Attention training method and device based on dynamic sensory stimulation

Abstract

The invention provides an attention training method and device based on dynamic sensory stimulation. The method comprises: generating and presenting a game interaction interface to a user, the game interaction interface comprising at least one game task for training an attention level of the user, the game task being presented by at least one sensory stimulus, at least one variation dimension of the at least one sensory stimulus is composed of at least one variable having a plurality of values; obtaining performance information when the user performs the at least one game task, wherein the performance information is used for indicating the attention concentration degree of the user under the condition of different values of at least one variable; and determining the attention level of the user according to the performance information. Through the scheme, the diversity and interestingness of the attention training process are improved.

Description

Technical Field

[0001] The present application relates to the field of attention training, and in particular to a method, device, electronic device and computer-readable storage medium for attention training based on dynamic sensory stimulation. Background Art

[0002] Attention-deficit / hyperactivity disorder (ADHD) is a common neurodevelopmental disorder characterized by inattention, hyperactivity, and impulsive behavior. The principle of neuroplasticity suggests that ADHD patients experience functional abnormalities in the prefrontal cortex (responsible for executive function) and basal ganglia (regulating attention). In recent years, studies have shown that repeated attention training, such as working memory training and concentration tasks, can alleviate symptoms by enhancing neuroplasticity and improving functional connectivity in relevant brain regions. This has been confirmed by numerous neuroscience studies to improve ADHD symptoms.

[0003] However, existing attention training systems have a single training format and a weak feedback mechanism. They mostly use repetitive tasks such as clicking on targets and sorting numbers. They generally have problems such as lack of fun and poor adaptability, making it difficult to maintain the training interest of users such as child patients. Summary of the Invention

[0004] In view of this, the present application provides an attention training method, device, equipment and computer-readable storage medium based on dynamic sensory stimulation, which improves the diversity and fun of the attention training process by combining the training tasks of attention training with a gamification design with dynamic sensory stimulation changes.

[0005] The present application is introduced below from multiple aspects, and the implementation methods and beneficial effects of the following multiple aspects can be referenced to each other.

[0006] In a first aspect, the present application provides an attention training method based on dynamic sensory stimulation, comprising: generating and presenting a game interaction interface to a user, the game interaction interface including at least one game task for training the user's attention level, the game task being presented through at least one sensory stimulation, and at least one change dimension of the at least one sensory stimulation being composed of at least one variable having multiple values; obtaining performance information of the user when performing the at least one game task, the performance information being used to indicate the user's degree of concentration under different values of at least one of the variables; and determining the user's attention level based on the performance information.

[0007] According to the implementation mode of the present application, attention training is combined with user interests, which is more interactive and has richer content, thereby improving the user's attention level.

[0008] In a possible implementation of the first aspect above, the performance information includes real-time game performance information, and the method further includes: adjusting the difficulty of the game task according to the real-time game performance information of the user when performing the game task.

[0009] In a possible implementation of the first aspect above, the real-time game performance information includes an ability level and a load level. The ability level is used to indicate the user's degree of completion of the game task, and the load level is used to indicate the user's ability to withstand stress from the game task.

[0010] According to the implementation method of the present application, the method of adjusting the difficulty of the game is more reasonable, avoiding short-term fluctuation interference (such as accidental mistakes) caused by a single algorithm relying only on behavioral data and prematurely increasing the difficulty (such as the player's cognitive load is not ready).

[0011] In a possible implementation of the first aspect, the load level includes one or more of the following: an error rate of the game task and a ratio between theta waves and beta waves of an electroencephalogram (EEG) signal of the user.

[0012] In a possible implementation of the first aspect above, the performance information includes the user's first EEG signal and brain topography, and determining the user's attention level includes: performing brain microstate analysis and / or spectral analysis on the user's first EEG signal, and performing visual analysis on the brain topography to obtain analysis results; and determining the user's attention level based on the analysis results.

[0013] According to the implementation of the present application, more features are extracted from the multiple dimensions of microstates, spectra, and brain topography, which can be said to be more comprehensive and the reliability of analysis and demonstration is higher.

[0014] In a possible implementation of the first aspect above, obtaining the performance information of the user when performing the at least one game task includes: obtaining a second EEG signal, the second EEG signal including the first EEG signal and an interference signal, the interference signal including a non-neurogenic artifact signal, the non-neurogenic artifact signal including a scalp blood flow artifact signal and a physiological artifact signal; removing the scalp blood flow artifact signal from the second EEG signal by a short-range regression (SSR) method, and removing the physiological artifact signal from the second EEG signal by an independent component analysis (ICA) method to obtain the first EEG signal.

[0015] According to the implementation mode of the present application, by combining SSR and ICA, the neural source correlation of the EEG signal is effectively improved, the interference of non-neural artifacts is reduced, and high-quality input data is provided for subsequent system analysis.

[0016] In a possible implementation of the first aspect above, a Canopy algorithm is introduced to automatically select the optimal number of microstates and extract corresponding microstate features, and then perform brain microstate analysis based on the k-means clustering algorithm.

[0017] According to the implementation mode of the present application, although there is some loss of accuracy, the calculation speed of the improved k-means clustering algorithm based on the present application has been greatly improved, and it can be applied to sample data volumes of tens of millions and applied to large-scale image / video clustering.

[0018] In a possible implementation of the first aspect above, the sensory stimulation includes visual stimulation and / or auditory stimulation.

[0019] In a possible implementation of the first aspect above, the game task includes one or more of the following: a dynamic target tracking task, a color and shape recognition task, a similar target object finding task, and a multi-target monitoring task.

[0020] In the second aspect, the present application provides an attention training device based on dynamic sensory stimulation, including: an interface presentation unit, used to generate and present a game interaction interface to a user, the game interaction interface including at least one game task for training the user's attention level, the game task being presented through at least one sensory stimulation, and at least one change dimension of the at least one sensory stimulation being composed of at least one variable with multiple values; a processing unit, used to obtain performance information of the user when performing the at least one game task, the performance information being used to indicate the user's concentration level under different values of at least one of the variables, and determining the user's attention level based on the performance information.

[0021] In a third aspect, the present application provides a head-mounted wearable device, which includes: a memory for storing instructions executed by one or more processors of the device, and a processor, which is one of the processors of the device, for executing the attention training method disclosed in any aspect of the first aspect above.

[0022] In a fourth aspect, a computer-readable storage medium is provided, wherein the storage medium stores instructions. When the instructions are executed on a computer, the computer executes the steps of the attention training method described in the first aspect.

[0023] In a fifth aspect, a computer program product comprising instructions is provided. When the instructions are executed on a computer, the computer executes the steps of the attention training method of the second aspect. Alternatively, a computer program is provided. When the computer program is executed on a computer, the computer executes the steps of the attention training method of the second aspect.

[0024] The possible implementation methods and technical effects obtained in the above-mentioned second to fifth aspects are similar to the corresponding technical means and technical effects obtained in the first aspect, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a schematic diagram of an attention training system provided in an embodiment of the present application;

[0026] Figure 2 is a schematic diagram of a neural mechanism provided in an embodiment of the present application;

[0027] Figure 3 This is a flow chart of an attention training method provided in an embodiment of the present application;

[0028] Figure 4 This is a flow chart of another attention training method provided in an embodiment of the present application;

[0029] Figure 5 This is a schematic diagram of the process of preprocessing an electroencephalogram (EEG) signal provided in an embodiment of the present application;

[0030] Figure 6 is a schematic diagram of a brain microstate analysis method based on an improved k-means clustering algorithm provided in an embodiment of the present application;

[0031] Figure 7 is a schematic diagram of an attention level analysis result provided by an embodiment of the present application;

[0032] Figure 8 is a schematic diagram of another attention level analysis result provided by an embodiment of the present application;

[0033] Figure 9 is a schematic diagram of another attention level analysis result provided by an embodiment of the present application;

[0034] Figure 10 It is a structural schematic diagram of a device provided in an embodiment of the present application;

[0035] Figure 11 It is a structural diagram of a system on chip (SoC) provided in an embodiment of the present application. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0037] First, the technical terms that may be used in this article are explained.

[0038] 1. Attention training:

[0039] Attention training refers to the process of helping individuals (such as children, ADHD patients, or people with attention deficits) improve their concentration, inhibit distractions, and enhance cognitive control through a systematic approach. Studies have shown that attention training can enhance the functional connectivity of the prefrontal cortex and parietal cortex, which are brain regions closely related to executive function and attention control. In addition, repeated attention training (such as working memory training and concentration tasks) can alleviate ADHD symptoms by enhancing neuroplasticity and improving the functional connectivity of related brain areas.

[0040] Currently, there are a variety of attention training methods, including behavioral intervention, cognitive training, and neurofeedback training. For example, behavioral intervention can include behavioral correction training to shape concentration through positive reinforcement (rewards) and negative reinforcement (reducing undesirable behavior); cognitive training can include the Stroop test (such as asking children to ignore the meaning of words and report the name of a color) and continuous performance tests (such as continuously monitoring reactions when a target appears); and neurofeedback training can regulate attention by observing children's brain waves in real time.

[0041] refer to Figure 1 , Figure 1 The basic mechanism of neuromodulation is shown. Figure 1 As shown, stimuli can be converted into nerve impulses by receptors, which are then transmitted to afferent nerves, nerve centers, transmission nerves, and effectors, resulting in a reaction and reflex. Simple and complex reflexes can be incorporated into attention training design in different ways. For example, using sudden visual stimulation to trigger the user's focused gaze response (simple reflex) and using virtual rewards after completing a task to trigger the user's focused state (complex reflex).

[0042] 2. Sensory stimulation:

[0043] Sensory stimulation refers to the process by which signals from the external environment or internal body parts are transmitted to the nervous system through sensory organs (such as the eyes, ears, and skin), triggering brain perception. Sensory stimulation is fundamental to human interaction with the outside world, directly influencing attention, emotions, behavior, and cognitive development.

[0044] Depending on the senses through which people receive information, sensory stimulation can include visual stimulation, auditory stimulation, tactile stimulation, etc. For example, visual stimulation can include dynamic images with dynamic light, color, shape, motion, or flashing light, auditory stimulation can include sounds or prompts with specific frequencies, rhythms, and volumes, and tactile stimulation can include tactile feedback of different pressures, temperatures, and textures.

[0045] Furthermore, each sensory stimulus can include multiple dimensions of variation, where a dimension of variation refers to a collection of variables that can independently regulate and influence the sensory stimulus experience. In other words, each dimension of variation is composed of specific, quantifiable variables that can be adjusted to modulate the sensory stimulus.

[0046] For example, Table 1 shows examples of varying dimensions, variables, and value ranges of several sensory stimuli.

[0047] Table 1

[0048]

[0049] It should be understood that Table 1 is only an example. In the embodiments of the present application, other sensory stimulations such as olfactory stimulation, visual stimulation, auditory stimulation, etc. may also be included, and other changing dimensions and parameters may also be included, which will not be elaborated in this article.

[0050] 3.EEG signal:

[0051] EEG is a non-invasive detection technology that records the electrical activity of brain neurons through scalp electrodes. The EEG signal reflects the macroscopic sum of the synchronized postsynaptic potentials of millions of neurons in the cerebral cortex. The signal amplitude is usually 10-100 microvolts (μV) and the frequency range is 0.5-100Hz. Among them, the bands and corresponding frequencies of EEG signals include delta waves (0.5-4Hz), theta waves (4-8Hz), alpha waves (8-13Hz), beta waves (13-30Hz), gamma waves (>30Hz), etc. EEG signals have a millisecond time resolution and can directly reflect neural electrical activity, but they are easily interfered with by noise such as electromyographic signals or eye movement artifacts.

[0052] The following combination Figure 2 A scenario in which an embodiment of the present application is applied is introduced. Figure 2 Schematic diagram of an attention training system 200 provided in an embodiment of the present application is shown. Figure 2As shown, the attention training system 200 includes an attention training device 210 , an EEG acquisition device 220 and a data recording and analysis device 230 .

[0053] In an embodiment of the present application, the attention training device 210 can guide the user (such as a child or an ADHD patient) to enter a state of concentration through the visual stimulation (such as dynamic images, flashing light spots) or auditory stimulation (such as sounds of a specific frequency, prompt sounds) described above. The training content of the attention training device 210 may include meditation guidance or reaction tests, etc., and may also include game tasks in the embodiments of the present application. Optionally, the attention training device 210 can be implemented by hardware devices such as interactive display screens and multi-sensory feedback devices. The multi-sensory feedback device may include an eye tracker for visual feedback, headphones for auditory feedback, a vibrator for tactile feedback, etc.

[0054] The EEG acquisition device 220 can be a portable EEG acquisition device such as a wireless EEG headset. Figure 2 As shown, when the user participates in attention training by looking at visual stimulation or listening to auditory stimulation, the EEG acquisition device 220 can continuously monitor and acquire the user's EEG signals non-invasively through an electrode cap or a head-mounted device, and transmit the acquired raw EEG signals to the data recording and analysis device 230.

[0055] The data recording and analysis device 230 can be responsible for receiving and storing data, i.e., EEG signals, sent from the EEG acquisition device 220. The data recording and analysis device 230 can also process the user's EEG signals in real time through algorithms. For example, it can extract features related to attention (such as alpha waves, beta waves, etc. of EEG signals) to evaluate the user's degree of concentration. In addition, based on the evaluated user attention analysis results, the data recording and analysis device 230 can also generate a training report or provide feedback to the attention training device 210 to adjust the content of subsequent training tasks. For example, when the data recording and analysis device 230 detects that the user's attention is distracted, it may add visual cues in the attention training device 210 or adjust the training tasks to help the user refocus.

[0056] The above-mentioned devices can be configured from different hardware devices. For example, the attention training device 210 and the data recording and analysis device 230 can be configured through the interactive display of a personal computer and electrically connected to the EEG acquisition device 220. The above-mentioned devices can also be integrated into the same device. For example, a virtual reality / augmented reality (VR / AR) device (such as a head-mounted VR / AR helmet or glasses) includes a head-mounted EEG acquisition device 220 and a screen or optical component that serves as the attention training device 210 and / or the data recording and analysis device 230.

[0057] The following introduces the technical problems existing in the existing technology and the Figure 3 Introduce the technical solution of this application.

[0058] Figure 2 The attention training system 200 shown can be applied to children's attention training or ADHD rehabilitation training. However, the current attention training methods have problems such as single training form, lack of interest and poor adaptability.

[0059] For example, the attention training system 200 may adopt a highly repetitive training task design, such as basic cognitive exercises such as monotonous click targets, number sorting, and color matching. Although these repetitive training tasks can temporarily activate attention-related brain areas, due to their fixed form, lack of change, and lack of fun, users are prone to psychological fatigue and boredom. For example, although the traditional "Stroop Test" or "continuous performance testing" (CPT) can monitor attention, long-term repeated use will lead to a decrease in the participation of child users, and child users are also prone to resistance to boring training tasks.

[0060] In order to solve the above technical problems, the present application proposes a method 300 for attention training based on dynamic sensory stimulation, which includes steps 310 to 330. The method 300 can be applied to Figure 2 The attention training system 200 shown improves the diversity and fun of the attention training process by combining the training tasks of attention training with a gamification design with dynamic sensory stimulation changes.

[0061] Step 310: Generate and present a game interaction interface to the user.

[0062] Specifically, the game interaction interface includes at least one game task for training the user's attention level, and the game interaction interface can be Figure 2The attention training device 210 is shown, for example, presented via an interactive display screen. The game task is presented via at least one sensory stimulus. As shown in Table 1, at least one variable dimension of the at least one sensory stimulus is composed of at least one variable having multiple values, meaning that the at least one sensory stimulus is dynamically variable. Users may include children with attention deficit disorder, adult users, or individuals with ADHD.

[0063] Exemplarily, the game task may include a gamified dynamic target tracking task, such as tracking an animal that is moving smoothly in a grassland. The dynamic target tracking task can be presented by visual stimulation, and the changing dimension of the visual stimulation includes motion. Referring to Table 1, motion can be composed of two variables: the speed and direction of the animal's smooth motion. In addition, the game task may also include other tasks presented by visual stimulation or auditory stimulation, such as a gamified color and form recognition task (presented by visual stimulation and the changing dimensions include color and form), and a sound positioning task (presented by auditory stimulation and the changing dimensions include loudness and spatial positioning).

[0064] As another example, the game task can include a gamified target detection task, such as identifying animals in a pond scene by the direction of the sound source and the trajectory of their movement. This target detection task can be presented through visual and auditory stimulation. The variation dimensions and variables involved in visual and auditory stimulation can be shown in Table 1. In other words, the game task can implement cross-sensory stimulation such as audio-visual interaction or visual-tactile binding, and the parameters that constitute the variation dimension can also be a combination of variables, such as the combination of sound source direction and movement direction.

[0065] The game tasks introduced above are only examples, and the following embodiments will introduce the game scenes of these game tasks in detail, so they will not be described in detail here.

[0066] Optionally, the game tasks in the game interaction screen can be game tasks related to user information or points of interest. In other words, embodiments of the present application can analyze user information or points of interest, such as using association rule algorithms based on the user's registered user information and the user's selected points of interest to recommend appropriate and interesting game tasks to the user. The following embodiments will introduce this part in detail.

[0067] Optionally, the game task may also include a feedback mechanism or incentive mechanism. For example, the feedback mechanism may include providing instant visual (score changes), auditory (prompt sound effects) and tactile (vibration feedback) multi-channel feedback, thereby strengthening the correct response. For another example, in the incentive mechanism, gamification elements such as point rewards and achievement badges can be set to enhance the continuity of training. In addition, virtual gold coin rewards can also be set, which can be used to purchase clothing, food, toys, etc. for the virtual pets in the game tasks, thereby enhancing the training enthusiasm of the patient children.

[0068] Step 320: Obtain performance information of the user when performing the at least one game task.

[0069] Specifically, the performance information is used to indicate the user's concentration level under different values of at least one variable. The different values of the at least one variable can refer to Figure 1 As shown, that is, Figure 1 At least one variable in the game can take different values within a range of values to achieve dynamic changes in sensory stimulation. In this way, users can get a better sense of environmental realism and continuous engagement based on dynamic sensory stimulation, avoiding sensory fatigue of players.

[0070] For example, in the target detection task introduced above, there are multiple combinations of variable combinations of sound source orientation and motion direction. The performance information can indicate the user's degree of concentration when the multiple combinations are taken. For example, the user's degree of concentration when the sound source orientation is right front (such as a value of 45°) and the motion direction is horizontal motion (such as a value of 90°), and the user's degree of concentration when the sound source orientation is left rear (such as a value of 225°) and the motion direction is vertical motion (such as a value of 0°).

[0071] Optionally, the performance information may include the user's real-time game performance information, that is, the user's real-time score, real-time decision-making, real-time completion rate, etc. when performing a game task. Furthermore, in an embodiment of the present application, the real-time game performance information can not only evaluate the user's performance, but can also be used to adjust the difficulty of the game task. In other words, the embodiment of the present application can adjust the difficulty of the game task based on the real-time game performance information presented by the user when performing the game task.

[0072] Exemplarily, in an embodiment of the present application, the real-time performance information of the game may include an ability level and a load level. Among them, the ability level can be used to indicate the user's degree of completion of the game task, such as instant scores, achievement progress, etc.; the load level can be used to indicate the user's stress tolerance for the game task, such as cognitive load (task completion time, error rate, operation pause frequency, number of repeated attempts, the ratio between the θ wave and the β wave of the EEG signal), physiological load (heart rate, pupil diameter, blink frequency, body temperature, respiratory rate, etc.). Furthermore, when the user's ability level is low (such as below a certain threshold) and / or the load level is high (such as above a certain threshold), the difficulty of the game task can be reduced, and vice versa, the difficulty of the game task can be increased. The following embodiments will introduce this process in detail, and this article will not go into details here.

[0073] Optionally, the performance information may include the user's EEG signal and brain topography. Among them, the EEG signal is as described above, and the brain topography is a method of presenting brain activity or functional distribution in the form of a two-dimensional / three-dimensional "map" through visualization technology (such as EEG, functional magnetic resonance imaging (fMRI), etc.), which can intuitively display the activation intensity, frequency characteristics or connection status of different brain regions. The EEG signal and brain topography can be used to subsequently evaluate the user's attention level, which will be described in detail in the following embodiments.

[0074] Step 330: Determine the user's attention level based on the performance information.

[0075] Optionally, as mentioned above, the performance information may include the user's EEG signal and brain topography, and then brain microstate analysis and / or spectrum analysis may be performed based on the user's EEG signal, as well as visual analysis of the brain topography to obtain analysis results.

[0076] Among them, brain microstate analysis can be used to indicate that the brain functional network maintains a stable spatial activation pattern within tens to hundreds of milliseconds and reflects the rapidly switching cognitive processing process. The abnormal duration or switching frequency it exhibits can indicate the user's attention fluctuations, which are usually divided into four or more microstates (such as microstate categories A to D, the positive and negative voltage centers of category A are located in the right frontal lobe and left occipital lobe, the positive and negative voltage centers of category B are located in the left frontal lobe and right occipital lobe, the positive and negative voltage centers of category C are located in the anterior frontal lobe and occipital lobe, and the positive and negative voltage centers of category D are located in the central frontal area and occipital lobe); spectral analysis can be used to reflect the rhythmic activity indication of neurovascular coupling based on low-frequency oscillations (0.01-0.1Hz) of EEG signals, and its specific frequency band power abnormalities can indicate the user's attention fluctuations; brain topography visualization analysis can be used to display the activation or connection patterns of different brain regions.

[0077] Seeking to extract more features from the multiple dimensions of microstates, spectra, and brain topography can be said to be more comprehensive and the reliability of analytical argumentation is higher.

[0078] Optionally, the EEG signal of the user included in the performance information may include a noise signal or an interference signal, and then, the embodiment of the present application can denoise and interfere with the original EEG signal. For example, the first EEG signal can be an EEG signal used for brain microstate analysis and / or spectrum analysis, and the second EEG signal is the original EEG signal collected by the EEG acquisition device 220. The data recording and analysis device 230 can remove interference signals such as non-neurogenic artifact signals in the second EEG signal through a short separation regression (SSR) method and an independent component analysis (ICA) method to obtain the first EEG signal. The following embodiments will introduce this process in detail, and this article will not go into details here.

[0079] Optionally, in some embodiments of the present application, brain microstate analysis can be implemented using a clustering algorithm such as a k-means clustering algorithm. For example, the present application can use an improved k-means clustering algorithm to pre-process the EEG signal using the Canopy algorithm, intelligently partition the original large EEG signal data set, and then perform brain microstate analysis using the k-means clustering algorithm. This process will be described in detail below and will not be elaborated on here.

[0080] Furthermore, the results of the aforementioned brain microstate analysis, spectral analysis, and brain topography visualization analysis can be used to determine the user's attention level. For example, this attention level can be assessed using standardized clinical scales, computerized attention tests, social adaptability, and other aspects. The information obtained from this assessment can be used for personalized adaptation, such as customizing a personalized training program.

[0081] For example, standardized clinical scales may include one or more of the ADHD Rating Scale, the Conners Scale, and the SNAP-IV Scale. The ADHD Rating Scale is used to assess the severity of inattention and hyperactivity / impulsivity symptoms, the Conners Scale is used to assess behavioral problems and attention performance, and the SNAP-IV Scale is used for screening ADHD symptoms in children.

[0082] For example, computerized attention tests can include the Continuous Performance Test (CPT) or the Stroop Task. CPTs can include the Test of Variables of Attention (TOVA) or the Integrated Visual and Auditory CPT-2 (IVA-2), which are used to quantify attention, response inhibition, and impulse control. The Stroop Task can be used to assess cognitive flexibility and inhibitory control.

[0083] For example, social adaptability can be assessed using the Weiss Functional Impairment Rating Scale (WFIRS), the Child Health Questionnaire (CHQ), or the Adult ADHD Quality of Life Scale (AAQoL). The WFIRS can be used to assess the impact of ADHD on family, school, social, and occupational functioning, while the CHQ and AAQoL can be used to measure quality of life.

[0084] Method 300 combines attention training with user interests by designing gamified scenarios, resulting in greater interactivity and richer content, thereby improving users' attention levels. Furthermore, by recommending appropriate game tasks, dynamically adjusting task difficulty, and adding feedback mechanisms, users' attention regulation abilities are enhanced.

[0085] The following combination Figures 4 to 8 An example of an embodiment corresponding to method 300 is introduced.

[0086] First, introduce the Figure 2The design of each device involved in the attention training system 200 is shown. In the embodiment of the present application, the system architecture of the attention training system 200 can adopt a hybrid mode of browser / server (B / S) architecture and client / server (C / S).

[0087] The attention training device 210 serves as the user end, and Table 2 shows an example of a user-end implementation technology. The data recording and analysis device 230 serves as the server end, and Table 3 shows an example of a server-end implementation technology. Furthermore, the system architecture may also include a database for storing user data, an algorithm engine for EEG data analysis, and an external application programming interface (API) required for data collection.

[0088] Table 2

[0089]

[0090] Table 3

[0091]

[0092] The EEG acquisition device 220 can use a combination of an EEG signal amplifier and an EEG recorder to collect EEG signals. Among them, 64 channels can be used, positioned according to the international 10-20 lead standard, with an analog / digital (A / D) sampling frequency of 256Hz to 1KHz, and the reference electrodes are the two earlobes. In addition, two acquisition channels can be used to synchronously record the electrooculogram (EOG) signals during the experiment to remove EOG noise. In order to avoid the problem of poor contact between the EEG acquisition device and the scalp, resulting in excessive measurement resistance and poor data collection effect, a pinhole can be used to apply conductive paste to the channel hole of the EEG acquisition device.

[0093] Experiments for testing the attention training system 200 can be conducted in a shielded room. For example, the experimental design may include: a pre-test phase, assessing baseline performance using a standard attention test; a training phase, 3-5 times per week, 20-30 minutes per session, for 8 weeks; a post-test phase, repeating the baseline test and collecting EEG data; and a follow-up phase, conducting follow-up assessments one and three months after the training.

[0094] Secondly, combined Figure 4 Introducing Method 400, Figure 4 The schematic flow of method 400 is shown, including steps 410 to 440. Method 400 corresponds to method 300 and is an example of an embodiment of method 300.

[0095] Step 410: Analyze the user information and points of interest when the user registers, and use the user's association rule algorithm to generate and present a suitable attention training game.

[0096] A user (e.g., a child or someone with ADHD) can log in to the user registration system displayed on the attention training device 210, enter the login interface, and complete identity verification. The system can then analyze the user's registration information or interests and, using an association rule algorithm, determine appropriate game tasks for the user's attention training game.

[0097] Optionally, methods such as label encoding and interest weighting can be incorporated into the association rule algorithm to improve recommendation accuracy. For example, the system can first perform one-hot encoding on the collected user information, constructing binary feature vectors to label different user attributes (such as age group, device type, and gaming duration). Then, leveraging the item-by-item search feature of the Apriori algorithm, interest weighting (e.g., the user's self-assessment of importance during registration) is introduced during the support calculation phase, along with a time decay factor. This exponentially weights the support calculation to assign a higher dynamic weight coefficient to recently participated game tasks. For example, if a user has not opened or selected a game within seven days, the weight decays. During the algorithm's iterations, candidate frequent itemsets (i.e., frequently occurring tasks and tags) are generated based on the weighted values. A layer-by-layer search and pruning strategy is employed to optimize computational efficiency. Furthermore, weighted confidence scores are calculated, and filtering criteria for lift and confidence scores (e.g., selecting game tasks with lift > 1.5 and confidence > 0.6) are added to recommend game tasks to users.

[0098] Weighted confidence refers to the percentage of users who previously played game tasks A and B. If the percentage is too small, it doesn't necessarily indicate that the user is interested in game task A. A high percentage, however, indicates that game task A is indeed a high-interest game for the user, and therefore, it will be recommended. Weighted confidence also factors in time and user personality.

[0099] Corresponding to step 310, after the game tasks of the attention training game are determined for the user, the attention training device 210 can generate and present a game interaction interface including the game tasks to the user, allowing the user to perform attention rehabilitation training based on dynamic sensory stimulation. Moreover, each of these game tasks can be combined with the game style, with different novel game backgrounds and dynamically changing sensory stimulation to attract the user's interest, and set up a reasonable feedback mechanism (such as feedback from sensory channels or setting up a growing virtual pet for users to observe and imitate), and an incentive system (such as point rewards, achievement badges, and virtual gold coins to purchase supplies for virtual pets), thereby reducing the boredom of traditional training systems. Optionally, the virtual pet can be selected when the user logs in for the first time, thereby preparing for subsequent training feedback.

[0100] The following are some examples of game tasks, including dynamic target tracking tasks, multi-target detection tasks, and similar target search tasks.

[0101] For example, the game task may include a dynamic target tracking task, which may be presented through visual stimulation, and the variation dimensions may include motion (such as different movement speeds and directions) and color (such as different brightness and saturation in different weather conditions), etc. The variables constituting the variation dimensions may refer to Table 1. In the embodiment of the present application, a prairie scene theme may be designed for the dynamic target tracking task, wherein the prairie scene may include animals (such as a cheetah, a target animal to be tracked, an elephant as a background animal, and a tiger as an interference animal), environmental interactions (such as tall grass swaying in the wind that can obscure the animal's body and a flock of birds that can distract attention), and weather changes (such as sunny days and dusk with different background contrasts).

[0102] For example, in dynamic target tracking tasks, animals in the grasslands, such as cheetahs, elephants, and tigers, will move in random paths and at gradually changing speeds. After receiving the task of tracking the target animal, the user needs to quickly focus their attention and continue to track the target animal through vision. Environmental interactions and weather changes can interfere with the user's visual tracking. When the target animal disappears, the user can provide feedback by pressing a button or touching the screen. In addition, when the user successfully and continuously tracks the target animal, the user can be provided with instant visual (such as score changes), auditory (such as prompt sound effects), and tactile (vibration feedback) multi-channel feedback to reinforce the correct response. Gamification elements such as point rewards, achievement badges, and virtual gold coins (which can be used to purchase clothing, food, toys, etc. for virtual pets) can also be set.

[0103] As another example, the game task may include a multi-target detection task, which may be presented through visual stimulation and auditory stimulation or through auditory stimulation alone. The variation dimension may include motion (such as different motion directions) and / or sound source orientation (such as different orientations), etc. The variables constituting the variation dimension may refer to Table 1. In the embodiment of the present application, a pond scene theme may be designed for the multi-target detection task, wherein the pond scene may include animals (such as birds, frogs, etc., which are target animals to be tracked) and environmental interactions (such as light and water ripples or the sound of raindrops and wind).

[0104] For example, in a multi-target detection task, multiple dynamic animals randomly appear in the scene, accompanied by the sounds of these animals. Environmental interactions such as light and wind can cause interference. Afterwards, all animals are cleared from the screen and the user is asked what animals are in each direction of the scene. The user can click on the moving animals captured visually and the animals captured auditorily on the screen. When the user successfully and continuously detects the target animal, the user can be provided with instant visual (such as score changes), auditory (such as prompt sound effects) and tactile (vibration feedback) multi-channel feedback to reinforce correct responses. Gamification elements such as point rewards, achievement badges, virtual gold coins, and virtual pets can also be set.

[0105] As another example, game tasks can include similar target object search tasks and color and form recognition tasks. These tasks can be presented through visual stimulation, and the dimensions of variation can include color, shape, position, etc. For example, in the similar target object search task, a large number of repeated distractors and a target item that is very similar to the distractors may appear on the screen. The subject is required to find the target item within a certain period of time, and the user can click to provide feedback. For another example, in the color and form recognition task, there is a special cannon at the bottom center of the screen that continuously fires geometric bullets. The target geometric shape is initially given, and when the target geometric shape is fired, the user is required to provide corresponding key feedback.

[0106] It should be understood that the above-mentioned game tasks such as the dynamic target tracking task are only examples, and the game tasks of the embodiments of the present application are not limited to this.

[0107] Step 420: While performing attention training, the game difficulty is adaptively adjusted according to the user's real-time ability level and load level.

[0108] Specifically, corresponding to step 320, the attention training system 200 can obtain the user's real-time game performance information. Furthermore, while the user is performing attention training, the attention training system 200 can adaptively adjust the game difficulty based on the ability level and workload level in the real-time game performance information.

[0109] The ability level may include the user's instant score, achievement progress, etc., and the load level may include cognitive load, physiological load, etc. For example, in a dynamic target tracking task, when the user's instant score remains unchanged or changes slowly and / or the load level is high, the speed of the target animal may be slowed down; when the score increases rapidly and / or the load level is high, the speed of the target animal may be accelerated.

[0110] It is worth noting that the embodiments of the present application can also adjust the game difficulty by considering other dimensions, such as the complexity of the mission scene. The above is only an example and the present application does not limit this.

[0111] For example, taking the game tasks mentioned above as an example, the game difficulty involved in the dynamic target tracking task can be related to the number of moving animals / interference animals, the movement speed of the target animal (such as slow, fast, variable speed, etc.), the proportion of tall grass occlusion (partial occlusion or multiple complete occlusions), and the task duration (2 minutes / round or 10 minutes / round). The game difficulty involved in the multi-target detection task can be related to the number of targets, the target presentation time, the target characteristics (single color, or the color and shape will constantly change, such as some animals have patterns or spots), and whether there are interference items (fallen leaves, water droplets or wind sounds).

[0112] Optionally, in an embodiment of the present application, the difficulty of the game task can be adjusted by a dynamic threshold algorithm and a cognitive balance algorithm.

[0113] Specifically, in the dynamic threshold algorithm, the initial threshold can be set to 0, that is, the initial difficulty is simple, and then the threshold can be dynamically adjusted, that is, the difficulty can be adjusted. For example, in short-term adaptation, the difficulty can be adjusted based on a sliding window of recent performance (such as the average of the past 5 tasks). If the user has correct feedback on the task for 3 consecutive times, the difficulty will be increased, such as increasing the threshold by 10%. In long-term adaptation, the threshold standard can be gradually increased in combination with the user's ability growth curve (such as the learning effect), and then the game difficulty can be set to avoid tasks that are too simple and make the subjects lose interest.

[0114] In the cognitive load balancing algorithm, by quantifying the user's current cognitive resource usage, the task difficulty is dynamically balanced with the user's load level and tolerance to optimize learning or operation efficiency, with the goal of avoiding "overload" (excessive pressure) or "underload" (boredom and distraction). The load level can include the task completion time, error rate, number of repeated attempts, the θ / β wave ratio of the EEG signal (degree of cognitive effort), etc. introduced above. In an embodiment of the present application, a weighted formula can be used to measure the load level by integrating multi-source data, and the value range can be set to be divided into low (0-0.3) load, medium (0.3-0.7) load, and high (0.7-1) load. If the load score is greater than 0.7, the task complexity is reduced (such as reducing the amount of information presented at the same time) or a rest prompt is inserted. If the load score is less than 0.3, the branch task or time pressure is increased. The weighted formula is shown in the following formula (1):

[0115]

[0116] Among them, Load score is the load level score, ErrorRate is the error rate, MaxError is the preset maximum error rate, EEG_θ / β is the θ / β wave ratio of the EEG signal, Baseline is the preset baseline ratio of the θ / β wave ratio, and a1 is the set parameter ratio.

[0117] The Dynamic Threshold Algorithm and the Cognitive Load Balance Algorithm are combined. For example, by setting different evaluation thresholds, the Dynamic Threshold Algorithm adjusts the game difficulty only when the accuracy rate within five feedbacks is 80% or above and the Cognitive Load Balance Algorithm calculation is above 0.6. This provides a more rational way to adjust the game difficulty, avoiding short-term fluctuations (such as accidental mistakes) caused by a single algorithm relying solely on behavioral data, as well as premature difficulty increases (such as when the player's cognitive load is not ready).

[0118] Step 430: Display the data collection status in real time, and perform pre-processing and data analysis on the data, including spectrum analysis, brain topography, and brain microstate analysis.

[0119] Optionally, corresponding to step 330, in an embodiment of the present application, the collected EEG signal data may be preprocessed using SSR and ICA methods, i.e., noise and interference signals in the EEG signal may be removed. For example, the EEG signal may include non-neural artifact signals, which may include scalp blood flow artifact signals and physiological artifact signals such as heartbeat and respiration.

[0120] Figure 5 A schematic flow chart of a data preprocessing method 500 provided in an embodiment of the present application is shown, wherein the method 500 includes steps 510 to 530.

[0121] Step 510: Acquire a second EEG signal, where the second EEG signal includes the first EEG signal and an interference signal, where the interference signal includes a non-neurogenic artifact signal, and the non-neurogenic artifact signal includes a scalp blood flow artifact signal and a physiological artifact signal.

[0122] Step 520: Remove the scalp blood flow artifact signal from the second EEG signal by using a short-range regression (SSR) method, and remove the physiological artifact signal from the second EEG signal by using an independent component analysis (ICA) method to obtain the first EEG signal.

[0123] Step 530: Perform brain microstate analysis and / or spectrum analysis on the user's first EEG signal, and perform visual analysis on the brain topography map to obtain analysis results.

[0124] As shown in step 520, in order to improve the quality of EEG signals and remove non-neural artifact signals, the SSR method is used to process the scalp blood flow artifact signal, and the ICA method is combined to remove physiological artifact signals such as heartbeat and breathing.

[0125] Specifically, the SSR method estimates and removes scalp blood flow artifact signals by constructing a long-short channel regression model. For example, a linear regression model can be used to remove short channel crosstalk. The calculation formula is as follows (2):

[0126]

[0127] Among them, Y long (t) is the second EEG signal in step 510, which includes the time series changes of the long channel recording EEG signal, Y short (t) is the spatial feature of the EEG signal recorded by the short channel. β is the regression coefficient, which represents the contribution of the short channel signal to the long channel signal. β can then be calculated by least squares regression, as shown in Equation (3):

[0128]

[0129] After SSR processing, That is, the time sequence change of the first EEG signal in step 520 mainly retains the change of the second EEG signal itself, reducing the influence of the artifact signal caused by brain movement.

[0130] The ICA method is a blind source separation method that can be used to decompose the independent components in the second EEG signal and identify and remove periodic physiological artifacts such as heartbeat and breathing. The signal matrix monitored by the EEG acquisition device is shown in the following formula (4):

[0131] X(t)=[x1(t),x2(t),…,x N (t)] T (4)

[0132] Among them, x i (t) represents the second EEG signal recording of the i-th channel, and N is the number of channels. Then, ICA decomposition is performed, assuming that the second EEG signal is generated by a linear mixture of multiple independent components S(t), as shown in the following equation (5):

[0133] X(t)=A*S(t) (5)

[0134] where S(t)=[s1(t),s2(t),...,s N (t)] T are independent source signals (including neural activity signals and physiological artifact signals), and A is the mixing matrix. The ICA method seeks a demixing matrix W such that S(t) = W*X(t), where W is estimated by maximizing non-gaussianity or minimizing mutual information.

[0135] Finally, we remove physiological artifacts. In the independent component S(t), high-frequency components (>0.5Hz) usually correspond to heartbeat (around 1Hz) and breathing (around 0.2-0.4Hz) artifacts, while low-frequency components (<0.01Hz) may come from blood pressure fluctuations and motion artifacts. Through power spectrum analysis and spatial distribution analysis, we identify and remove non-neurogenic components S artifact (t), and then reconstruct the EEG signal as shown in the following equation (6):

[0136]

[0137] Among them S clean (t) is the independent component after removing artifacts, This is the signal matrix of the first EEG signal. Finally, by combining SSR and ICA, the neural source correlation of the EEG signal is effectively improved, the interference of non-neural artifacts is reduced, and high-quality input data is provided for subsequent system analysis.

[0138] For brain microstate analysis, spectrum analysis, and visualization analysis of brain topography, please refer to the relevant introduction in step 330. Optionally, in an embodiment of the present application, a k-means-based clustering algorithm can be selected to perform brain microstate analysis, thereby resolving the problem of difficulty in analysis caused by the large size of the EEG dataset.

[0139] For example, the k-means-based clustering algorithm uses global map dissimilarity (GMD) as a metric for microstate classification. Based on the GMD metric, the k-means clustering algorithm can cluster maps with low GMD or high correlation into one category, and can obtain the number of categories that can optimally explain most of the data variance. For example, based on the calculated spatial correlation coefficient or GMD between the identified microstate map and the map at each time point, the microstate category of the map at each time point is determined.

[0140] Optionally, an embodiment of the present application proposes a brain microstate analysis method based on an improved k-means clustering algorithm, which can be used to automatically select the optimal number of microstates and extract corresponding microstate features by introducing the Canopy algorithm. Figure 6 A schematic diagram of the brain microstate analysis method is shown.

[0141] For example, Figure 6 As shown in the figure, during microstate analysis, the topography of all global field power (GFP) peaks is first extracted and incorporated into the Canopy and k-means clustering algorithms. The individual microstates are then classified to determine the microstate category corresponding to the EEG signal. GFP represents the instantaneous electric field strength of the brain and is therefore commonly used to measure whole-brain responses to events or to describe rapid changes in brain activity. In microstate analysis, the topography at the local maximum of the GFP curve (the time when the field strength is strongest and the signal-to-noise ratio of the topography is highest) represents the discrete state of the EEG signal.

[0142] For example, Figure 6 As shown, this method first analyzes the user's EEG signal data, calculates GFP, and selects GFP peaks. The Canopy clustering algorithm then uses its density-based clustering properties to intelligently partition the data space. A global explained variance weighted model is then established to systematically optimize the location of coarse cluster centers. Notably, this model possesses the core advantage of autonomously determining dual-threshold parameters, significantly improving the resolution of microstate temporal features while eliminating human intervention.

[0143] Specifically, the Canopy optimization framework introduces a dynamic parameter adjustment mechanism based on sample spacing statistics. The arithmetic mean of the distances between all sample points in the data set is defined as the key threshold parameter. This threshold parameter participates in both density calculation and neighborhood partitioning. Its mathematical expression is shown in the following formula (7):

[0144]

[0145] Then, the sample density ρ(i) is calculated using the threshold parameter. ρ(i) represents the number of samples around sample i.

[0146] The following formula (8) is the formula for calculating density, where d ij Represents the distance between samples i and j.

[0147]

[0148] Next, the local average distance a(i) of each sample and the distance s(i) to the nearest and denser sample are calculated and the weight w is calculated based on this, as shown in Equation (9):

[0149]

[0150] Finally, the sample points with the highest weights are selected as cluster centers in turn until there are no remaining samples in the data set. The coarse clustering results obtained by the Canopy clustering algorithm are used as the input data before initialization of the K-means clustering algorithm. This application does not elaborate on the specific process of the K-means algorithm.

[0151] Then, if Figure 6 As shown in Figure 2, after obtaining microstate clusters, the microstates can be mapped onto the original EEG signal to generate a microstate time series, a process known as remapping. The average duration and frequency of microstate features are then analyzed to classify attention deficit-related disorders.

[0152] Optionally, in the improved k-means clustering algorithm described above, locality-sensitive hashing (LSH) can be used to quickly approximate similarity calculations to accelerate the Canopy stage, and Elkan's algorithm (e.g., using triangle inequalities to reduce distance calculations) can be applied to accelerate the K-means stage. Furthermore, during the Canopy stage, data can be randomly sampled (e.g., retaining 70% of the data) to achieve data reduction, and GPUs can be used to parallelize the calculation of Canopy centers and k-means iterations.

[0153] Finally, the improved k-means clustering algorithm based on this application, although there is some loss of accuracy, the calculation speed is greatly improved, and it can be applied to tens of millions of sample data and large-scale image / video clustering.

[0154] Step 440: Obtain data analysis results, and determine the user's attention level based on the data analysis results.

[0155] Figures 7 to 9 An example of the data analysis results in the embodiment of the present application is shown. Figure 7 An example of brain microstate analysis results is shown, including the difference in transition probability (p-value) between a healthy control group and users with low concentration. Figure 8 shows an example of spectrum analysis, Figure 9 An example of brain topography analysis is shown.

[0156] exist Figure 7 In the table shown, Class AF represents the six states of microstates, PD (patient division) represents the patient group with attention deficits (such as ADHD), Control represents the control group of normal people, and t(p) refers to the time point of the microstate. The upper half of the table shows the duration of the six different microstates. It can be seen from the table that the duration of the six microstates in the PD group is significantly longer than that in the control group. Patients have slow reactions and difficulty switching tasks, resulting in long-lasting brain microstates. In contrast, the microstates of healthy people have shorter durations and more stable transition patterns, which meet the needs of efficient information processing. In addition, the lower half of the table shows the frequency of occurrence of the six different microstates. Due to the increased transition frequency caused by the patients' impulsivity and hyperactivity, the frequency of microstates is higher. Therefore, it can be seen from the table that the frequency of occurrence of the PD group in the experiment is higher than that of the control group.

[0157] For spectrum analysis, the Fast Fourier Transform (FFT) can convert time-series voltage signals into frequency signals, thereby analyzing the amplitude and phase of different frequency components. Furthermore, different cognitive processes can exhibit different rhythmic waves. Cognitive games that train attention stimulate the brain, causing higher alpha waves in the right prefrontal cortex, while memory-related experiments lead to higher beta wave activity in the occipital and parietal lobes at the back of the brain. Figure 8 This figure shows a spectrum analysis of a user with low concentration (such as someone with ADHD) during a cognitive experiment. The colors in the lower right corner distinguish different waves. As can be seen, the high-frequency beta and gamma waves (dark blue) have high power and account for the majority of the frequency range, while the mid-frequency alpha waves (green) are barely detectable. While alpha waves should increase during attention tasks, patients typically show no significant change in alpha waves. Therefore, based on this result, the user's attention level can be determined.

[0158] Figure 9 This is a conversion of EEG data collected from a cognitive experiment game designed for users with low concentration (such as ADHD patients). The left picture is the user's brain topography, and the right picture is the brain topography of a healthy control. Figure 9The results show that healthy controls showed higher activity in the right prefrontal cortex, consistent with the neurological theory that the prefrontal cortex controls attention. However, users with lower concentration levels showed relatively weaker activity compared to healthy controls, allowing for a quick assessment of their attention level.

[0159] Now refer to Figure 10 , shown is a block diagram of a device 1000 according to one embodiment of the present application. The device 1000 may include one or more processors 1001 coupled to a controller hub 1003. For at least one embodiment, the controller hub 1003 communicates with the processor 1001 via a multi-drop bus such as a front side bus (FSB), a point-to-point interface such as a quick path interconnect (QPI), or a similar connection 1010. The processor 1001 executes instructions that control general types of data processing operations. In one embodiment, the controller hub 1003 includes, but is not limited to, a graphics memory controller hub (GMCH) (not shown) and an input / output hub (IOH) (which may be on separate chips) (not shown), wherein the GMCH includes memory and a graphics controller and is coupled to the IOH.

[0160] The device 1000 may also include a coprocessor 1002 and a memory 1004 coupled to the controller hub 1003. Alternatively, one or both of the memory and the GMCH may be integrated within the processor, with the memory 1004 and the coprocessor 1002 directly coupled to the processor 1001 and the controller hub 1003, with the controller hub 1003 and the IOH being in a single chip. The memory 1004 may be, for example, a dynamic random access memory (DRAM), a phase change memory (PCM), or a combination of the two. In one embodiment, the coprocessor 1002 is a special-purpose processor, such as, for example, a high-throughput MIC processor (manyintegrated core, MIC), a network or communication processor, a compression engine, a graphics processor, a general purpose computing on GPU (GPGPU), or an embedded processor, etc. The optional nature of the coprocessor 1002 is indicated by a dotted line in Figure 10 middle.

[0161] The memory 1004, as a computer-readable storage medium, may include one or more tangible, non-transitory computer-readable media for storing data and / or instructions. For example, the memory 1004 may include any suitable non-volatile memory such as flash memory and / or any suitable non-volatile storage device, such as one or more hard-disk drives (HDD(s)), one or more compact disc (CD) drives, and / or one or more digital versatile disc (DVD) drives.

[0162] In one embodiment, device 1000 may further include a network interface controller (NIC) 1006. NIC 1006 may include a transceiver for providing a radio interface for device 1000, thereby communicating with any other suitable devices (e.g., a front-end module, an antenna, etc.). In various embodiments, NIC 1006 may be integrated with other components of device 1000. NIC 1006 may implement the functionality of the communication unit in the above-described embodiments.

[0163] Device 1000 may further include input / output (I / O) devices 1005. I / O 1005 may include: a user interface designed to enable a user to interact with device 1000; a peripheral component interface designed to enable peripheral components to interact with device 1000; and / or sensors designed to determine environmental conditions and / or location information related to device 1000.

[0164] It is worth noting that Figure 10 This is for illustrative purposes only. Figure 10 It is shown that the device 1000 includes multiple devices such as a processor 1001, a controller hub 1003, and a memory 1004. However, in actual applications, the device using the methods of the present application may only include a part of the devices of the device 1000, for example, it may only include the processor 1001 and the NIC 1006. Figure 10 The properties of the optional components are shown with dashed lines. According to some embodiments of the present application, the memory 1004 as a computer-readable storage medium stores instructions that, when executed on a computer, cause the device 1000 to perform the attention training method according to the above-described embodiment. For details, please refer to the method of the above-described embodiment and will not be repeated here.

[0165] Now refer to Figure 11 , which is a block diagram of a system on chip (SoC) 1100 according to an embodiment of the present application. Figure 11In FIG, similar components have the same reference numerals. In addition, the dashed boxes are optional features of more advanced SoCs. Figure 11 In the embodiment, SoC 1100 includes: an interconnect unit 1150 coupled to an application processor 1110; a system agent unit 1180; a bus controller unit 1190; an integrated memory controller unit 1140; a set of one or more coprocessors 1120, which may include integrated graphics logic, an image processor, an audio processor, and a video processor; a static random access memory (SRAM) unit 1130; and a direct memory access (DMA) unit 1160. In one embodiment, the coprocessors 1120 include specialized processors, such as, for example, a network or communication processor, a compression engine, a GPGPU, a high-throughput MIC processor, or an embedded processor.

[0166] The static random access memory (SRAM) unit 1130 may include one or more computer-readable media for storing data and / or instructions. The computer-readable storage medium may store instructions, specifically, temporary and permanent copies of the instructions. The instructions may include: when executed by at least one unit in the processor, causing the Soc 1100 to perform the attention training method according to the above embodiment. For details, please refer to the method of the above embodiment, which will not be repeated here.

[0167] The various embodiments of the mechanisms disclosed in this application can be implemented in hardware, software, firmware, or a combination of these implementation methods. The embodiments of the present application can be implemented as a computer program or program code executed on a programmable system, which includes at least one processor, a storage system (including volatile and non-volatile memory and / or storage elements), at least one input device, and at least one output device.

[0168] Program code can be applied to input instructions to perform the functions described herein and generate output information. The output information can be applied to one or more output devices in a known manner. For purposes of this application, a processing system includes any system having a processor such as, for example, a digital signal processor (DSP), a microcontroller, an application specific integrated circuit (ASIC), or a microprocessor.

[0169] Program code can be implemented with a high-level programming language or an object-oriented programming language to communicate with the processing system. Where necessary, program code can also be implemented in assembly language or machine language. In fact, the mechanism described in this application is not limited to the scope of any particular programming language. In either case, the language can be a compiled language or an interpreted language.

[0170] In some cases, the disclosed embodiments may be implemented in hardware, firmware, software, or any combination thereof. The disclosed embodiments may also be implemented as instructions carried or stored on one or more temporary or non-temporary machine-readable (e.g., computer-readable) storage media, which may be read and executed by one or more processors. For example, instructions may be distributed over a network or through other computer-readable media. Therefore, a machine-readable medium may include any mechanism for storing or transmitting information in a machine (e.g., computer) readable form, including but not limited to floppy disks, optical disks, optical discs, compact disc read-only memories (CD-ROMs), magneto-optical disks, read-only memories (ROMs), random-access memories (RAMs), erasable programmable read-only memories (EPROMs), electrically erasable programmable read-only memories (EEPROMs), magnetic or optical cards, flash memory, or a tangible machine-readable memory for transmitting information (e.g., carrier waves, infrared signals, digital signals, etc.) using the Internet in electrical, optical, acoustic, or other forms of propagation signals. Accordingly, machine-readable media includes any type of machine-readable media suitable for storing or transmitting electronic instructions or information in a form readable by a machine (eg, a computer).

[0171] In the accompanying drawings, some structural or method features may be shown in a particular arrangement and / or order. However, it should be understood that such a particular arrangement and / or order may not be required. Rather, in some embodiments, these features may be arranged in a manner and / or order different from that shown in the accompanying drawings. In addition, the inclusion of a structural or method feature in a particular figure does not imply that such a feature is required in all embodiments, and in some embodiments, such features may not be included or may be combined with other features.

[0172] It should be noted that the units / modules mentioned in the various device embodiments of the present application are all logical units / modules. Physically, a logical unit / module can be a physical unit / module, or a part of a physical unit / module, or can be implemented as a combination of multiple physical units / modules. The physical implementation of these logical units / modules themselves is not the most important. The combination of functions implemented by these logical units / modules is the key to solving the technical problems raised by this application. In addition, in order to highlight the innovative part of this application, the above-mentioned device embodiments of this application do not introduce units / modules that are not closely related to solving the technical problems raised by this application. This does not mean that other units / modules do not exist in the above-mentioned device embodiments.

[0173] It should be noted that in the examples and description of this patent, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "including a" does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0174] Although the present application has been shown and described with reference to certain preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the application.

Claims

1. A method for attention training based on dynamic sensory stimulation, characterized in that: include: Generating and presenting a game interaction interface to a user, the game interaction interface including at least one game task for training the user's attention level, the game task being presented through at least one sensory stimulation, at least one variation dimension of the at least one sensory stimulation being composed of at least one variable having multiple values; Acquiring performance information of the user when performing the at least one game task, the performance information being used to indicate the user's degree of concentration when at least one of the variables takes different values; An attention level of the user is determined based on the performance information.

2. The method according to claim 1, characterized in that The performance information includes real-time performance information of the game, and the method further includes: Adjust the difficulty of the game task according to the real-time game performance information of the user when performing the game task.

3. The method according to claim 2, characterized in that The real-time game performance information includes an ability level and a load level. The ability level is used to indicate the degree of completion of the game task by the user, and the load level is used to indicate the stress tolerance of the user for the game task.

4. The method according to claim 3, characterized in that The load level includes one or more of the following: The error rate of the game task and the ratio between the theta wave and the beta wave of the user's electroencephalogram (EEG) signal.

5. The method according to any one of claims 1 to 4, characterized in that The performance information includes a first EEG signal and a brain topography of the user, and determining the user's attention level includes: performing brain microstate analysis and / or spectrum analysis on the user's first EEG signal, and performing visual analysis on the brain topography to obtain analysis results; The user's attention level is determined based on the analysis result.

6. The method according to claim 5, characterized in that The obtaining of the performance information of the user when performing the at least one game task includes: Acquiring a second EEG signal, where the second EEG signal includes the first EEG signal and an interference signal, the interference signal includes a non-neurogenic artifact signal, and the non-neurogenic artifact signal includes a scalp blood flow artifact signal and a physiological artifact signal; The scalp blood flow artifact signal in the second EEG signal is removed by a short-range regression (SSR) method, and the physiological artifact signal in the second EEG signal is removed by an independent component analysis (ICA) method to obtain the first EEG signal.

7. The method according to any one of claims 1 to 4, characterized in that The sensory stimulation includes visual stimulation and / or auditory stimulation.

8. The method according to any one of claims 1 to 4, characterized in that The game tasks include one or more of the following: Dynamic target tracking tasks, color and shape recognition tasks, similar target finding tasks, and multi-target monitoring tasks.

9. An attention training device based on dynamic sensory stimulation, characterized in that: include: an interface presentation unit, configured to generate and present a game interaction interface to a user, the game interaction interface including at least one game task for training the user's attention level, the game task being presented via at least one sensory stimulus, wherein at least one variation dimension of the at least one sensory stimulus is composed of at least one variable having multiple values; A processing unit is used to: obtain performance information of the user when performing the at least one game task, the performance information is used to indicate the user's concentration level when at least one of the variables takes different values, and determine the user's attention level based on the performance information.

10. A head-mounted wearable device, characterized in that: include: a memory for storing instructions executable by the processor; A processor, wherein the processor is configured to implement the method according to any one of claims 1 to 8 when executing the instructions; A display screen is used to display game interaction images under the control of the processor.

Images