Method and system for attention function intervention based on multi-sensory music

Abstract

This invention discloses a method and system for attention function intervention based on multi-sensory music. The method includes: obtaining the user's attention function assessment results; establishing a decision tree model to obtain the classification probabilities of each set attention function impairment type corresponding to the user; obtaining the user's human-computer interaction scheme based on multi-sensory music; obtaining the user's human-computer interaction results and evaluating their effectiveness; iterating the decision tree model based on the effectiveness evaluation results to update the classification probabilities of each set attention function impairment type corresponding to the user, and updating the human-computer interaction scheme, until the user passes the effectiveness evaluation of the human-computer interaction scheme. This method can effectively improve the user's attention function, not only by training a specific attention type, but also by emphasizing comprehensive training of different dimensions of attention. Patients with different attention disorders can choose appropriate music therapy to achieve improvement. The content is rich and diverse, meeting the intervention needs of different attention disorders.

Description

Technical Field

[0001] This invention relates to a method for intervening in attentional function based on multi-sensory music, and also to a corresponding system for intervening in attentional function, belonging to the field of medical and health care information technology. Background Technology

[0002] Music therapy originated in the United States and is a comprehensive applied discipline integrating music, medicine, and psychology. Professor K. Bruscia of Temple University, in his book *Defining Music Therapy*, proposed that music intervention is a systematic process of disease intervention. In this process, therapists utilize various forms of musical experience, such as listening, singing, instrumental performance, music composition, lyric writing, improvisation, and dance, as well as the therapeutic relationship that develops during the treatment process and serves as the driving force for therapy, to help the recipient achieve health.

[0003] Music possesses characteristics such as fluency and rhythm. Music therapy can systematically utilize these features to influence the human body, improving both physical and neuropsychological conditions. These effects include lowering blood pressure, improving sleep and immune function, relieving pain, improving cognitive function, and reducing anxiety and depression. Music therapy enables patients to achieve psychological and physiological integration during the treatment of illness or disability, ultimately improving higher brain functions such as attention and memory.

[0004] From a neuropsychological perspective, exploring the mechanism by which music intervention improves attentional function reveals that listening to music can stimulate brain plasticity and lead to the activation of multiple brain regions, such as auditory mechanisms, attention, memory storage and retrieval, and sensory-motor integration. A growing body of research indicates that music activates many brain regions involved in attention processes.

[0005] Currently, most clinical research on attention problem interventions utilizes traditional music intervention methods. However, traditional music therapy requires highly skilled therapists and is not widely available; it also lacks systematic and scientific music intervention strategies for attention problems, making it impractical. Furthermore, research primarily focuses on single dimensions of attention function, with almost no studies addressing multiple specific types of attention deficit disorders. Summary of the Invention

[0006] The primary technical problem to be solved by this invention is to provide a method for intervening in attentional function based on multi-sensory music.

[0007] Another technical problem to be solved by the present invention is to provide an attention function intervention system based on multi-sensory music.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] According to a first aspect of the present invention, a method for intervening in attentional function based on multi-sensory music is provided, comprising the following steps:

[0010] Obtain the user's attention function evaluation results;

[0011] A decision tree model is established based on the user's attention function assessment results to obtain the classification probability of each set attention function impairment type corresponding to the user;

[0012] Based on the classification probability of each set attention function impairment type corresponding to the user, determine the priority of each set attention function of the user that needs to be improved, so as to obtain the human-computer interaction scheme of the user based on multi-sensory music;

[0013] Obtain the human-computer interaction results of the user performing human-computer interaction in sequence according to the priority, and evaluate the effectiveness of the human-computer interaction scheme based on the human-computer interaction results;

[0014] The decision tree model is iterated based on the effectiveness evaluation results to update the classification probabilities of each set attentional dysfunction type corresponding to the user, and the human-computer interaction scheme is updated until the user passes the effectiveness evaluation of the human-computer interaction scheme.

[0015] Preferably, the acquisition of the user's attention function evaluation results specifically includes:

[0016] Based on the user's diagnosis results, obtain the user's existing disease information;

[0017] Speech recognition technology is used to extract features from the user's recorded speech text information to obtain the word frequency of the user for each set attention impairment type;

[0018] Based on the user's task test results for each set attention impairment type, obtain the user's task evaluation score for each set attention impairment type;

[0019] The user's attention acquisition function needs to be improved;

[0020] By combining the symptom information, word frequency, task evaluation score, and improvement needs, the user's attention function evaluation result is obtained.

[0021] Preferably, each of the specified attentional dysfunction types includes at least: stable attentional dysfunction, selective attentional dysfunction, and persistent attentional dysfunction.

[0022] Preferably, the step of establishing a decision tree model based on the user's attention function assessment results to obtain the classification probability of each set attention function impairment type corresponding to the user specifically includes:

[0023] Multinomial logistic regression was used for classification, and the following three different decision trees were built for the three types of attentional dysfunction;

[0024]

[0025]

[0026]

[0027] Wherein, the input training set T = {(x1, y1), (x2, y2), ..., (x m y m The superposition objective of F(x) from 1 to M is to determine the parameters of the next decision tree by minimizing empirical risk, and by fitting the residual c. m Learning a regression tree, we obtain the leaf node region R of the m-th regression tree. m By finding the regression tree with the smallest error corresponding to c m To update the obtained regression problem boosting tree F M (x);

[0028] Based on the three established decision trees, the user's attention function assessment results are input into the following formula to obtain the classification probability p of the three types of attention function impairments corresponding to the user. i (x);

[0029]

[0030] Where i takes values ​​from 1 to 3; exp() represents the exponential function, F iM (x) represents the best-fit regression tree.

[0031] Preferably, based on the classification probability of each type of attentional impairment corresponding to the user, the priority of each attentional function of the user that needs to be improved is determined to obtain the human-computer interaction scheme based on multi-sensory music for the user, specifically including:

[0032] Based on the priority of each attention function set by the user, and according to the classification probability of each attention function impairment type corresponding to the user, human-computer interaction tasks based on multi-sensory music are sequentially matched from the human-computer interaction task library to the corresponding attention function impairment type.

[0033] Based on the classification probability of each set attentional dysfunction type corresponding to the user, calculate the number of human-computer interaction days for each set attentional dysfunction type.

[0034] The human-computer interaction tasks based on multi-sensory music corresponding to each type of attention deficit disorder are matched one-to-one with the number of human-computer interaction days for each type of attention deficit disorder, forming human-computer interaction sub-schemes corresponding to each type of attention deficit disorder.

[0035] The human-computer interaction sub-solutions are sorted according to their priority to form the human-computer interaction scheme based on multi-sensory music for the user.

[0036] Preferably, in the human-computer interaction scheme, after the current human-computer interaction sub-scheme is completed, the human-computer interaction of the next level human-computer interaction sub-scheme is started.

[0037] Preferably, the number of human-computer interaction days for any given attentional dysfunction type is obtained by multiplying the classification probability of that attentional dysfunction type by the optimal number of intervention days for the user to completely belong to that attentional dysfunction type.

[0038] According to a second aspect of the present invention, an attention function intervention system based on multi-sensory music is provided, comprising a processor and a memory, wherein the processor reads a computer program in the memory for performing the following operations:

[0039] Obtain the user's attention function evaluation results;

[0040] A decision tree model is established based on the user's attention function assessment results to obtain the classification probability of each set attention function impairment type corresponding to the user;

[0041] Based on the classification probability of each set attention function impairment type corresponding to the user, determine the priority of each set attention function of the user that needs to be improved, so as to obtain the human-computer interaction scheme of the user based on multi-sensory music;

[0042] Obtain the human-computer interaction results of the user performing human-computer interaction in sequence according to the priority, and evaluate the effectiveness of the human-computer interaction scheme based on the human-computer interaction results;

[0043] The decision tree model is iterated based on the effectiveness evaluation results to update the classification probabilities of each set attentional dysfunction type corresponding to the user, and the human-computer interaction scheme is updated until the user passes the effectiveness evaluation of the human-computer interaction scheme.

[0044] Compared with the prior art, the present invention has the following technical effects:

[0045] (1) Multisensory comprehensive intervention mode can effectively improve users' attention function. Multisensory music therapy is a method that integrates vision, hearing and movement and adopts multisensory comprehensive intervention. By stimulating multiple senses, it changes their psychological and physiological indicators, so as to fully improve attention disorders related to multiple information channels.

[0046] (2) Provides operable and systematic training programs. Digital music therapy, developed based on therapeutic theories such as Music Attention Control Training (MACT) and visual music therapy, makes up for the current shortage of professional therapists and also provides patients with a more convenient intervention method to improve attention problems.

[0047] (3) Training patients' attentional function in multiple ways. This is not limited to training a specific type of attention, but emphasizes comprehensive training across different dimensions of attention. Patients with different attention disorders can choose appropriate music therapy to achieve improvement; the content is rich and diverse, meeting the intervention needs of different attention disorders. Attached Figure Description

[0048] Figure 1 A flowchart illustrating an overall process for intervening in attentional function based on multi-sensory music, as provided in an embodiment of the present invention.

[0049] Figure 2 A flowchart illustrating a method for intervening in attentional function based on multi-sensory music, provided in an embodiment of the present invention;

[0050] Figure 3 This is a structural diagram of an attention function intervention system based on multi-sensory music, provided in an embodiment of the present invention. Detailed Implementation

[0051] The technical content of the present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0052] The overall concept of this invention lies in designing a user-friendly and widely applicable visual music intervention strategy that combines active and passive musical activities to improve attention problems in patients with attention deficit disorder. Therefore, as... Figure 1 As shown, the present invention provides a method for intervening in attentional function based on multi-sensory music, which includes at least steps S1 to S5:

[0053] S1: Obtain the user's attention function evaluation results.

[0054] Specifically, this includes sub-steps S11 to S15:

[0055] S11: Obtain the user's existing medical condition information.

[0056] When a user logs in for the first time, they need to fill in their diagnosed condition αn (n = 1, 2, 3, 4, 5...) in the list of common diagnosed conditions, based on their own medical diagnosis. This strengthens the weight of the corresponding condition category in the human-computer interaction scheme. If the user has no diagnosed condition, the system default value is selected.

[0057] S12: Obtain the word frequency for each set attention impairment type set by the user.

[0058] After the user selects their diagnosed condition, they need to describe their attention symptoms and record the conversation. Speech recognition technology is used to extract features from the user's recorded speech, thereby extracting keywords related to each set type of attention disorder. The frequency of words belonging to each category is statistically analyzed, and the frequency φn (n = 1, 2, 3…) in each category is used as one of the bases for recommending human-computer interaction solutions.

[0059] In one embodiment of the present invention, the various attention impairment types may include three categories: stable attention impairment, selective attention impairment, and persistent attention impairment. Therefore, the attention word frequency classification based on semantic extraction is mainly: stable attention impairment φ1, selective attention impairment φ2, and persistent attention impairment φ3. It is understood that in other embodiments, the various attention impairment types may also be divided into four or more categories as needed.

[0060] S13: Obtain the user's task evaluation score for each set attention impairment type.

[0061] After a user reports their attention symptoms, an attention function assessment is required. Based on the assessment score λn (n = 1, 2, 3, ...), the user's specific symptoms are clarified, which is beneficial for accurately pushing personalized human-computer interaction solutions. In this embodiment of the invention, stable attention λ1, selective attention λ2, and shifting attention λ3 are mainly assessed.

[0062] Among them, the Continuous Task Test (CPT) is used to assess the stability of attention. This method is presented through a computer program. Multiple stimuli are presented on the computer, and when the target stimulus is presented, the user is instructed to make a key response as quickly as possible. The computer records data such as average reaction time and number of hits, and calculates these data to reflect the ability to maintain attention.

[0063] The Stroop Color Word Test (SCWT) is used to assess selective attention. It is a computer program that presents four colored words (red, green, yellow, and blue) on a screen. Each word is displayed in a different color (red, green, yellow, and blue). Participants are asked to name the color of the word as quickly as possible, and the number of correct names given within one minute is recorded.

[0064] The Line Connecting Test B (TMT-B) is used to assess attention shift. The numbers 1-12 and the letters A-L are printed on the screen in a random order. The subject connects the numbers and letters in a line, and is required to do so in the correct order, i.e., 1-A-2-B-3-C-4-D...11-K-12-L. The time taken to complete the test is recorded (s).

[0065] S14: Improved requirements for user attention acquisition functionality.

[0066] After completing the attention function test, users also need to select and rank their attention function improvement needs, i.e., the degree of need μn (n = 1, 2, 3, ...), so as to obtain the attention problems that users urgently want to solve and provide a basis for pushing human-computer interaction solutions.

[0067] In one embodiment of the present invention, the user has three types of needs: improved stable attention μ1; improved selective attention μ2; and improved transferable attention μ3. The user inputs a rating of 1 to 9 according to the degree of their need, and the ratings are arranged from high to low. The higher the rating, the greater the degree of need.

[0068] S15: Based on comprehensive symptom information, word frequency, task evaluation score, and improvement needs, obtain the user's attention function evaluation results.

[0069] After obtaining the above information, the user's attention function is comprehensively evaluated based on four factors: symptom information, word frequency, task evaluation score, and improvement needs. This evaluation serves as the basis for pushing subsequent human-computer interaction solutions.

[0070] S2: Build a decision tree model based on the user's attention function assessment results to obtain the classification probability of each set attention function impairment type for the user.

[0071] In one embodiment of the present invention, the user's attention impairment is mainly classified into the following three types, which are described in detail below:

[0072] (1) Stable attention deficit: The ability to maintain attention on an activity or stimulus for a sustained period of time, also known as concentration, is related to alertness and depends on the level of sustained arousal. This is also the baseline for information processing, such as driving on the road, watching TV, and observing patients during functional training. Stable attention is mainly trained through listening to music, playing instruments, and learning rhythms.

[0073] (2) Selective attention deficit: The ability to choose relevant activities or tasks while ignoring irrelevant stimuli (such as external noise, internal worries, etc.); for example, while others are watching TV in the living room, you are reading a newspaper or doing homework. This is related to the intentional selection of a particular activity. Training includes: rhythm perception, timbre perception, pitch perception, and melody perception.

[0074] (3) Attention shifting disorder: The ability to flexibly shift attention between two activities; for example, if the phone rings while you are doing a task, you will pause your work to answer the phone and then resume your work. This is mainly trained using vocalizations and animated music.

[0075] Specifically, this includes steps S21 to S22:

[0076] S21: Using multinomial logistic regression for classification, and utilizing the log loss function of softmax regression, three different decision trees are established for the three types of attentional impairments.

[0077] Specifically, the Gradient Boosting Decision Tree (GBDT) algorithm in supervised learning is used to build a decision tree model for user classification. Specifically, user profiles (diagnosed symptoms, semantic attention feature extraction, attention function task evaluation, and user improvement needs) and task profiles (training time and benefit levels for three types of tasks: stable attention training, selective attention training, and shifting attention training) data from the previous training database are used. Since the goal is to classify users into three categories δ: stable attention impairment δ1, selective attention impairment δ2, and shifting attention impairment δ3, corresponding to stable attention training ξ1, selective attention training ξ2, and shifting attention training ξ3 in the training database, multinomial logistic regression (Softmax Regression) is used for classification. The logarithmic loss function of softmax regression is used to build three different decision trees as shown below:

[0078]

[0079]

[0080]

[0081] Wherein, the input training set T = {(x1, y1), (x2, y2), ..., (x m y m The superposition objective of F(x) from 1 to M is to determine the parameters of the next decision tree by minimizing empirical risk, where the parameters are obtained by fitting the residual c. m Let's learn a regression tree and obtain the leaf node region R of the m-th tree.m By finding the tree c with the smallest error m To update and obtain the boosting tree F for the regression problem M (x). F 1M (x), F 2M (x), F 3M (x) represents the decision trees built for each type of insomnia, R 1m Representing the leaf node region for the first type of insomnia, c 1m This represents minimizing the fitting residual for the first type of insomnia. Classification and regression trees (CART) are used respectively, recursively constructing binary decision trees. The mean squared error loss criterion is used to determine the split points for the regression tree, traversing all feature values ​​and using each feature value as a split point. An iterative boosting algorithm continuously adjusts the distribution of interactive samples based on the performance of the base learners, combining them into a strong learner to improve the accuracy of the regression algorithm in predicting each category.

[0082] S22: Based on the three decision trees established, input the user's sleep quality monitoring results into the following formula to obtain the classification probability of the user corresponding to the three attentional dysfunctions.

[0083] Specifically, using the three regression tree formulas above, input the information x of the newly arrived patient, and input the probability of the corresponding three categories: primary insomnia, secondary insomnia, and mood insomnia. The formulas are as follows:

[0084]

[0085] Where i takes values ​​from 1 to 3; exp() represents the exponential function, and Fim(x) represents the best-fit regression tree.

[0086] After calculating the classification probability of the current patient, based on the mapping relationship between the patient's classification attributes and the interaction module duration, the prediction variables for the three intervention modules are input: namely, the interaction duration of each module.

[0087] S3: Based on the classification probability of each user's set attention function impairment type, determine the priority of each user's set attention function that needs to be improved, so as to obtain the user's human-computer interaction solution based on multi-sensory music.

[0088] Specifically, this includes sub-steps S31 to S34:

[0089] S31: Obtain human-computer interaction tasks corresponding to each type of attentional dysfunction.

[0090] Specifically, after calculating the classification probabilities of the user's three types of attentional dysfunction in step S2, the higher the classification probability, the more severe the impairment of that type of attentional function, and therefore, the higher the priority for improvement. Thus, based on the classification probabilities, the order in which the three types of attentional dysfunction need to be improved can be determined. For example, if the calculated probability of the user having stable attentional dysfunction is 0.7, selective attentional dysfunction is 0.2, and shifting attentional dysfunction is 0.1, then the order in which the user's attentional function needs to be improved is: first improve stable attentional function—then improve selective attentional function—and finally improve shifting attentional function.

[0091] Based on the above priorities, and according to the classification probabilities of the user's corresponding to the three types of attentional dysfunction, multi-sensory music-based human-computer interaction tasks are sequentially matched from the human-computer interaction task library to each type of attentional dysfunction. Each human-computer interaction task includes at least: task type, task level, and number of tasks.

[0092] In one embodiment of the present invention, the human-computer interaction task library includes at least: listening to music, percussion, and rhythm learning tasks corresponding to stable attention deficit; rhythm perception, timbre perception, pitch perception, and melody perception tasks corresponding to selective attention deficit; and sound learning and animation music tasks corresponding to sustained attention deficit.

[0093] Specifically, the contents of this human-computer interaction task library are as follows:

[0094] Unit A: Stability Training

[0095] Specifically, it includes tasks such as listening to music, playing percussion instruments, and learning rhythm.

[0096] (1) Listening to music: Based on evidence-based research, soothing music and alpha music can improve attention. The music library contains two types of music, including soothing music with a beat of 60-70 (such as classical music, Baroque music, etc.) and alpha music. Users can choose to listen to music according to their preferences.

[0097] (2) Percussion instruments: Instruments include piano, electronic keyboard, drum kit, taiko drum, folk instruments, etc. Users first enter the instrument selection interface, select the instrument to learn, and then select the music to practice. They can choose music from three types of music: beginner, intermediate and advanced. Then the interface presents the instrument teaching mode, and users need to imitate and learn the basic percussion and playing methods of simple instruments.

[0098] (3) Rhythm Learning: This includes rhythm performance tasks, rhythm-based tasks, and rhythm-differentiated tasks. In the rhythm performance task, users need to play along with the notes on the interface. In the rhythm-based tasks, users need to listen to a rhythm while observing the positions for tapping on the interface and tapping along with the rhythm. In the rhythm-differentiated tasks, users listen to a rhythm and then listen to three different rhythms from the options, comparing which one is the same as the one they heard before.

[0099] Unit B: Selective Attention Training

[0100] Specifically, it includes rhythm perception tasks, timbre perception tasks, pitch perception tasks, and melody perception tasks.

[0101] (1) Rhythm perception: including rhythm detection task (the user needs to detect the presence or absence of rhythm) and rhythm discrimination task (the user needs to discern the speed and intensity of rhythm).

[0102] (2) Timbre perception: including the task of recognizing timbre (the user needs to recognize different timbres) and the task of distinguishing timbre (distinguishing between two or more timbres).

[0103] (3) Pitch perception: including pitch perception task (users need to recognize the pitch of a sound) and difference discrimination task (users need to perceive the difference between two or more pitches).

[0104] (4) Melody perception: including the task of perceiving melody (users need to perceive different melodies).

[0105] Unit C: Transfer Attention Training

[0106] Specifically, this includes tasks related to sound and momentum learning, and tasks related to animation soundtracks.

[0107] (1) Learning of sound and gestures: From the perspective of attention test design, learning of sound and gestures is to enable students to understand and identify the meaning of each sound and gesture symbol, and combine it with the music theory knowledge exercises mentioned above to accurately retrieve the sound and gesture actions that need to be performed, so as to achieve the purpose of accurately identifying symbols and accurately performing actions. This part of the experiment can cultivate the user's attention transfer and allocation.

[0108] Three symbols are marked on the musical score, representing stomping, clapping, and patting the shoulder, respectively. Students clap according to the meaning of the symbols, and during this process, users need to clap accurately and in sync with the music. The vocalization learning includes Vocalization Learning 1 (clapping, patting the shoulder), Vocalization Learning 2 (clapping, stomping), and Vocalization Learning 3 (clapping, patting the shoulder, stomping).

[0109] (2) Animation background music: Users need to select background music of the appropriate style according to the type of animation.

[0110] S32: Calculate the number of days of human-computer interaction corresponding to each set attention function impairment type.

[0111] Once the human-computer interaction tasks for the three types of attentional impairments are determined, it is necessary to calculate the number of human-computer interaction days corresponding to each of the three types of attentional impairments. Specifically, in one embodiment of the present invention, the number of human-computer interaction days for any given type of attentional impairment is obtained by multiplying the classification probability of that type of attentional impairment by the optimal number of intervention days for the user to completely belong to that type of attentional impairment.

[0112] For example: Based on the three decision tree models in step S2, the corresponding classification probabilities are calculated as follows: stable attention deficit 0.7, selective attention deficit 0.2, and shifting attention deficit 0.1. The severity of attention deficit is 0.7 compared to the norm, meaning it exceeds the attention deficit situation of more than 70% of people. Then, based on the classification output predicted response value calculated by the model, the training duration for stable attention intervention for this user is 0.7 * the optimal time for stable attention intervention (X days), the training duration for selective attention intervention is 0.2 * the optimal time for selective attention intervention (Y days), and the training duration for shifting attention intervention is 0.1 * the optimal time for shifting attention intervention (Z days). (X, Y, and Z are obtained by weighting the optimal mapping segmentation standard feature values ​​generated by the three decision tree models and the user's attention severity of 0.7, representing the optimal intervention time for this user's attention situation if it completely belongs to a certain type of attention deficit). A total of T days are used as the initial training course. Where T = X + Y + Z.

[0113] S33: Obtain human-computer interaction sub-schemes corresponding to each type of attentional dysfunction.

[0114] Specifically, the three types of human-computer interaction tasks obtained in step S31 are matched one-to-one with the three human-computer interaction days obtained in step S32, thereby forming three human-computer interaction sub-solutions for the three types of attentional dysfunction.

[0115] S34: Sort the various human-computer interaction sub-solutions according to priority to form a human-computer interaction scheme based on multi-sensory music for users.

[0116] Specifically, the three human-computer interaction sub-solutions obtained in step S33 are sorted according to priority to form an overall human-computer interaction solution, and the user's three types of attention functions are intervened and improved in sequence.

[0117] Among the three human-computer interaction sub-schemes, the highest priority sub-scheme (i.e., the user's weakest attention function) will be interacted with first. After the user completes the current intervention improvement, the next level of human-computer interaction sub-scheme (i.e., the user's relatively weak attention function) will then be interacted with, and so on. Moreover, the user's daily interaction content will be randomly selected from three human-computer interaction tasks in the human-computer interaction task library based on the human-computer interaction tasks corresponding to the currently improved attention function.

[0118] S4: Evaluate the effectiveness of the human-computer interaction solution.

[0119] Specifically, after the user completes the entire human-computer interaction process according to priority, the effectiveness of the human-computer interaction solution can be evaluated by obtaining the human-computer interaction results of the entire process.

[0120] Once all three human-computer interaction sub-solutions have passed the validity test, the current human-computer interaction process ends. If any one of the human-computer interaction sub-solutions fails to pass the validity test, it is necessary to proceed to step S5 to carry out the next stage of intervention and improvement.

[0121] S5: Moving to the next stage of intervention enhancement.

[0122] Specifically, in one embodiment of the present invention, when the effectiveness of any human-computer interaction sub-scheme fails, the decision tree model is iterated based on the effectiveness evaluation result to return to step S2, update the classification probabilities of each set attentional dysfunction type corresponding to the user, and then update the human-computer interaction scheme again through step S3. After the user has completed the interaction process of the updated human-computer interaction scheme, the effectiveness of the user is evaluated again until the user passes the effectiveness evaluation of the human-computer interaction scheme.

[0123] Based on the aforementioned method for intervening in attentional function using multi-sensory music, this invention further provides a system for intervening in attentional function using multi-sensory music. For example... Figure 3 As shown, the attention function intervention system includes one or more processors 21 and a memory 22. The memory 22 is coupled to the processors 21 and is used to store one or more programs. When the one or more programs are executed by the one or more processors 21, the one or more processors 21 implement the attention function intervention method as described in the above embodiments.

[0124] The processor 21 controls the overall operation of the intervention system to complete all or part of the steps of the aforementioned attention function intervention method. The processor 21 can be a central processing unit (CPU), graphics processing unit (GPU), field-programmable gate array (FPGA), application-specific integrated circuit (ASIC), digital signal processing (DSP) chip, etc. The memory 22 stores various types of data to support the operation of the intervention system. This data may include, for example, instructions for any application or method used to operate on the intervention system, as well as application-related data. The memory 22 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, etc.

[0125] In one exemplary embodiment, the intervention system may be implemented by a computer chip or physical entity, or by a product with certain functions, to perform the attention function intervention method described above and achieve the same technical effect as the method described above. A typical embodiment is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, an in-vehicle human-machine interaction device, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0126] In another exemplary embodiment, the present invention also provides a computer-readable storage medium including program instructions that, when executed by a processor, implement the steps of the attention function intervention method in any of the above embodiments. For example, the computer-readable storage medium may be a memory including program instructions that can be executed by the processor of the intervention system to complete the above-described attention function intervention method and achieve the same technical effects as the described method.

[0127] In summary, the attention function intervention method and system based on multi-sensory music provided by this invention have the following beneficial effects:

[0128] (1) Multisensory comprehensive intervention mode can effectively improve users' attention function. Multisensory music therapy is a method that integrates vision, hearing and movement and adopts multisensory comprehensive intervention. By stimulating multiple senses, it changes their psychological and physiological indicators, so as to fully improve attention disorders related to multiple information channels.

[0129] (2) Provides operable and systematic training programs. Digital music therapy, developed based on therapeutic theories such as Music Attention Control Training (MACT) and visual music therapy, makes up for the current shortage of professional therapists and also provides patients with a more convenient intervention method to improve attention problems.

[0130] (3) Training patients' attentional function in multiple ways. This is not limited to training a specific type of attention, but emphasizes comprehensive training across different dimensions of attention. Patients with different attention disorders can choose appropriate music therapy to achieve improvement; the content is rich and diverse, meeting the intervention needs of different attention disorders.

[0131] The above provides a detailed description of the attention function intervention method and system based on multi-sensory music provided by this invention. Any obvious modifications made by those skilled in the art without departing from the essence of this invention will constitute an infringement of the patent rights of this invention and will incur corresponding legal liability.

Claims

1. A method for intervening in attentional function based on multi-sensory music, characterized in that... Includes the following steps: Obtain the user's attention function evaluation results; A decision tree model is established based on the user's attention function assessment results to obtain the classification probability of each set attention function impairment type corresponding to the user; The types of attentional dysfunctions described above include at least: stable attentional dysfunction, selective attentional dysfunction, and persistent attentional dysfunction; The establishment of a decision tree model based on the user's attention function assessment results specifically includes: using multinomial logistic regression for classification, and establishing the following three different decision trees for the three types of attention function impairments; Wherein, the input training set T = {(x1, y1), (x2, y2), ..., (x m y m The goal of stacking F(x) from 1 to M is to determine the parameters of the next decision tree by minimizing empirical risk. A regression tree is learned by fitting the residual cm, and the leaf node region R of the m-th regression tree is obtained. m The boosting tree F for the regression problem is updated by finding the cm corresponding to the regression tree with the smallest error. M (x); The step of obtaining the classification probabilities of each set attentional dysfunction type corresponding to the user specifically includes: based on the three established decision trees, inputting the user's attentional function assessment results into the following formula to obtain the classification probabilities p of the user corresponding to the three types of attentional dysfunction. i (x); Where i takes values ​​from 1 to 3; exp() represents the exponential function, F i M(x) represents the best-fit regression tree; Based on the classification probability of each set attention function impairment type corresponding to the user, determine the priority of each set attention function of the user that needs to be improved, so as to obtain the human-computer interaction scheme of the user based on multi-sensory music; Obtain the human-computer interaction results of the user performing human-computer interaction in sequence according to the priority, and evaluate the effectiveness of the human-computer interaction scheme based on the human-computer interaction results; The decision tree model is iterated based on the effectiveness evaluation results to update the classification probabilities of each set attentional dysfunction type corresponding to the user, and the human-computer interaction scheme is updated until the user passes the effectiveness evaluation of the human-computer interaction scheme.

2. The attention function intervention method as described in claim 1, characterized in that, The evaluation results of the user's attention function specifically include: Based on the user's diagnosis results, obtain the user's existing disease information; Speech recognition technology is used to extract features from the user's recorded speech text information to obtain the word frequency of the user for each set attention impairment type; Based on the user's task test results for each set attention impairment type, obtain the user's task evaluation score for each set attention impairment type; The user's attention acquisition function needs to be improved; By combining the symptom information, word frequency, task evaluation score, and improvement needs, the user's attention function evaluation result is obtained.

3. The attention function intervention method as described in claim 1, characterized in that, Based on the classification probability of each type of attentional impairment corresponding to the user, the priority of each attentional function of the user that needs to be improved is determined, so as to obtain the human-computer interaction scheme based on multi-sensory music for the user, specifically including: Based on the priority of each attention function set by the user, and according to the classification probability of each attention function impairment type corresponding to the user, human-computer interaction tasks based on multi-sensory music are sequentially matched from the human-computer interaction task library to the corresponding attention function impairment type. Based on the classification probability of each set attentional dysfunction type corresponding to the user, calculate the number of human-computer interaction days for each set attentional dysfunction type. The human-computer interaction tasks based on multi-sensory music corresponding to each type of attention deficit disorder are matched one-to-one with the number of human-computer interaction days for each type of attention deficit disorder, forming human-computer interaction sub-schemes corresponding to each type of attention deficit disorder. The human-computer interaction sub-solutions are sorted according to their priority to form the human-computer interaction scheme based on multi-sensory music for the user.

4. The attention function intervention method as described in claim 3, characterized in that, In the human-computer interaction scheme, after the current human-computer interaction sub-scheme is completed, the human-computer interaction of the next level human-computer interaction sub-scheme is started.

5. The attention function intervention method as described in claim 3, characterized in that, The number of days of human-computer interaction for any given attention deficit disorder type is obtained by multiplying the classification probability of that attention deficit disorder type by the optimal number of intervention days for the user to completely belong to that attention deficit disorder type.

6. A multi-sensory music-based attention function intervention system, characterized in that... It includes a processor and a memory, wherein the processor reads a computer program from the memory for executing the attention function intervention method as described in any one of claims 1-5.

Images