for assisting in diagnosing Alzheimer's disease of a subject and provided with a preset dot matrix code, each handwriting recognition data is received in real time from the Bluetooth pen, wherein each set of the handwriting recognition data includes: the correspondence between the scale content data determined by the Bluetooth pen through real-time recognition of the dot matrix code and the handwriting data recorded in real time by the Bluetooth pen; the diagnostic scale score of the subject is calculated based on each of the handwriting recognition data generated when the assessor fills out all the diagnostic scales, and the first Alzheimer's disease score corresponding to the subject is determined based on the diagnostic scale score. Alzheimer's disease risk level; and, based on the handwriting data in each of the handwriting recognition data, a preset Alzheimer's disease risk prediction model is used to determine the second Alzheimer's disease risk level corresponding to the subject; based on the first Alzheimer's disease risk level and the second Alzheimer's disease risk level, the Alzheimer's disease risk warning result data of the subject is obtained, which can be suitable for the subject population who are inconvenient to use electronic scales, such as the elderly, and can ensure the efficiency and automation level of paper quality scale data processing, can effectively improve the accuracy and reliability of Alzheimer's disease risk warning, and thus can better realize the auxiliary diagnosis of Alzheimer's disease for the subject.

[0048] Additional advantages, purposes, and features of the present application will be described in part in the following description and will become apparent to those skilled in the art upon study of the following or may be learned from practice of the present application. The purposes and other advantages of the present application may be achieved and obtained by the structures specifically pointed out in the specification and drawings.

[0049] Those skilled in the art will understand that the purposes and advantages that can be achieved by the present application are not limited to the above specific description, and the above and other purposes that can be achieved by the present application will be more clearly understood based on the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The drawings described herein are intended to provide a further understanding of the present application, constitute a part of the present application, and do not constitute a limitation of the present application. The components in the drawings are not drawn to scale, but are only for the purpose of illustrating the principles of the present application. In order to facilitate the illustration and description of some parts of the present application, the corresponding parts in the drawings may be enlarged, that is, they may become larger than other components in the exemplary device actually manufactured according to the present application. In the drawings:

[0051] Figure 1 This is a first flow chart of an Alzheimer's disease risk warning method based on handwriting recognition in one embodiment of the present application.

[0052] Figure 2 This is a second flow chart of the Alzheimer's disease risk warning method based on handwriting recognition in one embodiment of the present application.

[0053] Figure 3 This is a third flow chart of the Alzheimer's disease risk warning method based on handwriting recognition in one embodiment of the present application.

[0054] Figure 4 This is a flowchart of the candidate box selection and judgment step in the Alzheimer's disease risk warning method based on handwriting recognition in one embodiment of the present application.

[0055] Figure 5 This is a schematic diagram of the first structure of an Alzheimer's disease risk warning device based on handwriting recognition in one embodiment of the present application.

[0056] Figure 6 This is a second structural diagram of an Alzheimer's disease risk warning device based on handwriting recognition in one embodiment of the present application. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail in conjunction with the embodiments and drawings. Here, the illustrative embodiments of this application and their descriptions are used to explain this application, but are not intended to limit this application.

[0058] It should also be noted here that in order to avoid obscuring the present application due to unnecessary details, the accompanying drawings only show structures and / or processing steps that are closely related to the scheme according to the present application, while other details that are not closely related to the present application are omitted.

[0059] It should be emphasized that the term "include / comprises" when used herein refers to the existence of features, elements, steps or components, but does not exclude the existence or addition of one or more other features, elements, steps or components.

[0060] It should also be noted that, unless otherwise specified, the term "connection" herein may refer not only to a direct connection but also to an indirect connection involving an intermediate.

[0061] Hereinafter, embodiments of the present application will be described with reference to the accompanying drawings. In the accompanying drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

[0062] In order to achieve a balance between traditional paper-based scales and electronic scales that are completely based on software systems, the embodiments of the present application respectively provide an Alzheimer's disease risk warning method based on handwriting recognition, an Alzheimer's disease risk warning device based on handwriting recognition for executing the Alzheimer's disease risk warning method based on handwriting recognition, an Alzheimer's disease risk warning system based on handwriting recognition, an electronic device, a computer-readable storage medium and a computer program product, which can recognize handwriting data on dot code paper through a smart Bluetooth pen, transmit it to a mobile device via Bluetooth for real-time display, and further transmit the relevant information to the server for data analysis and storage, extract corresponding features based on the handwriting characteristics to calculate the risk of the disease, and generate an evaluation report to better achieve auxiliary diagnosis of Alzheimer's disease for the subject.

[0063] The details are described in detail through the following examples.

[0064] Based on this, the embodiment of the present application provides an Alzheimer's disease risk warning method based on handwriting recognition that can be implemented by an Alzheimer's disease risk warning device based on handwriting recognition, see Figure 1 The Alzheimer's disease risk warning method based on handwriting recognition specifically includes the following contents:

[0065] Step 100: When the evaluator holds a Bluetooth pen and fills out a diagnostic scale with a preset dot matrix code for assisting in diagnosing the subject's Alzheimer's disease, various handwriting recognition data are received in real time from the Bluetooth pen, wherein each set of the handwriting recognition data includes: the correspondence between the scale content data determined by the Bluetooth pen through real-time recognition of the dot matrix code and the handwriting data recorded in real time by the Bluetooth pen.

[0066] In one or more embodiments of this application, a diagnostic scale is typically completed by both a physician and a patient. In most cases, the physician asks questions based on the scale and fills out a rating scale based on the patient's responses, similar to a medical interview and record-keeping. When connection and drawing tasks are involved, the patient completes the scale. The term "evaluator" in this application may include the physician and / or the patient. However, in one or more embodiments of this application, the term "subject" refers solely to the patient.

[0067] In one or more embodiments of the present application, a Bluetooth pen is used as a data collection tool to collect handwriting data from subjects filling out Alzheimer's disease-related diagnostic scales for subsequent diagnostic analysis. Specifically, an existing smart Bluetooth pen capable of connecting to a client device via Bluetooth and transmitting data can be used. The Bluetooth pen can also recognize dot codes on paper with dot codes upon contact and can collect handwriting data in real time through its own preset sensors. These are all existing functions of existing Bluetooth pens and are not subject to application or improvement. Therefore, the Bluetooth pen architecture will not be described in detail.

[0068] It is understood that the handwriting recognition data includes: the scale content data determined by the Bluetooth pen through real-time recognition of the dot matrix code, the handwriting data recorded in real time by the Bluetooth pen, and the correspondence between the scale content data and the handwriting data. The scale content data may include: the dot matrix code currently contacted by the Bluetooth pen on the diagnostic scale, the question number and the page area information, and the correspondence between the three;

[0069] The handwriting data includes: handwriting morphological feature data, writing pressure value, writing pause time, and the corresponding relationship between the three; wherein the handwriting morphological data includes: a timestamp, a sequence of coordinate points corresponding to the current handwriting in the diagnostic scale, a stroke thickness value, and the corresponding relationship between the three. The preset coordinate system is a two-dimensional coordinate system (x, y).

[0070] It should also be noted that although a paper diagnostic scale with a preset dot matrix code is used in step 100 of this application in order to better improve the scale filling experience and convenience for the elderly, the Alzheimer's disease risk warning method based on handwriting recognition proposed in this application is also applicable to the Alzheimer's disease electronic diagnostic scale developed based on a digital tablet.

[0071] In one or more embodiments thereof, the diagnostic scale may be referred to simply as a scale.

[0072] Step 200: Calculate the diagnostic scale score of the subject based on the handwriting recognition data generated when the evaluator fills out all the diagnostic scales, and determine the first Alzheimer's disease risk level corresponding to the subject based on the diagnostic scale score.

[0073] In step 200, the subject may fill out multiple different types of diagnostic scales. Examples of the types of diagnostic scales are shown in Table 1. Specifically, all or multiple of these scales may be used as the diagnostic scales required to be filled out by the assessor in this application. Furthermore, the diagnostic scales that may be used in this application are a combination of a series of neuropsychological assessment scales. The versions and number of diagnostic scales may be added, deleted, or modified according to actual needs.

[0074] Table 1: Examples of diagnostic scale types

[0075]

[0076]

[0077] Each scale consists of multiple sub-items, and the total score of the sub-items is calculated to represent the score of the corresponding scale. This scale assesses various abilities, such as cognitive level, depression level, and memory. The Mini-Mental State Examination (MMSE) is used to assist in identifying dementia, while the Montreal Cognitive Assessment or Basic (MoCA or MoCA-B) and the Addenbrooke's Cognitive Examination, Third Edition (ACE-III) are used to identify mild cognitive impairment. Other commonly used tests, such as the Auditory Verbal Learning Test (AVLT), the Boston Naming Test (BNT), the Verbal Fluency Test (VFT), the Symbol Digit Modalities Test (SDMT), and the Trail Making Test (TMT), are used to assess impairment in memory, language, attention, and executive function, respectively.

[0078] And, step 300: according to the handwriting data in each of the handwriting recognition data, a preset Alzheimer's disease risk prediction model is used to determine the second Alzheimer's disease risk level corresponding to the subject.

[0079] In one or more embodiments of the present application, the Alzheimer's disease risk prediction model can adopt a classification model for outputting an Alzheimer's disease risk score according to the input handwriting data. Specifically, the existing random forest algorithm can be adopted, and it can be implemented using a support vector machine, XGBoost or other integrated algorithms. It can also be implemented by integrating deep learning algorithms such as convolutional neural networks, recurrent neural networks, and multiple algorithms. The specific selection is based on actual application needs. It is only necessary to adopt these existing model frameworks and use handwriting data historical samples and labels used to represent Alzheimer's disease risk scores for training during the training phase.

[0080] Step 400: Acquire Alzheimer's disease risk warning result data of the subject based on the first Alzheimer's disease risk level and the second Alzheimer's disease risk level.

[0081] In step 400, the first Alzheimer's disease risk level and the second Alzheimer's disease risk level can be directly output as the Alzheimer's disease risk warning result data of the subject, so as to assist the diagnostic doctor in making a more accurate and comprehensive assessment of the subject's Alzheimer's disease risk based on the two risk levels.

[0082] In order to further improve the convenience and reliability of auxiliary diagnostic doctors in assessing the Alzheimer's disease risk of the subjects, in another implementation of step 400, the existing large model can also be used to generate the first Alzheimer's disease risk level, the second Alzheimer's disease risk level, or the diagnostic analysis report text data corresponding to the first Alzheimer's disease risk level and the second Alzheimer's disease risk level after the data is summarized.

[0083] From the above description, it can be seen that the Alzheimer's disease risk warning method based on handwriting recognition provided in the embodiment of the present application can be applicable to the test population such as the elderly who are inconvenient to use electronic scales, and can ensure the efficiency and degree of automation of paper quality scale data processing, and can effectively improve the accuracy and reliability of Alzheimer's disease risk warning, thereby better realizing the auxiliary diagnosis of Alzheimer's disease.

[0084] In order to further improve the accuracy and effectiveness of calculating the diagnostic scale score of the subject, in the Alzheimer's disease risk warning method based on handwriting recognition provided in the embodiment of the present application, see Figure 2 Step 200 of the Alzheimer's disease risk warning method based on handwriting recognition specifically includes the following:

[0085] Step 210: According to the preset test task types to which each of the diagnostic scales belongs, the handwriting recognition data corresponding to the diagnostic scale whose test task type is a selective test task are divided into a selective test group, and the handwriting recognition data corresponding to the diagnostic scale whose test task type is a drawing test task are divided into a drawing test group.

[0086] Step 220: Based on the handwriting recognition data in the selective test group, the total score corresponding to each diagnostic scale belonging to the selective test task is obtained, and the total score corresponding to each diagnostic scale belonging to the selective test task is summarized to obtain the selective test score corresponding to the subject.

[0087] And, step 230: based on each of the handwriting recognition data in the drawing type test group, obtain the total score corresponding to each of the diagnostic scales belonging to the drawing type test task, and summarize the total score corresponding to each of the diagnostic scales belonging to the drawing type test task to obtain the drawing type test score corresponding to the subject.

[0088] Step 240: Determine the selective test score and the drawing test score corresponding to the subject as the diagnostic scale score of the subject, and determine the first Alzheimer's disease risk level corresponding to the subject according to the diagnostic scale score of the subject from the diagnostic standard data used to record the correspondence between the diagnostic scale score and the Alzheimer's disease risk level.

[0089] In order to further improve the accuracy and effectiveness of calculating the test subject's score for the selective test task, in the Alzheimer's disease risk warning method based on handwriting recognition provided in the embodiment of the present application, see Figure 3 Step 220 of the Alzheimer's disease risk warning method based on handwriting recognition specifically includes the following:

[0090] Step 221: Execute the preset candidate box selection judgment step for each of the handwriting recognition data corresponding to each of the diagnostic scales belonging to the selective test task to determine the score of each selected candidate box in each of the diagnostic scales belonging to the selective test task, and summarize the score of each selected candidate box in each of the diagnostic scales belonging to the selective test task to obtain the total score corresponding to each of the diagnostic scales belonging to the selective test task.

[0091] Among them, see Figure 4 The candidate box selection judgment step specifically includes the following contents:

[0092] Step 010: Determine the page identifier in the diagnostic scale corresponding to the handwriting recognition data based on the scale content data in the current handwriting recognition data, and obtain the position information of each candidate box in the target page corresponding to the page identifier from the candidate box information corresponding to each of the preset diagnostic scales belonging to the selective test task.

[0093] Step 020: Based on the handwriting data in the current handwriting recognition data and the position information of each candidate box in the target page, determine whether the distance between the handwriting data and any candidate box in the target page is less than a preset distance threshold. If so, determine that the candidate box whose distance to the handwriting data is less than the distance threshold is checked and execute step 030; if not, determine that the handwriting data is invalid data and complete the candidate box checking judgment step for the current handwriting recognition data.

[0094] Step 030: Determine the score of the selected candidate box from the scores corresponding to the respective candidate boxes stored in the candidate box information and complete the candidate box selection judgment step for the current handwriting recognition data.

[0095] Specifically, the Alzheimer's disease risk warning device based on handwriting recognition achieves scoring by matching the handwriting position with the coordinates of the score candidate boxes. First, the information of each candidate box used for score determination in the scale is determined, including its center coordinates, border size, corresponding score and number. Secondly, the handwriting data collected by the Bluetooth pen is converted into a series of data points with continuous coordinates and time information. The coordinates of its center point are obtained by taking the mean of the horizontal and vertical coordinates of the data point, and the distance between the center point of the handwriting and all candidate boxes on the current page is calculated. If the shortest distance is less than the sum of the side length of the candidate box and the error distance, the handwriting is determined to correspond to this candidate box. Otherwise, the handwriting is determined to be invalid and does not participate in the total score calculation. By repeatedly iterating all the handwriting on a set of subscales, the check information of all candidate boxes on this scale can be obtained. The scores corresponding to the checked candidate boxes are added to obtain the total score of the sub-item.

[0096] In order to further improve the accuracy and effectiveness of calculating the test score of the subject's drawing test task, in the Alzheimer's disease risk warning method based on handwriting recognition provided in the embodiment of the present application, see Figure 3 Step 230 of the Alzheimer's disease risk warning method based on handwriting recognition specifically includes the following:

[0097] Step 231: Convert the handwriting data corresponding to each of the handwriting recognition data in the drawing-type test group into handwriting image data.

[0098] Step 232: Input each of the handwriting image data into a preset handwriting image scoring model to obtain the scores corresponding to each of the handwriting recognition data in the drawing type test group, and obtain the total scores corresponding to each of the diagnostic scales belonging to the drawing type test task based on the scores corresponding to each of the handwriting recognition data in the drawing type test group.

[0099] The handwriting image scoring model can be implemented using a convolutional neural network, or it can be solved by using a convolutional neural network combined with enhanced training data generated by a generative adversarial network (GAN).

[0100] Specifically, for drawing-type test tasks, the Alzheimer's disease risk warning device based on handwriting recognition converts the handwriting data collected by the Bluetooth pen into image information for data processing. Some relatively intuitive and simple scoring tasks are solved using functions such as polygon recognition of OpenCV (a cross-platform computer vision library), while other difficult scoring tasks are solved using convolutional neural networks combined with enhanced training data generated by generative adversarial networks (GAN). When training the model, taking advantage of the Bluetooth pen's ability to collect multi-dimensional handwriting features, information such as handwriting pressure and handwriting fluency related to the scoring rules can be used as features for training, so that the trained model can obtain results that are closer to human evaluation than previous classification tasks. Among them, intuitive and simple scoring tasks and difficult scoring tasks can be manually specified in advance.

[0101] In order to further improve the accuracy and effectiveness of handwriting risk calculation, in the Alzheimer's disease risk warning method based on handwriting recognition provided in the embodiment of the present application, see Figure 3 Step 300 of the Alzheimer's disease risk warning method based on handwriting recognition specifically includes the following:

[0102] Step 310: Generate handwriting morphological feature image data corresponding to each handwriting recognition data according to the handwriting morphological feature data in the handwriting data in each handwriting recognition data.

[0103] Step 320: setting data samples corresponding to each of the handwriting recognition data, wherein each of the data samples includes the handwriting morphological feature image data, the writing pressure value, and the writing pause time corresponding to the handwriting recognition data.

[0104] Step 330: Input the data sample into a preset Alzheimer's disease risk prediction model so that the Alzheimer's disease risk prediction model outputs the Alzheimer's disease risk score of the subject; the Alzheimer's disease risk prediction model is pre-trained and generated by a random forest model.

[0105] Step 340: Determine a second Alzheimer's disease risk level corresponding to the subject according to the Alzheimer's disease risk score.

[0106] Specifically, the Alzheimer's disease risk warning device based on handwriting recognition obtains the handwriting of the assessee, constructs features such as the original image, pressure, writing time, and completion, and uses a trained classifier to obtain the probability that the assessee has Alzheimer's disease, which is used as a risk score to assist in disease decision-making. This classifier is a random forest model. When training the model, the training data must first be divided into a training set and a test set. Nested cross-validation and grid search are used to optimize the model's hyperparameters and select the optimal hyperparameter combination. The features input to the model include but are not limited to the above features. The machine learning algorithm can also be implemented using deep learning methods such as convolutional neural networks and recurrent neural networks, or integrated algorithms.

[0107] In order to further improve the reliability and effectiveness of the generation of Alzheimer's disease risk warning results, in the Alzheimer's disease risk warning method based on handwriting recognition provided in the embodiment of the present application, see Figure 3 Step 400 of the handwriting recognition-based Alzheimer's disease risk warning method specifically includes the following:

[0108] Step 410: Input the first Alzheimer's disease risk level and the second Alzheimer's disease risk level into a preset diagnostic analysis report generation model, so that the diagnostic analysis report generation model outputs a corresponding diagnostic analysis report, and outputs the diagnostic analysis report as the Alzheimer's disease risk warning result data of the subject.

[0109] The diagnostic analysis report generation model can be implemented using an existing large model, such as the moonshot (dark side of the moon) large model.

[0110] Specifically, the Alzheimer's disease risk warning device based on handwriting recognition can first take the first Alzheimer's disease risk level and the second Alzheimer's disease risk level as parameters, and encapsulate them in the JSON format required by the moonshot big model API interface; then call the API interface to send the JSON data to; the big model generates a diagnostic analysis report containing potential cognitive problem analysis and treatment recommendations based on the received data and a pre-set template, and returns the results in JSON format; parses the returned JSON data, extracts keywords and corresponding content, and stores them in the corresponding location in the pre-made Excel template; finally, converts the completed Excel table into PDF format for storage, and feeds the generated PDF report back to the front end for the experimenter and the subject to provide a reference for subsequent diagnosis and treatment.

[0111] In order to further improve the accuracy and application efficiency of Alzheimer's disease risk warning based on handwriting recognition, in the Alzheimer's disease risk warning method based on handwriting recognition provided in the embodiment of the present application, see Figure 3 The Alzheimer's disease risk warning method based on handwriting recognition may further include the following contents after step 400:

[0112] Step 500: Classify and store the various handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects in a relational database, extract the various handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects from the relational database, and use the various handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects extracted from the relational database as training data to train the Alzheimer's disease risk prediction model or the handwriting image scoring model.

[0113] Specifically, the data stored in the Alzheimer's disease risk warning device based on handwriting recognition may include original handwriting, calculated total scale scores, and filled-in personal information. The above data are sorted according to different task types, scale types and other conditions, divided into multiple data tables, and stored in a MySQL relational database. This partitioned table storage method not only helps to quickly retrieve the required data, but also facilitates the extraction of different features for different types of tasks. For example, drawing tasks can extract features such as the smoothness of lines and the regularity of shapes, and multiple-choice tasks can extract features such as the accuracy of the check position, thereby improving the efficiency and accuracy of model training. In addition, it can also reduce data redundancy, avoid storing duplicate information, and save storage space.

[0114] The present application also provides an Alzheimer's disease risk warning device based on handwriting recognition for executing all or part of the Alzheimer's disease risk warning method based on handwriting recognition, see Figure 5 The Alzheimer's disease risk warning device based on handwriting recognition specifically includes the following contents:

[0115] The data acquisition module 10 is used to receive various handwriting recognition data in real time from the Bluetooth pen when the evaluator holds the Bluetooth pen and fills out a diagnostic scale with a preset dot matrix code for assisting in diagnosing Alzheimer's disease of the subject. Each set of the handwriting recognition data includes: the correspondence between the scale content data determined by the Bluetooth pen through real-time recognition of the dot matrix code and the handwriting data recorded in real time by the Bluetooth pen.

[0116] The data acquisition module 10 involves a smart Bluetooth pen, a mobile client, and a matching Alzheimer's disease-related scale with a dot matrix code. Users use the smart Bluetooth pen to complete an Alzheimer's disease-related diagnostic scale on paper printed with a specific dot matrix code. During the writing process, the smart Bluetooth pen can recognize the dot matrix code on the paper, determine the page currently being written on, the question number, and the page area information (i.e., each page is pre-divided into multiple different areas), and thus associate handwriting data with the scale content. Furthermore, the smart Bluetooth pen's built-in sensor records handwriting data in real time, including: the timestamp of the writing process, the (x, y) sequence of the handwriting points, the thickness of the strokes, etc., for reconstructing the handwriting form; the writing pressure value, the pause time, etc., for reflecting the strength and fluency of the writing. This information fully represents the user's writing characteristics on the paper. As the user writes with the Bluetooth pen, this information is transmitted in real time to the mobile client, where the scale base image and the corresponding writing trajectory are displayed in real time. The mobile client device can be an Android tablet or other similar device.

[0117] The scale score calculation module 20 is used to calculate the diagnostic scale score of the subject based on the various handwriting recognition data generated when the evaluator fills out all the diagnostic scales, and determine the first Alzheimer's disease risk level corresponding to the subject based on the diagnostic scale score.

[0118] The scale score calculation module 20 can automatically calculate the scores of each task after the scale test is completed.

[0119] For selective test tasks, the scale score calculation module 20 achieves scoring by matching the handwriting position with the coordinates of the score candidate box. First, determine the information of each candidate box used for score determination in the scale, including its center coordinates, frame size, corresponding score and number. Secondly, convert the handwriting data collected by the Bluetooth pen into a series of data points with continuous coordinates and time information. Take the mean of the horizontal and vertical coordinates of the data point to obtain the coordinates of its center point, calculate the distance between the center point of the handwriting and all candidate boxes on the current page, and if the shortest distance is less than the sum of the side length of the candidate box and the error distance, then the handwriting is determined to correspond to this candidate box, otherwise the handwriting is determined to be invalid handwriting and does not participate in the total score calculation. By repeatedly iterating all the handwriting on a set of subscales, the check information in all the candidate boxes on this scale can be obtained. Add up the scores corresponding to the checked candidate boxes to get the total score of the sub-item.

[0120] For drawing test tasks, the scale score calculation module 20 converts the handwriting data collected by the Bluetooth pen into image information for data processing. Some relatively intuitive and simple scoring tasks are solved using functions such as polygon recognition of opencv (a cross-platform computer vision library), and another part of the difficult scoring tasks is solved using convolutional neural networks combined with enhanced training data generated by generative adversarial networks (GAN). When training the model, taking advantage of the Bluetooth pen's ability to collect multi-dimensional handwriting features, we train the handwriting pressure, handwriting fluency and other information related to the scoring rules as features, so that the trained model can obtain results closer to human evaluation than previous classification tasks.

[0121] Finally, the scale score calculation module 20 can quantify the scores based on pre-acquired authoritative diagnostic standard data, and automatically classify and display the disease risk levels.

[0122] And, the handwriting risk calculation module 30 is used to determine the second Alzheimer's disease risk level corresponding to the subject according to the handwriting data in each of the handwriting recognition data using a preset Alzheimer's disease risk prediction model.

[0123] After acquiring the examinee's handwriting, the handwriting risk calculation module 30 constructs features such as the original image, pressure, writing time, and degree of completion. Using a pre-trained classifier, it determines the probability of the examinee having Alzheimer's disease, which serves as a risk score to assist in disease decision-making. This classifier is a random forest model. Model training requires first dividing the training data into a training set and a test set. Nested cross-validation and grid search are then used to optimize the model's hyperparameters and select the optimal hyperparameter combination. The input model features include, but are not limited to, those listed above. The machine learning algorithm can also be implemented using deep learning methods such as convolutional neural networks and recurrent neural networks, or using integrated algorithms.

[0124] The risk warning result acquisition module 40 is used to obtain the Alzheimer's disease risk warning result data of the subject based on the first Alzheimer's disease risk level and the second Alzheimer's disease risk level.

[0125] Among them, the risk warning result acquisition module 40 can also be called the assessment result analysis module. First, the calculated total scores of each subscale are used as parameters and encapsulated in the JSON format required by the moonshot large model API interface; then the API interface is called to send the JSON data to; the large model generates a diagnostic analysis report containing potential cognitive problem analysis and treatment recommendations based on the received data and the pre-set template, and returns the results in JSON format; parses the returned JSON data, extracts keywords and corresponding content, and stores them in the corresponding location in the pre-made Excel template; finally, the completed Excel table is converted into PDF format for storage, and the generated PDF report is fed back to the front end for the experimenter and the subject to provide a reference for subsequent diagnosis and treatment.

[0126] The embodiment of the Alzheimer's disease risk warning device based on handwriting recognition provided in this application can be specifically used to execute the processing flow of the embodiment of the Alzheimer's disease risk warning method based on handwriting recognition in the above-mentioned embodiment. Its functions will not be repeated here, and reference can be made to the detailed description of the above-mentioned embodiment of the Alzheimer's disease risk warning method based on handwriting recognition.

[0127] The part of the Alzheimer's disease risk warning device based on handwriting recognition that performs the Alzheimer's disease risk warning based on handwriting recognition can be executed in the server or completed in the client device. The specific selection can be based on the processing power of the client device and the limitations of the user's usage scenario. This application is not limited to this. If all operations are completed in the client device, the client device may also include a processor for the specific processing of the Alzheimer's disease risk warning based on handwriting recognition.

[0128] The client device may include a communication module (i.e., a communication unit) that can establish a communication connection with a remote server to implement data transmission with the server. The server may include a server on the task scheduling center side, and in other implementation scenarios, may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, a server cluster consisting of multiple servers, or a server structure of a distributed device.

[0129] The server and the client device may communicate using any suitable network protocol, including network protocols that have not yet been developed as of the filing date of this application. Examples of such network protocols include TCP / IP, UDP / IP, HTTP, and HTTPS. Furthermore, examples of such network protocols include RPC (Remote Procedure Call Protocol) and REST (Representational State Transfer) protocols, which are used on top of the aforementioned protocols.

[0130] From the above description, it can be seen that the Alzheimer's disease risk warning device based on handwriting recognition provided in the embodiment of the present application can be suitable for subjects such as the elderly who are not convenient to use electronic scales, and can ensure the efficiency and automation of paper quality scale data processing, and can effectively improve the accuracy and reliability of Alzheimer's disease risk warning, thereby better realizing auxiliary diagnosis of Alzheimer's disease for the subjects.

[0131] In order to further improve the accuracy and application efficiency of Alzheimer's disease risk warning based on handwriting recognition, in an Alzheimer's disease risk warning device based on handwriting recognition provided in an embodiment of the present application, see Figure 6 The Alzheimer's disease risk warning device based on handwriting recognition may further specifically include the following contents:

[0132] The data storage module 50 is used to classify and store the various handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects in a relational database, to extract the various handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects from the relational database, and to use the various handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects extracted from the relational database as training data to train the Alzheimer's disease risk prediction model or the handwriting image scoring model.

[0133] Specifically, data storage module 50 stores data including original handwriting, calculated total scale scores, and personal information. The above data is organized according to different task types, scale types, and other conditions, divided into multiple data tables, and stored in a MySQL relational database. This table-based storage method not only facilitates rapid retrieval of required data, but also facilitates the extraction of different features for different types of tasks. For example, drawing tasks can extract features such as line smoothness and shape regularity, while multiple-choice tasks can extract features such as the accuracy of checkbox placement, thereby improving the efficiency and accuracy of model training. It can also reduce data redundancy, avoid storing duplicate information, and save storage space.

[0134] That is to say, the Alzheimer's disease risk warning method and device based on handwriting recognition provided in the embodiment of the present application uses a smart Bluetooth pen as a data collection tool to collect handwriting data of the Alzheimer's disease-related diagnostic scale filled out by the subject for subsequent diagnostic analysis; calculates the score of each evaluation indicator in multiple scales based on the handwriting data, and calculates the total score of the scale; the handwriting data, total score of the scale and the case information of the evaluator are rated and classified according to the severity of the situation; extracts the relevant features of the evaluator's handwriting data, uses machine learning to obtain the evaluator's risk score for Alzheimer's disease, and generates an evaluation report in combination with the scale results for auxiliary diagnosis.

[0135] Compared with the prior art, the Alzheimer's disease risk warning method and device based on handwriting recognition provided by the embodiments of the present application have the following beneficial effects:

[0136] 1. Electronic scale: This does not change the paper-based Alzheimer's disease diagnosis process; instead, the pen and paper are replaced with a smart Bluetooth pen and paper with dot-matrix codes. This makes it easy to promote and use. Mobile-based data collection also facilitates remote diagnosis of Alzheimer's disease.

[0137] 2. Automatic calculation and storage of scores: effectively simplifying the operating process, reducing manual operations, and efficiently tracing and sharing data resources.

[0138] 3. Handwriting-based auxiliary recognition: Previous paper-based quality charts cannot effectively utilize handwriting data. The advantage of this system is that it can automatically evaluate the user's handwriting behavior characteristics, automatically determine their risk of Alzheimer's disease, and assist in disease screening.

[0139] 4. Assessment report generation: Based on the evaluator's scale assessment results and handwriting behavior analysis, an assessment report is generated that explains the patient's condition in detail, making it easier for the patient to understand their own condition.

[0140] Based on the above-mentioned Alzheimer's disease risk warning method and device based on handwriting recognition, the present application also provides an embodiment of an Alzheimer's disease risk warning system based on handwriting recognition. The provided Alzheimer's disease risk warning system based on handwriting recognition specifically includes the following contents:

[0141] A diagnostic scale with a preset dot matrix code for assisting in diagnosing Alzheimer's disease in a subject, a Bluetooth pen, and a mobile client device; the Bluetooth pen and the mobile client device are in communication connection;

[0142] The mobile client device is used to execute the Alzheimer's disease risk warning method based on handwriting recognition provided by the aforementioned embodiment;

[0143] The Bluetooth pen is provided with a sensor for collecting the handwriting data in real time.

[0144] The developed system can include an Android client, a web page display terminal and a data storage terminal, which are respectively used to receive Bluetooth pen data, display and modify data, and store data.

[0145] That is to say, the embodiment of the present application designs an electronic scale system based on a smart Bluetooth pen. Compared with the current situation where there is no electronic scale solution for Alzheimer's disease related to a smart pen, the present application has developed a set of Android, web and server-side software systems based on the Bluetooth pen; it can obtain the score items corresponding to the handwriting by matching the candidate boxes with coordinates; it can automatically calculate the scale scores; it can use handwriting data to assist in the diagnosis of Alzheimer's disease, extract the handwriting of the evaluator, and use machine learning methods to automatically determine whether he has the risk of Alzheimer's disease, to assist in decision-making; and it can generate an evaluation report based on a large model.

[0146] The present application also provides an electronic device that may include a processor, a memory, a receiver, and a transmitter. The processor is configured to execute the Alzheimer's disease risk warning method based on handwriting recognition described in the above embodiment. The processor and the memory may be connected via a bus or other means, with bus connection being used as an example. The receiver may be connected to the processor and the memory via a wired or wireless means.

[0147] The processor may be a central processing unit (CPU). The processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0148] The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the handwriting recognition-based Alzheimer's disease risk warning method in the embodiments of the present application. The processor executes the non-transitory software programs, instructions, and modules stored in the memory to perform various processor functions and data processing, thereby implementing the handwriting recognition-based Alzheimer's disease risk warning method in the above-mentioned method embodiments.

[0149] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely located relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0150] The one or more modules are stored in the memory, and when executed by the processor, perform the Alzheimer's disease risk warning method based on handwriting recognition in the embodiment.

[0151] In some embodiments of the present application, the user equipment may include a processor, a memory and a transceiver unit, and the transceiver unit may include a receiver and a transmitter. The processor, memory, receiver and transmitter may be connected through a bus system. The memory is used to store computer instructions, and the processor is used to execute the computer instructions stored in the memory to control the transceiver unit to send and receive signals.

[0152] As an implementation method, the functions of the receiver and transmitter in this application can be considered to be implemented through a transceiver circuit or a dedicated transceiver chip, and the processor can be considered to be implemented through a dedicated processing chip, a processing circuit or a general-purpose chip.

[0153] As another implementation method, it is possible to use a general-purpose computer to implement the server provided in the embodiments of the present application. That is, the program code for implementing the functions of the processor, receiver, and transmitter is stored in a memory, and the general-purpose processor implements the functions of the processor, receiver, and transmitter by executing the code in the memory.

[0154] The present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the aforementioned Alzheimer's disease risk warning method based on handwriting recognition. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a floppy disk, a hard disk, a removable storage disk, a CD-ROM, or any other form of storage medium known in the art.

[0155] An embodiment of the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the Alzheimer's disease risk warning method based on handwriting recognition.

[0156] It should be understood by those skilled in the art that the various exemplary components, systems and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software or a combination of the two. Whether it is specifically performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application. When implemented in hardware, it can be, for example, an electronic circuit, an application-specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of this application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier.

[0157] It should be understood that the present application is not limited to the specific configurations and processes described above and illustrated in the figures. For the sake of brevity, a detailed description of known methods is omitted here. In the above embodiments, several specific steps are described and illustrated as examples. However, the method process of the present application is not limited to the specific steps described and illustrated. Those skilled in the art can make various changes, modifications, and additions, or change the order of the steps after understanding the spirit of the present application.

[0158] In this application, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or replace features of other embodiments.

[0159] The above description is merely a preferred embodiment of the present application and is not intended to limit the present application. Those skilled in the art will appreciate that various modifications and variations of the present embodiment are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A method for early warning of Alzheimer's disease risk based on handwriting recognition, characterized in that: include: When an assessor uses a Bluetooth pen to fill out a diagnostic scale with a preset dot matrix code for assisting in diagnosing a subject's Alzheimer's disease, the Bluetooth pen receives in real time various handwriting recognition data, wherein each set of the handwriting recognition data includes: a correspondence between scale content data determined by the Bluetooth pen through real-time recognition of the dot matrix code and handwriting data recorded in real time by the Bluetooth pen; the scale content data includes information about the page, question number, and page area of ​​the diagnostic scale on which the dot matrix code currently contacts the Bluetooth pen; Calculating the test subject's diagnostic scale score based on each of the handwriting recognition data generated when the evaluator fills out all the diagnostic scales, and determining a first Alzheimer's disease risk level corresponding to the test subject based on the diagnostic scale score; and, determining a second Alzheimer's disease risk level corresponding to the subject using a preset Alzheimer's disease risk prediction model based on the handwriting data in each of the handwriting recognition data; Acquire Alzheimer's disease risk warning result data of the subject based on the first Alzheimer's disease risk level and the second Alzheimer's disease risk level; The step of calculating the test subject's diagnostic scale score based on the handwriting recognition data generated when the evaluator fills out all the diagnostic scales, and determining the first Alzheimer's disease risk level corresponding to the test subject based on the diagnostic scale score, includes: According to the preset test task types to which the respective diagnostic scales belong, the handwriting recognition data corresponding to the diagnostic scales whose test task type is a selective test task are divided into a selective test group, and the handwriting recognition data corresponding to the diagnostic scales whose test task type is a drawing test task are divided into a drawing test group; Executing a preset candidate box selection judgment step for each of the handwriting recognition data corresponding to each of the diagnostic scales belonging to the selective test task, respectively, to determine the score of each selected candidate box in each of the diagnostic scales belonging to the selective test task, and summarizing the score of each selected candidate box in each of the diagnostic scales belonging to the selective test task to obtain the total score corresponding to each of the diagnostic scales belonging to the selective test task, and summarizing the total score corresponding to each of the diagnostic scales belonging to the selective test task to obtain the selective test score corresponding to the subject; and, based on each of the handwriting recognition data in the drawing-type test group, obtaining a total score corresponding to each of the diagnostic scales belonging to the drawing-type test task, and summarizing the total score corresponding to each of the diagnostic scales belonging to the drawing-type test task to obtain a drawing-type test score corresponding to the subject; The selective test score and the drawing test score corresponding to the subject are determined as the diagnostic scale score of the subject, and the first Alzheimer's disease risk level corresponding to the subject is determined according to the diagnostic scale score of the subject from the diagnostic standard data used to record the correspondence between the diagnostic scale score and the Alzheimer's disease risk level.

2. The Alzheimer's disease risk warning method based on handwriting recognition according to claim 1, characterized in that: The handwriting data includes: handwriting morphological feature data, writing pressure value and writing pause time; wherein, the handwriting morphological feature data includes: timestamp, coordinate point sequence of the current handwriting in the preset coordinate system corresponding to the diagnostic scale and stroke thickness value.

3. The Alzheimer's disease risk warning method based on handwriting recognition according to claim 1, characterized in that: The candidate box selection judgment step includes: Determine, based on the scale content data in the current handwriting recognition data, a page identifier in the diagnostic scale corresponding to the handwriting recognition data, and obtain, from the candidate box information corresponding to each of the preset diagnostic scales belonging to the selective test task, position information of each candidate box in the target page corresponding to the page identifier; Based on the handwriting data in the current handwriting recognition data and the position information of each candidate box in the target page, determining whether the distance between the handwriting data and any candidate box in the target page is less than a preset distance threshold, and if so, determining that the candidate box whose distance to the handwriting data is less than the distance threshold is selected; The score of the selected candidate box is determined based on the scores corresponding to the respective candidate boxes stored in the candidate box information.

4. The Alzheimer's disease risk warning method based on handwriting recognition according to claim 1, characterized in that: The obtaining of the total score corresponding to each diagnostic scale belonging to the drawing type test task based on each of the handwriting recognition data in the drawing type test group includes: converting the handwriting data corresponding to each of the handwriting recognition data in the drawing type test group into handwriting image data respectively; Each of the handwriting image data is respectively input into a preset handwriting image scoring model to obtain the score corresponding to each of the handwriting recognition data in the drawing type test group, and based on the score corresponding to each of the handwriting recognition data in the drawing type test group, the total score corresponding to each of the diagnostic scales belonging to the drawing type test task is obtained.

5. The Alzheimer's disease risk warning method based on handwriting recognition according to claim 2, characterized in that: Determining the second Alzheimer's disease risk level corresponding to the subject using a preset Alzheimer's disease risk prediction model based on the handwriting data in each of the handwriting recognition data includes: generating handwriting morphological feature image data corresponding to each of the handwriting recognition data according to the handwriting morphological feature data in the handwriting data in the handwriting recognition data; Setting data samples corresponding to each of the handwriting recognition data, wherein each of the data samples includes the handwriting morphological feature image data, the writing pressure value, and the writing pause time corresponding to the handwriting recognition data; Inputting the data sample into a preset Alzheimer's disease risk prediction model so that the Alzheimer's disease risk prediction model outputs an Alzheimer's disease risk score for the subject; the Alzheimer's disease risk prediction model is previously trained and generated by a random forest model; A second Alzheimer's disease risk level corresponding to the subject is determined according to the Alzheimer's disease risk score.

6. The Alzheimer's disease risk warning method based on handwriting recognition according to claim 1, characterized in that: The step of obtaining the Alzheimer's disease risk warning result data of the subject based on the first Alzheimer's disease risk level and the second Alzheimer's disease risk level includes: The first Alzheimer's disease risk level and the second Alzheimer's disease risk level are input into a preset diagnostic analysis report generation model so that the diagnostic analysis report generation model outputs a corresponding diagnostic analysis report, and the diagnostic analysis report is output as the Alzheimer's disease risk warning result data of the subject.

7. The Alzheimer's disease risk warning method based on handwriting recognition according to claim 4, characterized in that: Also includes: The handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects are classified and stored in a relational database, and the handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects are extracted from the relational database. The handwriting recognition data generated when the evaluator fills in all the diagnostic scales, the diagnostic scale scores of the subjects, and the pre-acquired personal information of the subjects are extracted from the relational database as training data to train the Alzheimer's disease risk prediction model or the handwriting image scoring model.

8. An Alzheimer's disease risk warning device based on handwriting recognition, characterized in that: include: A data acquisition module is configured to receive, in real time, individual handwriting recognition data from a Bluetooth pen while an assessor is filling out a diagnostic scale with a preset dot matrix code used to assist in diagnosing a subject's Alzheimer's disease using the Bluetooth pen. Each set of handwriting recognition data includes: a correspondence between scale content data determined by the Bluetooth pen through real-time recognition of the dot matrix code and handwriting data recorded in real time by the Bluetooth pen; the scale content data includes information about the page, question number, and page area of ​​the diagnostic scale with the dot matrix code currently contacted by the Bluetooth pen; a scale score calculation module, configured to calculate the diagnostic scale score of the test subject based on the handwriting recognition data generated when the evaluator fills out all the diagnostic scales, and determine the first Alzheimer's disease risk level corresponding to the test subject based on the diagnostic scale score; and a handwriting risk calculation module, configured to determine a second Alzheimer's disease risk level corresponding to the subject using a preset Alzheimer's disease risk prediction model based on the handwriting data in each of the handwriting recognition data; a risk warning result acquisition module, configured to acquire Alzheimer's disease risk warning result data of the subject based on the first Alzheimer's disease risk level and the second Alzheimer's disease risk level; The step of calculating the test subject's diagnostic scale score based on the handwriting recognition data generated when the evaluator fills out all the diagnostic scales, and determining the first Alzheimer's disease risk level corresponding to the test subject based on the diagnostic scale score, includes: According to the preset test task types to which the respective diagnostic scales belong, the handwriting recognition data corresponding to the diagnostic scales whose test task type is a selective test task are divided into a selective test group, and the handwriting recognition data corresponding to the diagnostic scales whose test task type is a drawing test task are divided into a drawing test group; Executing a preset candidate box selection judgment step for each of the handwriting recognition data corresponding to each of the diagnostic scales belonging to the selective test task, respectively, to determine the score of each selected candidate box in each of the diagnostic scales belonging to the selective test task, and summarizing the score of each selected candidate box in each of the diagnostic scales belonging to the selective test task to obtain the total score corresponding to each of the diagnostic scales belonging to the selective test task, and summarizing the total score corresponding to each of the diagnostic scales belonging to the selective test task to obtain the selective test score corresponding to the subject; and, based on each of the handwriting recognition data in the drawing-type test group, obtaining a total score corresponding to each of the diagnostic scales belonging to the drawing-type test task, and summarizing the total score corresponding to each of the diagnostic scales belonging to the drawing-type test task to obtain a drawing-type test score corresponding to the subject; The selective test score and the drawing test score corresponding to the subject are determined as the diagnostic scale score of the subject, and the first Alzheimer's disease risk level corresponding to the subject is determined according to the diagnostic scale score of the subject from the diagnostic standard data used to record the correspondence between the diagnostic scale score and the Alzheimer's disease risk level.

9. An Alzheimer's disease risk warning system based on handwriting recognition, characterized in that: include: A diagnostic scale, a Bluetooth pen, and a mobile client device provided with a preset dot matrix code for assisting in diagnosing Alzheimer's disease in a subject; The Bluetooth pen is in communication with the mobile client device; The mobile client device is used to execute the Alzheimer's disease risk warning method based on handwriting recognition according to any one of claims 1 to 7; The Bluetooth pen is provided with a sensor for collecting the handwriting data in real time.

Images