Intelligent and integrated I-DID system for online evaluation and intervention of Chinese reading disorder based on strong and weak items

Through an intelligent multimodal dyslexia assessment and intervention system, combining strengths and weaknesses, the problem that existing tools cannot fully evaluate creativity, music and social emotional skills is solved, and accurate assessment and personalization of children's multi-dimensional abilities are achieved. Intervention, early identification of dyslexia, improve assessment efficiency and child development.

CN119092091APending Publication Date: 2024-12-06THE UNIVERSITY OF HONG KONG
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410654996.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2023-06-05
Filing Date
2024-05-24
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

Existing dyslexia assessment and intervention tools fail to fully identify children’s creativity, music and social emotional skills, ignore family background, and have a single, expensive approach to assessment and inability to diagnose potential dyslexia early, resulting in children’s failure to fully utilize their potential.

Method used

An intelligent integrated dyslexia assessment and intervention system has been developed, using multimodal displays and artificial intelligence to provide assessments based on Chinese, English and Chinese-English bilingualism through multi-dimensional assessment and personalized intervention, combining strengths and weaknesses, and Treatment options, support mobile devices, enable online assessment and intervention.

Benefits of technology

It has achieved accurate assessment and personalized intervention for children's multi-dimensional ability, early identification of reading disorders, reduced testing anxiety, improved assessment efficiency, and applied to different age groups to promote the all-round development of children.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119092091A_ABST
    Figure CN119092091A_ABST
Patent Text Reader

Abstract

The I-DID is an intelligent integrated system based on strong-weak reading disorder assessment and intervention, and is suitable for children, middle school students and adults with reading disorder of Chinese monolingual, English monolingual and Chinese-English bilingual users. An integrated database of the system comprises five fields: language and read-write, cognition, creativity, music and social emotion skills. A subgroup generator based on artificial intelligence machine learning can create differently characterized subgroups, each of which is recommended and provided with a personalized intervention, such that users in different subgroups receive a read-write capability intervention treatment based on their strong-weak characteristics. Intervention measures include general cognitive and music skill training in one or more fields, or listening, speaking, reading, writing or combination training in language-specific fields. The method of combining the theory of the I-DID system and data driving integrates multi-layer online multi-aspect evaluation, and classifies strong and weak characteristics of monolingual and bilingual users, so as to carry out self-management and personalized game-based intervention treatment.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to an intelligent / integrated dyslexic interface design (I-DID), which is a multi-dimensional and multi-modal evaluation and intervention system for Chinese, English and Chinese-English bilingual dyslexia for children aged 4 to 12, middle school students aged 12 to 18, and adults over 18 and their parents. I-DID uses a parallel Chinese-English bilingual and strong-weak combination approach to divide children's behavior patterns into five areas, including language and literacy, cognition, creativity, music and social-emotional skills, and conducts systematic and continuous longitudinal tracking and evaluation, thereby creating a growth and development curve and profile based on the five dimensions of personal cognition, emotion, language and reading, creativity, and music, and providing personalized intervention treatment plans. Background Art

[0002] Dyslexia is one of the most common learning disabilities among school-age children, who usually have normal intellectual development levels but have persistent difficulties in learning to read and spell. Dyslexia is widespread, affecting 5%-10% of the population in all language areas. Although researchers in different countries have developed various tools to identify cognitive and language deficits in children with dyslexia, all current reading assessments and intervention treatment programs are based on the deficits of dyslexia rather than its strengths. This may cause children, parents, and teachers to emphasize the deficiencies of children with dyslexia and ignore their underutilized talents and undeveloped potential. In addition, currently available assessment, diagnosis, and intervention tools focus on: (1) single language, ignoring the fact that most school-age children are learning to read and write in both Chinese and English; (2) only assessing the child’s own reading, writing, and cognitive abilities, without considering the direct relationship between the parents’ language, reading, writing, and cognitive abilities and the development of the child’s reading and writing abilities; (3) only assessing and testing language, reading, writing, and cognition, while ignoring the development and cultivation of creativity, music, and social-emotional abilities of children with potential dyslexia; (4) only assessing the child’s behavior once, without considering changes in the child’s growth and development; and (5) being expensive and inconvenient. The face-to-face manual service delivery model limits the ability to simultaneously assess a large number of children at risk of dyslexia; (6) the content and format of the assessment are monotonous, similar to children's school classroom tests and examinations, which causes certain test anxiety and pressure for children with potential dyslexia who are being assessed, reducing their interest and motivation to participate in assessment and treatment; (7) the assessment is oriented towards children's academic performance and cannot diagnose the specific difficulties children have in the learning process; and (7) it is only designed for preschool children and there is no systematic and objective assessment of dyslexia in preschool children, so the time for prevention and optimal intervention for children with a family history of dyslexia may be missed.

[0003] In view of the inadequacy of current assessment, diagnosis and treatment tools for dyslexia, I-DID has taken into account the following key factors that have been proven by a large amount of experimental evidence in designing its multilingual dyslexia screening and intervention tools: (1) the comprehensiveness and effectiveness of strengths-based assessment, diagnosis and intervention measures; (2) the close connection between children at risk of dyslexia and their parents' speech, reading and literacy skills; (3) the potential strengths and talents of dyslexic children in non-verbal areas such as creativity, music and social-emotional skills; (4) the age of the assessees, such as preschool or elementary school-age children and adults; (5) the real-time nature of the assessment test and the flexibility of the intervention treatment; the ability to conduct online assessment, diagnosis and intervention treatment without time and place restrictions; (6) the direct connection between the assessment and treatment content and the students' daily life, and to make it interesting and gamified. And (7) the use of the latest models and technologies of artificial intelligence to better predict the growth stages of cognitive, creativity, speech and literacy, music, and social-emotional development abilities of dyslexic children and normally developing children. Therefore, it is very important to integrate these key factors into a system of assessment and intervention training. Only by better integrating these aspects into an assessment and intervention training system can we more comprehensively and accurately identify the underutilized talents and undeveloped potential of children with dyslexia, so that new opportunities can be created and these children can be integrated into the community. In addition, a system that integrates these factors can also timely diagnose and provide early intervention for dyslexia.

[0004] Therefore, there is an urgent need for an intelligent tool that integrates the assessment and intervention of strengths and weaknesses based on individual development to conduct early screening, diagnosis and systematic intervention for children at risk of dyslexia. Such a tool needs to comprehensively consider the factors mentioned above, especially the creativity, music, and social-emotional potential strengths and talents of children in addition to verbal literacy, age and family background. In order to address this need in education and clinical practice, the I-DID invention not only provides the possibility of systematic and comprehensive assessment and diagnosis of the development of strengths and weaknesses of early children and early intervention combining strengths and weaknesses, but also creates new opportunities for children with dyslexia to successfully integrate into the group. Summary of the invention

[0005] One object of the present invention is to provide an intelligent, mobile-friendly, integrated Chinese, English, and Chinese-English dyslexia assessment, identification, and treatment intervention system, based on an intelligent classification method that integrates strengths and weaknesses, to address the above-mentioned limitations of existing reading diagnostic tools and the unmet needs of children with dyslexia. The system includes computer hardware, servers, and software for single or multiple users to assess their language and literacy, cognitive, creativity, music, and social-emotional skills. The system is controlled by an intelligent analysis engine and is described as follows: (1) A test and scoring interface is provided for single or multiple users through a multi-module and multimedia display, including a multi-dimensional assessment test educational game and a learning game for intervention therapy. (2) The response performance of users who complete all items in each game stage is summarized so that the system generates scores in five areas for each user; (3) Once the artificial intelligence-assisted screening model describes the performance scores of individual users in the five areas, the individual users are divided into different subgroups of strengths and weaknesses; (4) For each subgroup, personalized adaptive intervention based on personal data is provided to individual users, so that users of different subgroups can receive different intervention plans through multi-module multimedia games. Specifically, the intelligent, digital, mobile device-enabled intervention program can be implemented through music and simulated narrative assisted training in reading, writing, listening, speaking, or a combination thereof for individuals and multiple users.

[0006] According to a first aspect of the present invention, an intelligent digital education and clinical tool / teaching aid for strength-weakness-based assessment and intervention is developed. The teaching aid includes a multi-module multimedia display, a program model, and a display model. The display is configured to provide a scoring interface to individual and group users. The program model is configured to digitally communicate with the display and score one or more user performances and answers at each task stage. Then, the program model summarizes the scores of users who have completed all assessment games in each of the five areas (language and literacy, cognition, creativity, music, and social-emotional skills). The screening model is configured to digitally communicate with the program model and screen users according to their scores in various areas, so that users are automatically classified into one or more subgroups. The program model is also configured to provide personalized, adaptive interventions that combine strengths and weaknesses for each subgroup, so that users in different subgroups receive different intervention programs through a multimedia gamification display. The intervention plan is implemented by training participants in music or simulated narrative, gamification-based reading, writing, listening, speaking, or a combination thereof.

[0007] Through the embodiments of the present invention, through the combination of theory-driven and data-driven, the I-DID assessment and intervention tool can optimize and most accurately identify children at risk of dyslexia, and accurately reflect the subgroups of similarities and differences between neurodiverse users and neurotypical users. Through cross-validation, and using hierarchical clustering and validation, the reliability and validity of the subgroups can be further enhanced. Therefore, participants in the corresponding subgroups can receive adaptive interventions based on their individual strengths and weaknesses to improve and optimize the effectiveness of the intervention treatment.

[0008] In one aspect, the present invention provides an intelligent and integrated system for clinical assessment and intervention of dyslexia based on strengths and weaknesses. The system includes a database, which includes a large-scale data archive of multiple user performance sets across different domains. The first subgroup generator performs a theory-driven machine learning method on the database by obtaining a layer with nodes, wherein the nodes are calculated by summing the factor loadings of all component values ​​multiplied by the component values ​​using exploratory factor analysis, wherein subgroups are generated based on whether the nodes are activated. The second subgroup generator performs a theory- and data-driven integrated machine learning method on the database by training a neural network through an encoder and a decoder applying principal component analysis (PCA). The automatic encoder and decoder convert the data set of each user into a three-dimensional vector for classifying users by hierarchical clustering. This process can group users according to the user similarity or their distance calculated by the algorithm.

[0009] The encoder converts each user’s dataset into a three-dimensional vector, which is used to classify the users by hierarchical clustering. This process groups users based on the similarity or distance calculated between them.

[0010] The comparator compares the theory-driven machine learning method with the theory- and data-driven integrated machine learning method to determine whether the theory-driven machine learning method is consistent with the theory- and data-driven integrated machine learning method. The scoring module calculates the scores of various domains for users clustered in different subgroups. The intervention module provides personalized adaptive intervention for each subgroup, so that users in different subgroups can receive different interventions based on individual strengths and weaknesses. Intervention measures include general cognition and music skills in one or more domains or listening, speaking, reading, writing or other groups and interventions for specific languages.

[0011] On the other hand, the second subgroup generator iterates the user clustering analysis in the hierarchical clustering until only two clusters exist and then terminates.

[0012] In another aspect, the system includes an updating unit for dynamically updating the sub-groups when new users enter the database.

[0013] In another aspect, the intervention module includes visual motor games.

[0014] In another aspect, the visual-motor game instructs the user to perform one or more tasks related to pattern search, object matching, or sequence memory.

[0015] On the other hand, the intervention module consisted of having Yonghua recognize and distinguish subtle differences between speech and non-speech.

[0016] On the other hand, the difficulty levels of the training games in the intervention module gradually increased from level one to level ten.

[0017] On the other hand, the intervention module included vocabulary games.

[0018] On the other hand, the vocabulary games in the intervention module included both vocabulary breadth and depth.

[0019] On the other hand, the intervention module includes music assessment and training games, which assess musical emotions and preferences for different instruments while incorporating basic musical elements.

[0020] In another aspect, the user representation set includes data from dyslexic first language Chinese and second language English users.

[0021] On the other hand, the representation of users includes Chinese monolingual, English monolingual, and Chinese-English bilingual dyslexic bilingual users.

[0022] The present invention also provides a clinical assessment and intervention device and tool, which includes a system for performing strengths-weaknesses-based clinical assessment and intervention for dyslexia, as well as a user interface, a display and a processor.

[0023] On the other hand, the advantage of this integrated system for assessment, diagnosis and intervention treatment is also reflected in its flexibility and convenience in use. I-DID can be used simultaneously with one or more mobile devices such as mobile phones, tablets or laptops to perform clinical assessment and intervention treatment anytime and anywhere. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Embodiments of the present invention will be described in more detail below with reference to the accompanying drawings, in which:

[0025] Figure 1A A schematic diagram of a strengths-and-weaknesses-based assessment and intervention system according to an embodiment of the present invention is shown;

[0026] Figure 1B A schematic diagram of an integrated system for evaluation, diagnosis and intervention treatment of strengths and weaknesses of intelligent digital dyslexia is presented, which performs evaluation and intervention based on strengths and weaknesses according to one aspect of the present invention, thereby achieving Figure 1A system;

[0027] Figure 2 A schematic diagram of a method for performing strength-based assessment and intervention according to one aspect of the present invention is shown, which can be performed by Figure 1A system execution;

[0028] Figures 3A-3C , Figures 4A-4E , Figures 5A-5C , Figures 6A-6E , Figures 7A-7E , Figures 8A-8D , Fig. 9 , Figures 10A-10C , Fig.11 , Figures 12A-12B as well as Figures 13A-13B An exemplary scoring interface having questions for evaluation according to an aspect of the present invention is shown;

[0029] Figure 14-18 Presenting an example interface of I-DID evaluation according to one aspect of the present invention;

[0030] Fig.19 presents an example relationship between various fields according to an aspect of the present invention;

[0031] Fig. 20 A schematic diagram of a machine learning method for generating subgroups according to one aspect of the present invention is shown;

[0032] Fig.21 A schematic diagram of a theoretical driving method according to one aspect of the present invention is shown;

[0033] Fig. 22 A schematic diagram of an integrated theory and data-driven approach according to one aspect of the present invention is presented;

[0034] Fig.23 A flowchart of an automatic recommendation system for effective dyslexia intervention according to an embodiment of the present invention is shown.

[0035] Figure 24-28 An example interface of an I-DID game according to some embodiments of the present invention is presented. DETAILED DESCRIPTION

[0036] In the following description, devices, systems and / or methods for providing assessment diagnostic tools and interventions through assessment and intervention based on strengths and weaknesses are listed as preferred examples. Clinical and educational practitioners and technicians in this field may make appropriate adjustments to the specific implementation, including additions and / or substitutions, without departing from the scope and spirit of the invention. Specific details may be omitted here to avoid confusing the present invention; however, this disclosure is written to enable those skilled in the art to correctly apply the present assessment and intervention tools without systematic training.

[0037] Figure 1A A system 1000 is described for strengths-based clinical assessment and intervention for dyslexia. Figure 1A is a system overview with further details of various system components that are described in further detail below. System 1000 can be implemented in hardware, software, or a combination of hardware and software.

[0038] The system 1000 includes a database 1100 that can be dynamically updated according to user usage, which includes large-scale data or digital archives of multiple user performance sets across different fields. The first subgroup generator 1200 performs a theoretically driven machine learning method on the database 1100 by obtaining a layer with nodes, and the nodes are calculated by multiplying all component values ​​by the sum of the factor loadings of the component values ​​using exploratory factor analysis (described in detail below). Different subgroups can be generated depending on whether the node is activated. The second subgroup generator 1250 trains the neural network by applying principal component analysis (PCA) via an encoder and a decoder, thereby performing a machine learning method that integrates theory and data drive for the database 1100. The encoder encodes each user's data set into a three-dimensional vector and clusters the users into hierarchical clusters to classify and group the users based on the user similarity or distance calculated between the users. In one embodiment, the second subgroup generator can cluster the users in the hierarchical clusters more than once and terminate until only two clusters exist.

[0039] The comparator 1300 compares the theory-driven machine deep learning method with the theory- and data-driven machine deep learning method to determine whether the theory-driven machine deep learning method is consistent with the theory- and data-driven machine deep learning method. This system adopts a self-verification method to ensure accuracy. The scoring module 1400 determines the scores of each field relative to users gathered in different subgroups.

[0040] The intervention module 1500 provides personalized intervention for each subgroup so that users in different subgroups can receive interventions suitable for the different developmental characteristics of individual strengths and weaknesses. Interventions include general cognitive and musical skills training in one or more fields, or listening, speaking, reading, writing or a combination thereof in a specific language. Various interventions can be selected. Visual motion games are also included, which can guide users to perform one or more of pattern search, object matching or sequence memory. Alternatively, intervention can include detecting and distinguishing and identifying the difference between speech and non-speech. Training games for the understanding and use of vocabulary for vocabulary breadth and depth can also be provided. Music assessment games that evaluate musical emotions and instrument preferences while combining basic musical elements can be selected. As users successfully complete various intervention tasks, levels that gradually increase in difficulty can be provided.

[0041] To dynamically update the database, the update unit 1600 can update the subgroups when a new user enters the database 1100. The update unit 1600 may receive the user information via a wired or wireless connection to a network such as an intranet, or through the Internet.

[0042] Typically, the user performance set includes data from Chinese monolingual, English monolingual, and Chinese-English bilingual users with dyslexia, who are different from assessment systems that only assess a single language during the assessment, diagnosis, and intervention process. In one aspect, these users are more likely to learn two languages ​​at the same time, such as Chinese as the first language and English as the second language. Since Chinese characters are a pictographic writing system that uses unique square characters to represent semantics, while English is a phonetic writing system based on letters, there is a direct correspondence between English letters and sounds. Based on the differences between Chinese characters and English words, this poses a great challenge to the diagnosis of dyslexia in Chinese children who are native Chinese speakers and English as a second language. This requires considering the characteristics of Chinese children's native Chinese characters and systematically assessing and diagnosing their English language skills, so as to more comprehensively assess whether these children have dyslexia in one or both languages, and provide empirical evidence for targeted and precise intervention.

[0043] The clinical assessment and intervention device 2000 includes the strengths-based clinical assessment and intervention system 1000 as well as a user interface 2300 , a display 2100 , and a processor 2200 .

[0044] In another aspect, the system enables flexible and convenient clinical assessment and intervention in a mobile device selected from one or more of a mobile phone, tablet, or laptop.

[0045] Figure 1B A schematic diagram depicting an implementation of a device 2000 according to an aspect of the present invention that can be used for strengths-based assessment and intervention. The device 100 is an intelligent digital system 100 that includes a display 102, a program module 104 for implementing intervention 1500, and an assessment screening module 106 for implementing a dyslexia diagnostic test. The program module 104 is configured to electronically communicate with the display 102 and the assessment and screening module 106, and to send a diagnostic assessment to an update module 1600 to update a database 1100. The assessment device can interactively communicate and update with the system 1000 via a network such as the Internet.

[0046] Figure 2A schematic diagram of a method 200 for strength-weakness based assessment and intervention according to an aspect of the present invention is depicted, which can be implemented by the system 1000 in combination with the device 2000 and / or the device 100. When the method 200 is executed, single and multiple users can simultaneously obtain access to the digital electronic assessment system 1000. The electronic system 1000 in combination with the method 200 can be referred to as an intelligent / integrated dyslexic interface design (I-DID) system. The method 200 includes steps S202, S204, S206, S208, and S210.

[0047] In step S202, a scoring interface is provided to one or more participants and users 10 through display 102 or display 2100, and is used to communicate with the scoring module 1400. The scoring interface can use different evaluation formats associated with the task characteristics. A summary of the interface types can be found in Table 1. In one embodiment, the I-DID system is suitable for preschoolers, elementary school students, middle school students and adults. The 4-12 children's module covers five different assessment areas, covering Chinese and English language and literacy, cognition, creativity, music and social-emotional skills; while the middle school student and adult modules cover Chinese and English language and literacy and music assessments. Then, in step S204, the scoring module 1400 scores one or more answers of the participants 10 at each task stage. Table 1: Overview of test interface types for screening assessments for children aged 4-12 years, secondary school students, and adults

[0048] Figure 3A-3B A user initiated audio recording test format is shown. Figure 3A The cognitive assessment interface of the rapid number naming game is shown. There are two trials in the assessment, and the system automatically records the responses in the two trials. After the user completes the first trial, it will automatically go to the second trial. Figure 3B An example of a reading assessment format is shown, where stimulus material is presented on the screen a single time. While responses are recorded throughout the test, the user can click on the left and right arrows to proceed to the next stimulus or return to the previous stimulus. Figure 3C The test interface of the short passage reading comprehension task is shown.

[0049] Figures 4A-4E The audio autostart recording evaluation test format is shown. Figure 4AAnother reading assessment test format involving a graphic display is shown. The user is required to read each real or false word aloud and stop the automatically initiated recording after each response. Figure 4B An example of an assessment test format involving audio playback is shown. This format is applied to the digit span test and the speech test to assess the user's speech combination ability. After the stimulus is presented in each trial, the user's response is automatically recorded. Figure 4C-4D The evaluation test format applied to another type of speech test, which mainly consists of a two-part audio playback and recording, is shown. The evaluation test requires the user to remove a single sound from an auditory presented target word and give the ability of the remaining sound. Figure 4C The audio recording process is shown starting after the first audio is played. Figure 4D The audio recording process is shown starting after the second audio is played. Figure 4E An interface involving audio recording, visual images, and textual knowledge assessment is shown. In each trial, after playing the audio, the system automatically records the user's verbal response.

[0050] Figures 5A-5C A multiple choice assessment test format using images as key responses is shown. Figure 5A Shows an example of an assessment where items are displayed as images in the Word Knowledge and Orthographic Awareness assessments. Figure 5B-5C The image and audio matching assessment test format is shown in the voice test, which assesses the user's ability to select options (shown in the second row) that are phonetically similar to the target word (shown in the first row). When the corresponding audio is played, the image of each word will activate and light up to reduce reliance on the user's memory ability. Figure 5B A two-item forced-choice judgment assessment test format is shown. Figure 5C Three forced-choice judgment assessment test formats are shown.

[0051] Figures 6A-6E Test formats involving two, four, and six multiple choice assessments are shown. Fig. 6A A two-choice prosodic discrimination test format is shown that requires the user to select the best option based on audio stimuli. Figure 6B Four-choice and multiple-choice formats for assessing morphemes and vocabulary are shown. Figure 6C Showing the six-choice task format, this assessment game allows users to select the picture that best represents the audio stimulus from a six-choice list. Fig.6D The format of the Music Emotion Perception Game Assessment Test is shown. This test requires users to select the music that best represents the user's perception of four emotions based on the stimulus material presented by the music auditory presentation. Fig. 6E Shows the test format for a sentence comprehension assessment that requires the user to select the image that best represents the paragraph on each trial. A progress bar in this assessment test game allows the user to monitor their test completion.

[0052] Figures 7A-7E Shows a multiple choice test format using text displays as responses. Fig. 7A A format for one type of test that may be used for vocabulary and morpheme assessment for secondary school students and adults is shown, where the test format displays questions in text form and prompts the user to complete the question using one of four response options displayed below. Figure 7B-7C The test format for the vocabulary, reading comprehension and passage reading comprehension assessments is shown, which displays passages that the user needs to read and multiple-choice questions following each passage to assess the user's understanding of the passage. Figure 7B An example of a paragraph chapter display within an assessment task is shown. Figure 7C A question section following a passage is shown, requiring the user to select an answer from four options. Figure 7D-7E A test format is shown in which the user is required to judge the frequency stimulus and select "check" or "cross". Fig.7D Shows the format used in a prosody assessment test where two stimuli per item are played and the user is asked to select a "check" or "cross" based on whether they perceive the stimuli to sound the same or different. Fig. 7E Another binary assessment test format was shown, requiring users to judge whether two pieces of music audio were the same or different. Figures 8A-8D The test format presented as an audio stimulus requiring the user to respond to two or three options is shown. Fig. 8A Displays the test format used in spoken language comprehension and a type of adult speech assessment that plays a sentence and asks the user to select an answer from three options. Figure 8B A similar format for showing musical ability assessments. Figure 8C Displays the format used in the Morpheme Assessment Test for Adults, where stimuli in both text and audio are presented simultaneously and the user is asked to select a response from four multiple-choice options. Fig.8D Another image and audio display format for morpheme assessment is shown, in which an image is displayed while an audio stimulus is played, and the user is asked to respond from two forced options.

[0053] Fig. 9 Shown is the matching format for an assessment of an adult vocabulary test, where the user matches a target word with its description.

[0054] Figures 10A-10C Shows the format of the handwriting test in the Chinese dictation assessment, where the system plays the pronunciation of the target word and requires the user to write the heard target word in the box on the screen. Fig. 10A , 10B 10C and 10C respectively show the test interfaces for one, two, and four-word dictation assessments. Fig.11The handwriting test format of the English spelling assessment is shown, which requires users to write their English words on the screen after listening to an audio recording of the English words. Users can modify or rewrite their answers using the eraser and reset functions on the screen.

[0055] Figures 12A-12B The format of a test called the Verbal Sentence Span Working Memory Assessment is shown. First, Fig. 12A As shown in , the user is asked to confirm whether the sentence played in each trial is correct by pressing the corresponding key. The user is then asked to repeat the last word of each sentence. Fig. 12B As shown, the audio recording function is automatically activated to record the user's verbal response.

[0056] Figures 13A-13B An example showing the format of a drawing nonverbal creativity assessment test. Fig.13A A test interface including timer, eraser and reset functions is shown. The user can create a drawing in the frame provided on the screen. Fig. 13B The test interface when the time limit is reached is shown. It requires the user to verbally provide the name of their creation.

[0057] Figure 14-18 Examples of I-DID assessment test interfaces according to various embodiments of the present invention are shown. The entire interface makes comprehensive use of advanced graphic design, animation and multimedia display, which is used to create an intelligent integrated test system for users based on online real-time assessment tests and intervention tours that are educational, interesting and challenging.

[0058] After the evaluation test, the scoring module 2400 aggregates the total scores of the users 10 who have completed all items in each task stage in step S206, thereby assigning scores on various fields to each user 10. In this regard, Fig.19 An exemplary relationship between various domains according to one aspect of the present invention is shown. The I-DID system integrates and synchronizes and optimizes five basic elements of human development, including language and literacy, cognition, creativity, music, and social-emotional skills. Fig.19 It is a theoretical model of I-DID, namely the 2C-SEM-2L learning and training model of the brain, which includes five main areas: language and literacy, cognition, creativity, music and social-emotional abilities.

[0059] The existing diagnostic assessment and intervention training tools in clinical and educational practice are designed based on the theory of language and cognitive deficits of dyslexia, focusing mainly on the deficits of dyslexics in cognitive, speech and literacy development while ignoring the strengths and advantages of these individuals in the development of other abilities. Unlike these existing tools, the I-DID system is based on the 2C-SEM-2L model proposed based on the latest research in brain science. Combined with the optimization principles of brain learning, the system covers the five most important basic abilities of human development in the intelligent era, namely language and literacy, cognition, creativity, music and social-emotional skills. Moreover, unlike other traditional standardized assessment, diagnosis and intervention training tools for dyslexia, the design of I-DID not only evaluates five different abilities separately, but also emphasizes the integrity of personal development determined by the connection, interaction and coordination between the five abilities. Fig.19 As shown in Figure 1, I-DID emphasizes the connection and coordination mechanism between these five elements and uses them as the guiding principles for the design of I-DID assessment tests and intervention training content. The details of the assessment design for each domain are described below.

[0060] 1. Cognitive ability

[0061] The I-DID system assesses three key cognitive abilities: memory, executive function, and statistical learning. Based on the results of our research and other international and domestic research teams on dyslexia over the past decade, these three cognitive abilities play the most fundamental role in the language and literacy development of children in different cultural and linguistic backgrounds. Children with dyslexia have difficulties with working memory (WM), especially the ability to remember a series of information instantly and use this information to complete tasks. Existing behavioral research evidence shows that children with dyslexia have lower verbal and non-verbal working memory than children with normal development (Zou et al., 2022). And this conclusion is supported by the latest research evidence in cognitive neuroscience. Studies have found that for children with dyslexia, brain areas related to working memory are less activated (Beneventi et al., 2010). Based on these empirical research evidence, I-DID is a three-type online working memory game designed to assess visual spatial, phonological, and verbal working memory by using visual objects, numbers, and words, respectively.

[0062] Executive function is another important cognitive ability in I-DID assessment, diagnosis and intervention training. Executive function (EF) is the three core cognitive processes of inhibition, updating and shifting related to goal-oriented behavior. Current research shows that children with dyslexia perform poorly in inhibition and updating ability, but their switching ability is better than that of children with normal development, which is a potential strength and advantage (Doyle et al., 2018). However, the study also found that when the speed of completing the problem needs to be considered, these children are reduced in accuracy. In the I-DID system, based on the existing executive function research paradigm, we design multiple executive function online games related to the real world and life to evaluate the strengths and weaknesses of individuals in different executive function abilities.

[0063] Implicit statistical learning (SL) is the third basic cognitive ability to be assessed in I-DID. Implicit statistical learning refers to the cognitive process of discovering and extracting potential rules such as transition probability in a continuous stream of stimuli. According to our recent meta-analysis of dyslexics using different research paradigms and different languages ​​using statistical learning, children with dyslexia have moderate difficulties in statistical learning compared with children with normal development (Lee, Tong, & Cui, 2022). Moreover, our series of experimental studies on statistical learning found that although there was no difference in the overall performance of school-age children with Chinese dyslexia and children with normal development in statistical learning of radical position rules, their performance in left-right structure characters was lower than that in upper-lower structure characters, and they had difficulties in matching phonology-grapheme and semantics-grapheme, which indicates that children with Chinese dyslexia have difficulties in handling and processing inconsistent Chinese character input, especially these children have difficulty inhibiting the influence of interference items (e.g., Tong et al., 2020). At the same time, we also found that after giving a large amount of stimulus input, children with Chinese dyslexia can also perceive the statistical rules contained in the continuous series of materials (He & Tong, 2017). From the perspective of overall development based on individual strengths and weaknesses, these findings suggest that when exposure increases, children with dyslexia demonstrate specific types of statistical learning performance comparable to their typically developing peers. Therefore, based on the research evidence on statistical learning of children with dyslexia in China, the I-DID system contains two statistical learning games for assessing children's statistical learning of verbal and non-verbal regularities in different contexts.

[0064] 2. Language and Literacy

[0065] The I-DID system is different from the existing language and literacy measures in dyslexia assessments or screenings, which only focus on single-task measures of metalinguistic awareness, Chinese word recognition, and reading and writing. However, I-DID uses a multi-level and multi-component approach to design multiple different games to systematically assess oral listening and speaking, which are closely related to word recognition and dictation, from a micro and macro perspective.

[0066] Specifically, it is based on our research team's past empirical research evidence on children's vocabulary and reading development (e.g., Tong & Tong, 2022). Oral comprehension ability includes two closely related and different key components, namely vocabulary ability and narrative ability. The former focuses on children's understanding of word meaning, while the latter assesses children's ability to independently organize expressive language by generating stories. The vocabulary game design independently developed in our I-DID system includes comprehension vocabulary tasks and expression production vocabulary tasks. The two types of response tasks assess and test both the breadth and depth of vocabulary. The breadth is mainly based on the amount of vocabulary that can be understood and produced, and the depth mainly tests the extent to which children understand the vocabulary. Oral narrative ability is mainly assessed by assessing children's ability to independently organize language to construct and tell a coherent story based on the presented pictures.

[0067] In addition to the above-mentioned key vocabulary and oral language ability assessment and testing, the language games developed in-house by the present invention also include metalinguistic skills, such as phonological and morpheme awareness abilities. The design of the phonological awareness assessment test invented by I-DID differs from existing standardized tests and tools for dyslexia in three key ways. First, the existing metalinguistic awareness assessment format is similar to the tests and exams used in schools to assess children's language skills, which measure more memory ability than the ability to use language in real life. However, the phonological and morpheme awareness assessment games of the I-DID system use children's perception and understanding of the potential language ability of different levels of language structure to design a series of progressive mini-games that allow children to gradually understand the segmentation or integration of phonology and morphemes. Secondly, the mode of conducting the assessment test is different. Unlike the existing screening and assessment standardized tests that use traditional paper and pencil formats, the metalinguistic ability test assessment of phonological and morpheme awareness invented by I-DID is embedded in an online game environment. This not only stimulates children's interest in language learning; it can also reduce children's anxiety and stress caused by assessments that are mainly in the form of exams. Third, the content of I-DID's language and literacy test games is based on my research results and experience on children's speech, language and reading abilities for more than ten years. Through systematic analysis of the distribution information of Chinese and English databases and the different age development stages of children's acquisition of different language abilities, a series of key factors in the design of language stimulus materials such as word frequency, complexity, concreteness and abstractness are systematically considered. In addition, the design of our assessment and intervention content in Chinese and English also takes into account the asymmetric and nonlinear development characteristics of Chinese-English bilingual children in learning Chinese and English at the same time.

[0068] 3. Creativity

[0069] Creativity is characterized by a multi-component structure, which is a process consisting of generating new ideas through divergent thinking and exploring and evaluating the feasibility of these ideas through convergent thinking (Barbot et al., 2015). At present, growing research evidence shows that creativity is a major advantage shown by individuals with dyslexia, but this advantage depends more on the form of the assessment task. For example, a recent study of 124 dyslexic children in grades 2-5 in Hong Kong and 122 age-matched typically developing (TD) children showed that Chinese children with dyslexia, regardless of high or low grades, were significantly better than their typically developing non-dyslexic peers in non-verbal creativity and performed comparable to typically developing children in graphic creativity; moreover, we found no significant correlation between creativity and children's literacy (Lam & Tong, 2021). Other studies have also pointed out that high school students with dyslexia show strong creativity compared with their typically developing peers, for example, they have the ability to connect unusual ideas and recognize novel relationships between concepts in the verbal domain (Cancer et al., 2016).

[0070] In view of the fact that children with dyslexia may have different creativity characteristics, the I-DID system comprehensively evaluates creativity through games developed internally. Unlike traditional paper-and-pencil creativity measurements, the I-DID creativity game is conducted through an online interface, which can automatically record the entire process of children's creation in real time. Through the evaluation tests of non-verbal and literacy skills such as creativity of the present invention, the I-DID system is not only intended to change the currently used standardized tests that only emphasize the "defects" of children with dyslexia, but also hopes that teachers, parents, clinicians and educators can change their one-sided understanding of children and adults with dyslexia by using the I-DID intelligent integrated evaluation test and training intervention platform, and can better combine the strengths of dyslexics such as creativity into the design of intervention training content, so as to stimulate the internal motivation of dyslexics to learn and change their own literacy level. Moreover, the I-DID system develops effective advantage-based learning strategies to help children with dyslexia better realize their potential and personal value in school and society.

[0071] 4. Music

[0072] The design of the music assessment game in the I-DID system is based on existing research on the musical development of children with dyslexia and typically developing children. According to existing research evidence, children with dyslexia differ significantly from typically developing children in all aspects of musical ability. In terms of rhythm and prosody, children with dyslexia show difficulties in musical stress (Caccia & Lorusso, 2020), melody discrimination (Forgeard et al., 2008), and temporal perception of musical rhythm transformation (Overy et al., 2003). Moreover, evidence from longitudinal follow-up studies of children with dyslexia also shows that rhythm perception and production ability are closely related to reading ability, especially phonological awareness, which plays an important role in literacy development (Reifinger, 2019).

[0073] Existing assessments of music perception ability, including Primary Measures of Music Audiation (PMMA), Intermediate Measures of Music Audiation (IMMA; Gordon, 1979; 1986), and the Montreal Battery of Evaluation of Musical Abilities (MBEMA; Peretz et al., 2013), only assess basic music perception. Unlike these existing assessments and tests of music ability, the I-DID Music Assessment Game is the first to measure other key factors, especially musical emotion perception and preference for different instruments, while also incorporating the perception and understanding of basic musical elements including pitch, rhythm, and tempo. The I-DID's assessment of musical emotion perception is based on current research evidence that music is closely related to emotional expression. Moreover, an individual's ability to perceive emotional information in music, understand different performance characteristics, and recognize various musical timbres is important for a comprehensive interpretation of music. Therefore, these unique features make the I-DID Assessment Game more novel and comprehensive than previous assessment tools. In addition, the I-DID system takes into account the special learning needs of children with dyslexia when designing music assessment and intervention games. Specifically, all music assessment materials are gamified through visual and auditory displays, so no additional reading and language support is required. In addition, I-DID's user-friendly online format makes the administration of assessment tests and intervention treatments simple, requiring only the child to press keys on a computer keyboard. Therefore, I-DID games represent the first, most comprehensive, and unique online music assessment and intervention treatment system designed to be user-friendly for dyslexics.

[0074] 5. Social-Emotional Skills

[0075] Unlike most existing diagnostic assessment tools that completely ignore the emotional development and resilience of dyslexic patients, the I-DID system comprehensively and systematically assesses the social-emotional skills of dyslexic children and their parents / caregivers. Specifically, the I-DID system's online game is equipped with 14 parent and child social-emotional development questionnaires and tests, which assess children's social-emotional skills and resilience from micro and macro levels by considering individual personality traits, social relationships, and educational background. More importantly, this is the first online assessment system for Chinese children's social-emotional skills and parents' happiness.

[0076] The design of the assessment game, questionnaire and test of the present invention targets social-emotional skills, and its theoretical basis is Vygotsky's (1978) social learning theory, which emphasizes the connection between emotions and language and regards language as an active and advanced tool to promote emotion regulation. This theory has been verified by a large number of empirical studies. A large number of empirical studies have shown that an individual's social-emotional skills, that is, the ability to perceive, express and regulate emotions, can play a key role in language and literacy acquisition. However, despite the connection between social emotions and language and literacy skills, current diagnostic and screening tools for dyslexia seriously ignore the development of social emotions in children with dyslexia. In response to this problem, the I-DID assessment program systematically integrates two perspectives on the social-emotional development of dyslexia to better understand the strengths and weaknesses of children with dyslexia. On the one hand, studies have found that children with dyslexia will find themselves different in the development of academic abilities in school learning and are affected by the stigma of dyslexia. This in turn leads to low self-esteem and poor emotional health in these dyslexic children in terms of academic abilities (Livingston et al., 2018; Terras et al., 2009). On the other hand, research based on resilience points out that some children with dyslexia, especially those with strong coping ability to adversity, show less negative emotions than those with low resilience (Catts & Petscher, 2022). Therefore, social-emotional health can be understood as a potential protective factor to cope with the academic challenges faced by children with dyslexia (Terras et al., 2009). The protective role of social-emotional health, especially resilience, has been demonstrated in relevant intervention programs. Specifically, after systematic resilience training, middle school students with and without dyslexia have comparable social-emotional development abilities and the ability to cope with adversity (Firth et al., 2013). By integrating these two aspects of research evidence, the I-DID system not only assesses all negative and positive social-emotional outcomes that may result from dyslexia, but also takes into account the potential negative social-emotional consequences caused by dyslexia and the interactive relationship with the development of resilience.

[0077] Return to reference Figure 1A , Figure 1B and Figure 2 In one embodiment, after aggregating the total scores of the users 10, the total scores of the users 10 may be displayed via the display 102 or the display 2100. A radar chart may be displayed via the display 102 or the display 2100 to show the performance of the users in the five domains. Thereafter, in step S208, the screening model 106 screens the users 10 according to the scores of the users 10 relative to the five domains, so that the users 10 are assigned to one or more subgroups. The subgroups are related to three key functional domains: (1) cognition and creativity (C-domain); (2) language and music (L-domain); and (3) social and emotional skills (S-domain). The C-domain refers to a set of high-order executive processes used for abstract reasoning, decision-making, and problem solving. The L-domain refers to a set of domain-specific knowledge used to decode and encode linguistic and non-verbal information with rule-controlled prosodic structures. The S-domain aims to achieve a set of social-emotional skills for external interaction and internal regulation by integrating ideas, knowledge, and feelings.

[0078] In each CLS domain, users are divided into a high (+) group and a low (-) group. Therefore, in one embodiment, corresponding to the C domain, L domain, and S domain, eight subgroups are generated: (1) C+L+S+; (2) C+L+S-; (3) C+L-S+; (4) C+LS-; (5) C-L+S+; (6) C-L+S-; (7) CL-S+; and (8) CLS-. For example, the subgroup "C+L+S+" means that the C, L-, and S- domains are all at a high level; the subgroup "C+L+S-" means that the C- and L- domains are at a high level, while the S- domain is at a low level.

[0079] In one embodiment, the screening model 106 is configured to generate and save a plurality of subgroups. The generation of subgroups will be described as follows. Fig. 20 3 is a schematic diagram of a machine learning method 300 for generating subgroups. The machine learning method 300 includes steps S302, S304, S306, S308 and S310.

[0080] The database at step S302 includes a large-scale archive of multiple user performance sets across different domains. The database can be used as training data and fed into the screening module 106 for training purposes. The database can be used as a starting point for generating subgroups based on confirmed user performance patterns. The data stored in the database can provide information for developing a powerful and accurate screening module, so that the module can be used to support personalized intervention treatment work.

[0081] Step S304 shows a method for executing theory driving. Fig.21 The theoretically driven method according to one embodiment of the present invention is shown. The theoretically driven method is performed on the database, and its purpose is to Fig.21Get 3f in the diagram s Layer. In the basic form function 3f s In the layer, the C, L, and S domains are interrelated and interdependent, where: node f1 represents "C", which is the basic top-down process of the L domain; node f2 represents "L", which represents the formal structure of the C domain and the S domain; f3 represents "S", which expresses the functional skills of the L-domain.

[0082] The input layer consists of a set of normalized eigenvalues ​​x of size n. To reduce the data dimensionality based on theory-driven techniques, the first step is to perform exploratory factor analysis (EFA) to calculate the covariance between the observed variables explained by a set of latent factors F of size m. i and loaded into F j On, and the factor loading is λ ij , plus the error term δ i :x i =λ ij *F j +δ i Therefore, n inputs and m factors can be represented by matrix 402.

[0083] Based on the EFA results, F j The component value c j is calculated by summing all inputs multiplied by the corresponding factor loadings represented by matrix 404. Subsequently, another set of EFAs is performed to obtain factor loadings corresponding to the component values ​​of the three second-order latent factors of f1, f2, and f3. Similarly, the f1, f2, and f3 values ​​are calculated by summing all component values ​​multiplied by their factor loadings on the second-order latent factors, which are represented by matrix 406.

[0084] When 3f s Each node in the layer (i.e., f1, f2, and f3) is activated when its value is greater than a cutoff value, such as the 25th percentile of the proportion score in some embodiments. The 2×2×2 activation pattern of these three nodes can produce or generate eight subgroups in the output layer as described above.

[0085] In step S306, an integrated theoretical and data-driven approach is performed, such as Fig. 22 As shown, Fig. 22 An example diagram is shown according to one embodiment of the present invention, which is an integrated theory and data driven approach. The integrated theory and data driven approach is performed by applying principal component analysis (PCA) using a neural network 502 and grouping users in hierarchical clustering.

[0086] Specifically, the first stage is to reduce the dimensionality of the large-scale profile of the database by applying PCA using an unsupervised autoencoder neural network 502, which includes encoding and decoding processes. The user's large-scale profile, including performance in multiple cognitive, literacy, music, creativity and social emotional assessments, can be converted into a set of smaller principal components in the compression layer by multiplying the input matrix with the principal component matrix. The principal components can be decompressed by dividing by the principal component matrix.

[0087] During encoding, the input layer is mapped to a Φ The hidden layer F of . The input layer includes a set of normalized feature values ​​x, whose set size is n.

[0088] For any x∈X,f∈F,f=E Φ (x) = σ(Wx + b), where: σ is the element-wise activation function: W is the weight matrix; and b is the bias matrix.

[0089] In the decoding process, it can be based on the function D θ The encoder is used to reconstruct the output layer X'.

[0090] For any f∈F,x′∈X′,x′=Dθ(f)=σ′(W′f+b′), where: σ′ is the element activation function: W′ is the weight matrix; and b′ is the bias matrix.

[0091] Next, the difference d(x, x′) existing between X and X′ is measured, and the loss function L(Φ, θ) is obtained.

[0092] For x∈X,x′∈X′,

[0093] and

[0094] Further training is to minimize L by obtaining the optimal Φ and θ. After training, for each user i, the encoder encodes the data set of each user to obtain a three-dimensional vector, f i =(f i1 ,f i2 ,f i3 ), which in theory would include C, L and S scores.

[0095] The second stage is to generate the original distance matrix D by grouping users in hierarchical clustering, which is Fig. 22Marked as 504.

[0096] Based on the results of the autoencoder, an N×N distance matrix D504 is obtained, which represents the Euclidean distance between each of the two users, including the C, L, and S domains. For each two users (e.g., i and j), the sum of the Euclidean distances in terms of C, L, and S is considered to be:

[0097] The matrix D 504 may be optimized 506 as a process to obtain a distance matrix D*.

[0098] This stage involves grouping users in hierarchical clustering based on the similarity or distance of users calculated between all users. Specifically, the first step is to search for the minimum value (e.g., d ij ), and cluster the two users (e.g., i and j) into one group. The distance matrix D* is then updated by recalculating the distances between the users / merged groups. When only two clusters exist, the search and merge process is repeated, and clustering is terminated, as shown at the end of process 506.

[0099] For the distance d r(ij) Between user r and the merged group (ij), there exists:

[0100] For the distance d between two merged groups (e.g., group R containing n1 users and group I containing n2 users), RI , then there exists: For r∈R,i∈I,

[0101] Return to reference Fig. 20 ,In step 308, the theory-driven approach is compared with the integrated theory and data-driven approach to check whether the results of the former are consistent with the results of the latter, which can serve as a result verification mechanism, for example, to see whether the subgroups classified using the two methods are the same. Then, the scores of users clustered in different subgroups in the language and literacy, cognition, creativity, music, and socio-emotional domains are calculated.

[0102] In step 310, as new users participate in the I-DID system, they are entered into the database, at which point the screening module 106 can automatically and continuously update the subgroups, which can serve as validation of the clustering. The update mechanism not only ensures that the subgroups remain relevant and accurate over time, but also allows for the identification of new patterns and trends in user performance. Finally, as a result of machine learning, the trained I-DID system can automatically provide each user with its customized intervention.

[0103] Return to reference Figure 2 In step 210, personalized intervention is provided for each subgroup, so that users 10 in different subgroups can receive different intervention programs through display 102 or 2100, wherein the intervention program is implemented by music and simulated narrative training for reading, writing, listening, speaking or a combination thereof. In one embodiment, the I-DID system can automatically recommend intervention plans and recommend users with dyslexia to receive appropriate intervention programs, such as I-DID game intervention and music intervention. For example, users 10 in different subgroups can receive training in the same domain but at different levels.

[0104] The I-DID intervention game, called I-DID Carnival, is a self-developed mobile application intervention program. This intervention uses interactive game design techniques and meets individual learning needs by engaging dyslexic children of different ages in a series of mini-games in a mobile application. These games are designed to improve children's cognitive, language, literacy, music, and socio-emotional skills, which are related to the development of reading ability. Although there are mobile applications for children and adults with dyslexia on the market, most of them focus only on reading and writing skills. The invented I-DID Carnival is the first to target a variety of cognitive skills such as working memory, visual motor and auditory skills, and music skills. This bottom-up design enables children with dyslexia to explore their untapped potential.

[0105] The entire intervention on the mobile phone computer device can be divided into four parts, each targeting a different skill set: (1) general visual; (2) general auditory; (3) language-based; and (4) Fig.23 The language complex shown in FIG. 1 shows exemplary automatic recommendations for effective dyslexia intervention according to an embodiment of the present invention. The general layer includes the basic components of language, both oral and written. Written language is highly related to visual perception in learning and the motor skills required to write letters and characters. Spoken language is built using basic auditory sensory skills, such as pitch and rhythm, which are further combined with language-specific units (such as phonemes) to form vocabulary and sentences. In addition, visual and auditory working memory are also trained as basic cognitive skills.

[0106] Both the perception and production of language involve basic sensory skills. Auditory skills are very important for speech perception and oral expression, while visual motor skills are very important for Chinese character dictation and English spelling. The grapheme-sound connection represents the correspondence between spoken and written language. Morphemes are the representation of semantics in the language system. These components are organically integrated into the intervention program of the present invention.

[0107] The output layer targets the perception and production of oral and written language, especially listening, speaking, reading and writing skills, which are the most basic language and literacy skills in the development of children's academic and learning abilities.

[0108] Figure 24-27 I-DID intervention games according to various embodiments of the present invention are shown. All mini-game designs in the mobile application contain an engaging theme interface that children are curious and interested in. Visual motion mini-games require users to search for patterns (such as the flight direction pattern of fireflies) or perform object matching, while working memory games require users to remember sequences. In addition, the mini-game design of the present invention takes into account the different development levels of participating users and sets ten different difficulty levels by using the principle of incremental design. In order to immerse children in the training of cognitive, language and literacy, creativity, music, and social-emotional skills, each game includes 10 different difficulty levels, which are designed and operated according to the target training skills and game formats. For example, for the game of visual search for the flight direction of fireflies, users are required to detect the main flight direction of most fireflies, and levels 1 to 5 only manipulate the left or right two directions, while levels 6 to 10 manipulate the left, right, up and down four different directions. In addition, the time given by the user to respond is gradually reduced, starting from 3000ms at level 1 and ending at 500ms at level 10. Similarly, for the auditory musical pitch discrimination game, the difficulty level was manipulated by increasing the difference between the two sounds, in units of cents (1 / 100 of a semitone), which represents the ratio difference between the two notes calculated using the formula 1200×log2(f2 / f1).

[0109] Therefore, the step-by-step, incremental framework of the present invention realizes a comprehensive and targeted approach to improve children's sensory and cognitive skills related to reading (e.g., visual motor skills and working memory) and language skills. In summary, using a systematic bottom-up design, the I-DID Carnival of the present invention not only helps children with dyslexia, improves and treats specific difficulties related to language and reading, but also enhances and improves related skills in cognition, literacy, music, creativity, and social emotional assessment in children with normal development.

[0110] Another I-DID intervention treatment program is a music-based language intervention for children with and without dyslexia. This intervention is designed to enhance children's pitch perception in both verbal and non-verbal contexts. Specifically, a series of structured step-by-step games that combine gamification, multi-sensory stimulation input, and a multi-channel approach provide a foundation for children to learn the concepts of pitch, vocabulary tone, and speech rhythm. This intervention is supported by empirical research evidence. Studies have shown that music training can effectively improve a range of reading-related abilities in elementary school students with dyslexia, such as phonological awareness and spelling (Willats, 1994, Reifinger, 2019). In addition, a large number of previous studies have shown that musicians will show higher abilities than non-musicians in processing pitch and tone information associated with speech signals (e.g., Besson et al., 2011, Cooper & Wang, 2012).

[0111] In addition, a study by my team showed that Chinese dyslexics showed significant difficulties in Chinese tone perception compared with children of the same age with typical development (Tong, Tong, & Yiu, 2017). The music-based Chinese tone intervention therapy of the present invention is the first to date to transform empirical research on tone perception into clinical practice to improve Chinese tone perception by training with musical pitch. The training content of both musical pitch and Chinese tones follows a step-by-step, progressive approach from easy to difficult. I-DID's music-based Chinese tone perception training and learning are not only widely applicable to children and adults, but also to any potential learners who exceed the expected age-related ability and learners of Chinese as a second language. Each game training has a specific goal, and the learning goal is achieved through guided multimedia learning activities. In addition, various pre-intervention and post-intervention assessment games are embedded in each training course to accurately evaluate the results of each user's completion of each stage of game training. This function allows users to adjust the difficulty level according to their personal learning progress, thereby achieving more personalized intervention therapy. In short, I-DI's music language intervention is to enhance the potential musical power of children with dyslexia through gamification and animated virtual reality, stimulate their interest in language learning, and thus alleviate their learning difficulties in reading and writing.

[0112] In summary, as a clinical assessment and diagnosis and educational teaching auxiliary tool, the newly developed I-DID intelligent and integrated dyslexia clinical diagnosis and intervention interface has the following characteristics.

[0113] First, the I-DID system uses various gamification and multimedia technologies to digitize and intelligentize all assessment, diagnosis, and intervention treatment tests, which can promote large-scale early screening and assessment of children who are at potential risk of dyslexia who are monolingual in Chinese, monolingual in English, and bilingual in Chinese and English, to ensure early detection and intervention of children with dyslexia.

[0114] The I-DID system combines artificial intelligence, cognitive neuroscience brain learning mechanisms, and rigorous diagnostic assessment and treatment design, representing an intelligent and integrated clinical assessment and educational training product that integrates interdisciplinary research evidence. Specifically, by utilizing graphic design, animation technology, game mechanics, and multimedia displays in advanced computational science artificial intelligence, the I-DID system creates an interesting and challenging future-oriented virtual testing and learning environment. All assessment games incorporate multi-sensory and multi-channel input and output modes of images, audio, video, animation, and text. The I-DID Carnival Intervention Treatment System is a multi-game online learning platform, and each challenge training game includes 10 levels with instant feedback and reward systems. This digital, intelligent, and gamified design not only changes traditional exam- or test-based assessment and intervention by maximizing learners' engagement and motivation, but also improves the efficiency and popularity of early screening and treatment for all children at risk of dyslexia.

[0115] Secondly, for the self-developed content and design innovation, the I-DID system adopts a new concept of assessment and intervention combining strengths and weaknesses, an interface for multilingual users of Chinese, English, and Chinese and English, and children and parents can be assessed and trained simultaneously. This design concept is completely different from the current single emphasis on standardized assessment and training of children with dyslexia. This design can better take into account the needs of users of different languages, and store positive and family-oriented assessment and treatment of dyslexia.

[0116] The I-DID system is the first digital and intelligent system to date that assesses the strengths of children with dyslexia in creativity, music, cognition and social-emotional skills, and the weaknesses in language and literacy through a computer interface. Moreover, I-DID simultaneously assesses the first language, Chinese, and the second language, English. At the same time, I-DID conducts synchronous assessments of children and parents. More importantly, the I-DID system includes 92 self-developed and internally programmed educational games. The assessment includes five major areas, namely language and literacy, cognition, creativity, music, and social-emotional skills. This system is not only suitable for children with dyslexia, but can also be used to assess the overall development profile of the different abilities of generally developing children.

[0117] Third, regarding user experience innovation, the I-DID system meets the different learning needs of neurodiversity and general learners through progressive web applications, and is a continuous developmental assessment and training software.

[0118] The I-DID system utilizes user-oriented design to meet the diverse and personalized needs of learners of different ages through a selection of website and application modes. Since young children and primary school students may not be mobile phone users, the I-DID website supports an assessment system on a touch screen, which can be easily accessed by young children and primary school students. At the same time, the mobile application is convenient for middle school students and adults, and even novice mobile phone users can easily operate independently according to the guidance. In addition, all games follow the concept of sustainable assessment, in which the designed activities not only have the purpose of immediate diagnosis, but all game programs are designed with users as the theme, and learners are able to self-monitor and conduct assessments at any time, and track and record their continuous progress and improvement in intervention training.

[0119] The specific aspects of the innovation and novelty of the I-DID invention will be further elaborated below.

[0120] 1. Intelligent online diagnosis and intervention system:

[0121] The I-DID intelligent computer system is designed to provide timely diagnosis and early intervention for children at risk of developmental dyslexia. The user-friendly interface and low cost of the I-DID system enable the current system to be ahead of the traditional and more expensive face-to-face standardized model and process, and is more suitable for large-scale early screening and evaluation of children at potential risk of dyslexia who are monolingual in Chinese, monolingual in English, and bilingual in Chinese and English. Moreover, this system can be further applied to other monolingual and bilingual populations, such as the Chinese communities in Singapore, Malaysia, and the Philippines.

[0122] 2. Scope of application:

[0123] The I-DID system can be widely applied at three levels, including individuals, groups, and school classrooms. First, the above-mentioned clinical practice aspects are direct applications because the I-DID is tailored for Chinese children and adults and children learning English as a second / foreign language, and can provide a comprehensive profile of all children's abilities, namely language and literacy, cognition, creativity, music, and social-emotional skills. Second, the I-DID system can be implemented by teachers in local mainstream primary schools for effective assessment at the classroom level. In addition, the I-DID can be applied to other non-local Chinese mainstream schools in other societies. Moreover, this system can be appropriately adjusted and modified to provide continuous assessment of the ability development of dyslexic and typically developing children in other countries with different languages. Third, parents who raise children whose first language is Chinese can also use this system to more comprehensively and accurately assess and detect the strengths and weaknesses of their children's development, no matter where they are in the world, so as to formulate personalized education plans suitable for their individual development.

[0124] 3. I-DID is designed based on strengths and weaknesses:

[0125] The integrated design concept of the I-DID system strengths-weaknesses assessment diagnosis and intervention treatment guides parents to consciously identify their children's strengths and weaknesses and prioritize targeted family literacy training practices at home. More importantly, the personalized, counting and gamification functions of the I-DID system can be easily operated and executed by parents or caregivers. Finally, the strengths-weaknesses assessment diagnosis and intervention treatment products of I-DID are modified for use in other Chinese and non-Chinese societies, which helps to reduce the pressure on clinical and educational professionals to conduct repeated assessments and repetitive intervention treatments for a large number of children, allowing them to focus more on self-improvement and humanistic care and care for the physical and mental development of children with dyslexia.

[0126] 4. Improve well-being:

[0127] The goal of the I-DID system is to improve the reading and writing difficulties of children at risk of dyslexia by identifying their underutilized talents and undeveloped potential, thereby empowering these children and individuals to better realize their own value. Ultimately, the I-DID system is to support all children to learn, grow and develop the ability to meet future challenges.

[0128] In addition to its unique features, the implementation of the I-DID system will also address the following long-standing needs in the current market.

[0129] First, the I-DID system provides a comprehensive set of assessments and interventions that are more suitable for diagnosing Chinese monolingual, English monolingual, and Chinese-English bilingual dyslexia, specifically designed for Chinese native speakers. Currently, there is no digital, intelligent, and systematic online assessment tool that can provide diagnosis of developmental dyslexia in Chinese and English. In addition, through online implementation based on the web and mobile applications, the I-DID system can automatically provide individual and group users with an overview of strengths and weaknesses in the areas of language and literacy, cognition, creativity, music, and social-emotional.

[0130] In addition, because the I-DID system can promptly identify dyslexia symptoms and provide personalized intervention treatment for children at risk of dyslexia, it can also address the frequent delays in treatment after long waiting times for clinical assessment and follow-up services, which is common in different countries and regions. The online function of the I-DID system enables all families to access assessment and intervention services anytime and anywhere, which is especially necessary for families who have difficulty accessing expensive face-to-face diagnosis and treatment services.

[0131] More importantly, the digital intelligent clinical assessment and intervention system is designed for individuals of different age groups, from preschoolers to adults. Considering that most formal diagnoses of dyslexia are usually made in children aged 7 to 8 (primary school), the I-DID system can accurately identify behavioral symptoms and development trends before that age, such as preschoolers, which helps to make early diagnosis more accurate.

[0132] A brief description of the I-DID invention relative to existing dyslexia assessment methods is as follows.

[0133] (1) All existing reading assessment diagnosis and intervention treatment are mainly based on defects. Different from these diagnostic tools, the assessment diagnosis and intervention treatment system of the I-DID system of the present invention is the first to abandon the emphasis on the defects and deficiencies of dyslexics, and comprehensively evaluate the strengths and weaknesses of children with potential dyslexia and children with normal development in five areas (language and literacy, cognition, creativity, music and social emotion) in multiple dimensions and levels, so as to destigmatize dyslexics and make them and everyone more aware of the overall balance of strengths and weaknesses in individual development. More importantly, this I-DID effectively utilizes the potential advantages of dyslexic children and designs a series of personalized intervention treatment games. At present, the existing tools for clinical and educational practice assessment and diagnosis are limited to the assessment of cognitive and language abilities, emphasizing the defects and weaknesses of dyslexic children. This not only stigmatizes dyslexic children at home, but also largely causes children, parents and teachers to emphasize the deficiencies of dyslexic children, while ignoring their underutilized talents and undeveloped potential, which affects the self-confidence and physical and mental health of these children in their academic development.

[0134] (2) Most existing dyslexia assessments and interventions use individual face-to-face consultations. Unlike these existing tools, the online I-DID system of the present invention effectively addresses the limitations of time and physical location, making the early diagnosis and intervention process simpler, more timely and more accurate. More importantly, the I-DID system of the present invention is an intelligent system that can identify and integrate assessment information from all five domain levels and generate a comprehensive strength and weakness profile for children and adults at risk, so as to recommend children at risk of dyslexia to an I-DID treatment model that is suitable for their individual characteristics of ability development.

[0135] (3) Most existing assessments and interventions are "one-size-fits-all" standardized assessments and interventions. Unlike these tools, the I-DID system of the present invention takes into account the heterogeneity of individuals with dyslexia. The system assesses the strengths and weaknesses of each user and provides a personalized intervention plan to meet the different learning needs of children with dyslexia and children with typical development.

[0136] (4) Most existing diagnostic and assessment tools only focus on high-risk children with a family history of dyslexia and the assessment and intervention of their first language. Unlike these tools, the I-DID system of the present invention assesses, diagnoses and intervenes in the first language of children and their parents, Chinese, and the second language, English.

[0137] (5) Most existing assessment and intervention programs focus on limited cognitive and literacy skills. Unlike these existing tools, the I-DID system of the present invention systematically incorporates language and literacy, cognition, creativity, music, and social-emotional skills into its diagnosis and intervention system. This can enhance the capabilities of high-risk children with a family history of dyslexia through positive assessment diagnostic experience and training results, thereby alleviating their potential emotional, behavioral, and reading and writing problems.

[0138] The functional units and modules of the embodiments described above, including the above-mentioned electronic clinical diagnosis and intervention systems and programs, can be implemented using computer processors or electronic circuits, including but not limited to application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), microcontrollers, and other programmable logic electronic circuits configured or programmed according to the present disclosure. The machine instructions or software codes running in the electronic clinical diagnosis tool can be easily prepared by a technician in the software or electronics field based on the present disclosure.

[0139] All or part of the methods and systems according to the embodiments may be executed in one or more computing devices, including server computers, personal computers, laptop computers, and mobile computing devices, such as smartphones and tablet computers.

[0140] Embodiments may include computer storage media, transient and non-transient memory having computer instructions or software code stored therein, which may be used to program or configure the electronic teaching device to perform any of the processes of the present invention. The storage media and transient and non-transient memory may include, but are not limited to, floppy disks, optical disks, Blu-ray disks, DVDs, CD-ROMs and magneto-optical disks, ROMs, RAMs, flash memory, or any type of medium suitable for storing instructions, codes, and / or data.

[0141] Each functional unit and module according to various embodiments may also be implemented in a distributed computing environment and / or a cloud computing environment, where all or part of a machine instruction is executed in a distributed manner by one or more computing devices interconnected by a communication network, such as an intranet, a wide area network (WAN), a local area network (LAN), the Internet, and other forms of data transmission media.

[0142] The above description of the present invention is provided for the purpose of illustration and description. It is not intended to list or limit the present invention to the disclosed description. Clinical and educational practice technicians can appropriately modify the configuration of this invention to adapt to the operating system of local servers and devices.

[0143] The embodiments were chosen and described in order to best explain the principles of the invention and its practical applications, thereby enabling others skilled in the art to understand the invention for various embodiments as well as with the modifications and adaptations that are suited to the particular use contemplated.

Claims

1. An intelligent and integrated system for clinical assessment and intervention of dyslexia based on strengths and weaknesses, characterized by: include: Databases, including large-scale archives of multiple user performance collections across different domains; a first subgroup generator for performing a theory-driven machine learning method on the database by obtaining a layer having nodes, the nodes being calculated by summing all component values ​​multiplied by factor loadings of the component values ​​using exploratory factor analysis, wherein subgroups are generated according to whether the nodes are activated; a second subgroup generator for performing a theoretical and data-driven combined machine learning method on the database by training a neural network model by applying principal component analysis (PCA) via an encoder and a decoder, wherein a data set of each user is encoded into a three-dimensional vector by the encoder, and the users are clustered into hierarchical clusters based on similarities or distances of the users calculated between the users to group the users; a comparator that compares the theory-driven machine learning method with the theory- and data-driven combined machine learning method to determine whether the theory-driven machine learning method is consistent with the theory- and data-driven combined machine learning method; A rating module that calculates the rating of each domain relative to users aggregated in different subgroups; and An intervention module, configured to provide personalized intervention for each of the subgroups, so that users in different subgroups receive interventions of different methods; The intervention includes general cognitive and music skills in one or more areas, or language-specific intervention training in listening, speaking, reading, writing, or a combination thereof.

2. The system according to claim 1, characterized in that The second subgroup generator continues clustering users in the hierarchical clustering until only two clusters exist.

3. The system according to claim 1, characterized in that An updating unit is also included, which is used to dynamically update the subgroup when a new user enters the database.

4. The system according to claim 1, characterized in that The intervention included a visual-motor intervention game.

5. The system according to claim 4, characterized in that The visual-motor game guides the user to perform one or more game exercises in pattern search, object matching pairs, or sequence memory.

6. The system according to claim 1, characterized in that The intervention involves detecting and distinguishing the difference between speech and non-speech.

7. The system according to claim 1, characterized in that The intervention system described herein gradually increases the difficulty level.

8. The system according to claim 1, characterized in that The intervention included vocabulary games.

9. The system according to claim 8, characterized in that The vocabulary games include vocabulary breadth and depth.

10. The system according to claim 1, characterized in that The intervention included a music assessment game that assessed perception of basic musical elements as well as musical emotions and preference for different musical instruments.

11. The system according to claim 1, characterized in that The user representation set includes data from dyslexic monolingual and bilingual users whose first language is Chinese and whose second language is English.

12. The system according to claim 11, characterized in that The dyslexic bilingual user is a Chinese and English dyslexic bilingual user.

13. A clinical assessment and intervention device, characterized in that: include: The system according to claim 1; user interface; monitor; and processor.

14. A mobile device selected from one or more of a mobile phone, a tablet computer or a laptop computer, characterized in that: Comprising the clinical assessment diagnosis and intervention training device according to claim 13.

Citation Information

Cited By

  • Modeling and intervention simulation method and system for children with difficulty in reading and writing Chinese based on digital twin brain technology

    CN121983246A

  • A method and system for modeling and intervention simulation of children with Chinese reading and writing difficulties based on digital twin brain technology

    CN121983246B