Method, device, equipment and medium for rating child cognitive and motor dual task

Abstract

The present application provides a kind of child cognitive and motor dual task rating method, device, equipment and medium, comprising: obtaining the subject data collected by piezoelectric touch panel;Wherein, subject data is in the case where the hand of subject contacts piezoelectric touch panel, subject executes at least one hand movement subtask when following the task prompt of child cognitive and motor dual task, and is collected;Task rating model is obtained by pre-training, and the hand movement of subject is rated based on subject data, and task rating result is obtained.The present application has the characteristics of simple, easy to execute, etc., so it can be better adapted to children with different motor abilities, and significantly improve the measurement accuracy and reliability, also has higher safety and comfort.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a rating method, apparatus, device, and medium for children's cognitive and motor dual tasks. Background Technology

[0002] Childhood motor assessment can help evaluate a child's motor development level, identify problems and deficiencies in the motor development process in a timely manner, provide objective evidence for determining the causes and environmental factors of these problems and deficiencies, evaluate the effectiveness of implemented interventions, guide clinical interventions and rehabilitation training, improve parents' home training skills, cultivate a positive lifestyle, and make exercise an essential component of children's habits and parenting culture.

[0003] Motor-cognitive dual-task training refers to activities that simultaneously involve physical movement and cognitive tasks, such as memorizing words while running. Training children in motor-cognitive dual-task training can have a positive impact on their development, primarily in the following aspects: Improving cognitive abilities: Motor-cognitive dual-task training can improve children's cognitive abilities, such as attention, memory, reaction speed, and mental flexibility. These cognitive abilities are crucial for children's learning and daily life. Enhancing physical fitness: Motor-cognitive dual-task training can simultaneously improve children's physical fitness, such as coordination, balance, and cardiopulmonary function. These physical qualities are critical to children's health and quality of life. Promoting nervous system development: Motor-cognitive dual-task training can promote children's nervous system development and enhance its plasticity. This helps children better adapt to environmental changes and face challenges in learning and daily life.

[0004] Traditional dual-task assessment methods for motor and cognitive abilities may face the following problems when applied to children: (1) Differences in motor abilities. Children's motor abilities may differ from adults, such as balance and coordination. Traditional dual-task assessment methods may need to be appropriately adjusted to take into account the differences in children's motor abilities. (2) Measurement accuracy and reliability: Children's motor and cognitive abilities may be more difficult to assess accurately than adults' because they may not be able to understand or follow the instructions of the assessment task. In addition, since children's cognitive and motor abilities are still developing, the reliability of the assessment results may be affected by factors related to the children themselves (such as emotions, fatigue, etc.). (3) Safety issues during the assessment process: Due to children's age and smaller size, traditional dual-task assessment methods may need to be adjusted for safety to ensure the safety and comfort of the assessment process. For example, to avoid children getting injured while performing motor tasks, it is necessary to ensure the safety and stability of the sports equipment. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a rating method, device, equipment and medium for children's cognitive and motor dual tasks, which is simple and easy to implement, and therefore can be well adapted to children with different motor abilities, and significantly improves measurement accuracy and reliability, and also has high safety and comfort.

[0006] In a first aspect, embodiments of the present invention provide a rating method for children's cognitive and motor dual tasks, including:

[0007] Acquire subject data collected by a piezoelectric touch panel; wherein, the subject data is collected when the subject's hand touches the piezoelectric touch panel and the subject performs at least one hand movement sub-task according to the task prompts of a child's cognitive and motor dual task;

[0008] The task rating model, which is obtained through pre-training, is used to rate the hand movements of the subject based on the subject data, and the task rating result is obtained.

[0009] In one embodiment, the piezoelectric touch panel is configured with a pressure sensor array; acquiring subject data collected by the piezoelectric touch panel includes:

[0010] During the process of the subject performing each of the hand movement sub-tasks, pressure data collected by each pressure sensor in the pressure sensor array is acquired;

[0011] Based on the time dimension, task dimension, and sensor dimension, the pressure data collected by each pressure sensor is constructed into a pressure data cube; wherein, each data in the pressure data cube is used to characterize the pressure data collected by the current pressure sensor when the subject performs the current hand movement sub-task at the current time.

[0012] The pressure data cube is used as subject data.

[0013] In one implementation, the task rating model includes an input embedding layer, a feature extraction unit, and a decoder layer connected in sequence; using the pre-trained task rating model, a task rating is performed on the subject's hand movements based on the subject data to obtain a task rating result, including:

[0014] The pressure data cube is linearly embedded through the input embedding layer to map the pressure data cube to a two-dimensional vector corresponding to each time dimension.

[0015] The feature extraction unit extracts features from the two-dimensional vector corresponding to each time dimension to obtain the target feature vector corresponding to the pressure data cube.

[0016] The task rating result is determined based on the target feature vector through the decoder layer; wherein the task rating result includes rating results corresponding to multiple rating indicators.

[0017] In one embodiment, the feature extraction unit includes a convolutional layer, a first encoder layer, a context encoder layer, and a fusion layer; through the feature extraction unit, features are extracted from the two-dimensional vector corresponding to each time dimension to obtain the target feature vector corresponding to the stress data cube.

[0018] The convolutional layer performs a convolution operation on the two-dimensional vector corresponding to each time dimension to extract the initial feature vector corresponding to each time dimension.

[0019] The initial feature vector is mapped and transformed through the first encoder layer to extract the high-dimensional feature vector corresponding to each time dimension.

[0020] By capturing long-range dependencies and contextual information in the high-dimensional feature vector through each second encoder layer in the context encoder layer, the context feature vector corresponding to each time dimension is extracted.

[0021] Through the Fusion layer, time features, metadata features, and external environment features are obtained. Based on the time features, metadata features, and external environment features, feature fusion is performed on the context feature vector corresponding to each time dimension to obtain the target feature vector corresponding to the stress data cube.

[0022] In one implementation, the rating indicators include hand motor skills, attention span, hyperactivity, inhibitory control, cognitive flexibility, and working memory refresh rate.

[0023] In one embodiment, the method further includes:

[0024] A rating result radar chart is generated based on the rating result corresponding to each rating indicator in the task rating result.

[0025] In one embodiment, the task prompt includes: a color touch sequence and a color ring, wherein the description text and display color corresponding to each color in the color touch sequence are the same or different, and the color ring includes multiple ring-shaped graphics, each of which corresponds to a color;

[0026] The child's cognitive and motor dual task includes: the subject changes the opening and closing degree of an approximate ring formed by the fingertips and the lower edge of the palm according to the color touch sequence, so as to match the ring-shaped pattern corresponding to one or more colors in the color touch sequence.

[0027] Secondly, embodiments of the present invention also provide a rating device for children's cognitive and motor dual tasks, comprising:

[0028] The data acquisition module is used to acquire subject data collected by the piezoelectric touch panel; wherein, the subject data is collected when the subject's hand touches the piezoelectric touch panel and the subject performs at least one hand movement sub-task according to the task prompts of the child's cognitive and motor dual tasks;

[0029] The task rating module is used to rate the subject's hand movements based on the subject's data using a pre-trained task rating model, and obtain the task rating result.

[0030] Thirdly, embodiments of the present invention also provide an electronic device, including a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement the method described in any of the first aspects.

[0031] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the method described in any of the first aspects.

[0032] This invention provides a method, apparatus, device, and medium for rating children's cognitive and motor dual tasks. It acquires subject data collected by a piezoelectric touch panel. This subject data is collected when the subject's hand touches the piezoelectric touch panel and the subject performs at least one hand movement sub-task according to the task prompts of the children's cognitive and motor dual tasks. Then, a pre-trained task rating model is used to rate the subject's hand movements based on the subject data to obtain the task rating result. The above method designs a cognitive and motor dual task rating method based on a piezoelectric touch panel. The piezoelectric touch panel detects children's hand movements, with grasping as the primary form of hand movement, ensuring safety and comfort. Furthermore, this method is simple and easy to implement, reducing errors caused by children's emotions and fatigue due to complex steps. Moreover, the excellent mechanical properties and high force-voltage responsiveness of the piezoelectric touch panel provide a guarantee of reliability and measurement accuracy for subsequent dual-task quantification data processing, thus ensuring high reliability and measurement accuracy of the task rating results output by the task rating model.

[0033] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0034] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0035] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0036] Figure 1 A flowchart illustrating a rating method for children's cognitive and motor dual tasks provided in an embodiment of the present invention;

[0037] Figure 2 A schematic diagram of a piezoelectric touch panel provided in an embodiment of the present invention;

[0038] Figure 3 A schematic diagram illustrating a task prompt provided in an embodiment of the present invention;

[0039] Figure 4 A schematic diagram of a hand movement provided in an embodiment of the present invention;

[0040] Figure 5 A schematic diagram of a pressure data cube provided in an embodiment of the present invention;

[0041] Figure 6 This is a schematic diagram of the structure of a task rating model provided in an embodiment of the present invention;

[0042] Figure 7 A schematic diagram of the structure of a rating device for children's cognitive and motor dual tasks provided in an embodiment of the present invention;

[0043] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] Currently, traditional dual-task assessment methods for motor and cognitive abilities may face the following problems when applied to children: (1) differences in motor abilities; (2) measurement accuracy and reliability; and (3) safety issues during the assessment process. Based on this, the present invention provides a rating method, device, equipment, and medium for children's cognitive and motor abilities, which is simple and easy to implement. Therefore, it can be well adapted to children with different motor abilities, significantly improves measurement accuracy and reliability, and also has high safety and comfort.

[0046] To facilitate understanding of this embodiment, a detailed description of the rating method for children's cognitive and motor dual tasks disclosed in this embodiment of the invention will be provided first. (See [link to relevant documentation]). Figure 1 The diagram shows a rating method for children's cognitive and motor dual tasks. The method mainly includes the following steps S102 to S104:

[0047] Step S102: Obtain subject data collected by the piezoelectric touch panel.

[0048] The task prompts include a color touch sequence and a color ring. In the color touch sequence, the description text and display color corresponding to each color may be the same or different. The color ring consists of multiple ring-shaped graphics, and each ring-shaped graphic corresponds to a color.

[0049] The cognitive and motor dual task for children can include multiple hand movement sub-tasks. For example, one hand movement sub-task instructs the subject to touch a ring-shaped graphic corresponding to a color displayed on a piezoelectric touch panel. The process of the subject performing the cognitive and motor dual task for children is as follows: the subject changes the opening and closing degree of the approximate ring formed by the fingertips and the lower edge of the palm in accordance with the color touch sequence to match one or more ring-shaped graphics corresponding to the colors in the color touch sequence.

[0050] The subject data is collected when the subject's hand is in contact with the piezoelectric touch panel, and the subject performs at least one hand movement sub-task according to the task prompts for a dual cognitive and motor task for children. Optionally, the subject data can be a pressure data cube, which characterizes the pressure data subjected to the piezoelectric touch panel in terms of time, sensor, and task dimensions.

[0051] Step S104: Using the pre-trained task rating model, the task rating is performed on the subject's hand movements based on the subject data to obtain the task rating result.

[0052] Among them, the task rating model can adopt the CTFCE (Convolutional Temporal Fusion with Context Encoding) model.

[0053] The task rating results include ratings for multiple rating indicators, including hand motor skills, attention span, hyperactivity, inhibitory control, cognitive flexibility, and working memory refresh.

[0054] In its implementation, subject data is input into the CTFCE model. The CTFCE model proposed in this embodiment of the invention simultaneously considers the analysis and processing of temporal dimensions, local patterns, and inter-task influences. First, the subject data is mapped into a continuous vector space, enabling the CTFCE model to better learn patterns and correlations in the temporal dimension. A convolutional layer is added before the encoder layer of the CTFCE model to capture local patterns in the channel dimension, thereby better understanding the correlations between channels and improving modeling capabilities. To enhance the analysis of inter-task influence processes, a context encoder layer is added after the encoder layer, considering the mutual influences and contextual information between different tasks, providing more accurate regression predictions, and ultimately outputting high-precision task rating results.

[0055] This invention provides a rating method for children's cognitive and motor dual tasks. It designs a rating method based on a piezoelectric touch panel, using the piezoelectric touch panel to detect children's hand movements, with grasping as the primary form of hand movement, ensuring safety and comfort. Furthermore, this method is simple and easy to implement, reducing errors caused by children's emotions and fatigue due to complex steps. Moreover, the excellent mechanical properties and high force-voltage responsiveness of the piezoelectric touch panel provide a guarantee of reliability and measurement accuracy for subsequent dual-task quantification data processing, thus ensuring high reliability and measurement accuracy of the task rating results output by the task rating model.

[0056] For ease of understanding, this embodiment of the invention provides a specific implementation of a rating method for children's cognitive and motor dual tasks.

[0057] First, in this embodiment of the invention, the piezoelectric touch panel is configured with a pressure sensor array, which includes multiple pressure sensors. For a specific implementation, see [link to relevant documentation]. Figure 2 The diagram shown is of a piezoelectric touch panel, as follows: Figure 2The piezoelectric touch panel shown in (a) consists of five layers. The first layer is a glass substrate serving as a protective layer. The lower glass cover is a layer of patterned (4×4) standard indium tin oxide (ITO) electrodes, laser-patterned. ITO was chosen as the electrode material due to its good light transmittance and low resistivity. Each electrode element has a side length of 10 mm and a spacing of 3 mm (e.g., ...). Figure 2 (As shown in (b)). The third layer is a commercially available PVDF (polyvinylidene fluoride) film, which not only converts the applied force into a charge signal but also possesses high light transmittance, high flexibility, good mechanical properties, and high force-voltage response (d33 = 30 pC / N). The fourth layer is a continuous ITO electrode, serving as the ground reference, with a PET (polyethylene terephthalate) layer at the bottom, acting as the substrate for the ground (GND) ITO layer. The thickness of each layer is as follows. Figure 2 As shown in (a). These films were laminated using a stitching and defoaming machine (SMD, Shenzhen, China) at a pressure of 0.4 MPa and a temperature of 100 °C. Compression lasted for 0.8 s.

[0058] Additionally, task prompts can be based on the Stroop paradigm. Specifically, the Stroop paradigm is a classic psychological experimental paradigm used to study psychological processes such as attention, automaticity, and inhibitory control in humans. This experimental paradigm was initially proposed by American psychologist John Ridley Stroop in 1935. In his experiments, participants were asked to quickly read a set of words whose colors differed from their literal meanings within a certain time limit; for example, the word "red" was written in blue. Participants were required to quickly identify the color of the words. The results showed that participants were distracted by the literal meaning when identifying the color, leading to a longer reaction time. This interference phenomenon is known as the Stroop effect. Besides the emotional Stroop paradigm, there are other variations, such as the word Stroop paradigm and the graphic Stroop paradigm. These variations are all improvements and developments based on the original Stroop paradigm.

[0059] Therefore, in the task prompts of this embodiment of the invention, the descriptive text and display color (i.e., font color) corresponding to each color in the color touch sequence may be the same or different. For example, see Figure 3 The diagram shows a task prompt, in which the order of the descriptive text is "yellow, blue, green, red". The actual display color of the descriptive text "yellow" is blue, the actual display color of the descriptive text "blue" is red, the actual display color of the descriptive text "green" is yellow, and the actual display color of the descriptive text "red" is green.

[0060] Additionally, the color wheel consists of multiple ring-shaped patterns, each corresponding to a specific color. Please continue reading... Figure 3 , Figure 3The illustration shows that the color ring contains multiple ring-shaped graphics with different radii, each nested within the other, and each ring-shaped graphic is displayed using a single color.

[0061] The cognitive and motor dual task for children can include multiple hand movement sub-tasks. For example, one hand movement sub-task instructs the subject to touch a ring-shaped graphic corresponding to a color displayed on a piezoelectric touch panel. The process of the subject performing the cognitive and motor dual task for children is as follows: the subject changes the opening and closing degree of the approximate ring formed by the fingertips and the lower edge of the palm in accordance with the color touch sequence to match one or more ring-shaped graphics corresponding to the colors in the color touch sequence.

[0062] Based on the above embodiments, subjects can utilize the piezoelectric touch panel described above, in conjunction with color touch sequence and color rings, to perform dual tasks of cognitive and motor skills in children. See also Figure 4 The diagram shows a hand gesture, where the subject's hand grasping motion is as follows: Figure 4 As shown, the hand-closing action is as from action A to action F, and the hand-opening action is as from action F to action A. During the hand-opening and closing process, the subject should try to maintain contact between the fingers and palm and the touchscreen.

[0063] For example, the subject observes the content on a screen or cue card. The subject imagines a color ring under their hand, forming an approximate circle with their fingertips and the bottom edge of their palm. By opening and closing their palm, they change the size of the circle to match the color ring in the image. In practical applications, while avoiding interference from text, the subject should complete the colors blue, red, yellow, and green in that order.

[0064] During the process of the subject performing the cognitive and motor dual tasks for children, the following steps (1) to (3) can be followed to obtain the subject data collected by the piezoelectric touch panel:

[0065] (1) During the process of the subject performing each hand movement sub-task, pressure data collected by each pressure sensor in the pressure sensor array is obtained.

[0066] For example, when the subject touches the blue ring displayed on the piezoelectric touch panel, the first hand movement subtask is executed. During this process, each pressure sensor will collect pressure data at various locations on the piezoelectric touch panel. When the subject touches the red ring displayed on the piezoelectric touch panel, the second hand movement subtask is specified. During this process, each pressure sensor will also collect pressure data at various locations on the piezoelectric touch panel. The above process is repeated until the subject executes the last hand movement subtask.

[0067] (2) Based on the time dimension, task dimension and sensor dimension, the pressure data collected by each pressure sensor is constructed into a pressure data cube.

[0068] Each data point in the pressure data cube represents the pressure data collected by the pressure sensor at the current time when the subject performs the current hand movement sub-task. In practical applications, this can be understood as establishing a three-dimensional coordinate system with time, task, and sensor dimensions as coordinate axes. Any point in the three-dimensional coordinate system represents the pressure data collected by the pressure sensor at that point when the subject performs the corresponding hand movement sub-task at that time.

[0069] For example, see Figure 5 The diagram shows a stress data cube, which records the performance of subjects on each individual task (there are N tasks in total, such as...). Figure 5 The pressure values ​​of p channels (sampling frequency 30Hz) of the hand movement subtasks (with a total of 4 subtasks) are recorded as a pressure data cube (3D) M.

[0070] (3) Use the stress data cube as subject data.

[0071] After acquiring the subject data, the data can be input into a task rating model to determine the task rating result. The key point of this embodiment is that, under the prompting of the Stroop paradigm, the subject performs a grasping motion (hand opening and closing) to satisfy the paradigm prompting conditions (i.e., the aforementioned task prompts). Simultaneously, the touchscreen precisely captures subtle changes in hand force. Data from each touchscreen channel, along with parameters such as paradigm completion time, accuracy, and path, are input into a trained CTFCE model network. After data regression, an assessment of the subject's dual-task completion ability—both cognitive and hand motor—is obtained.

[0072] The CTFCE model proposed in this invention simultaneously considers the analysis and processing of time dimension, local patterns, and inter-task influences. First, time-series data is mapped to a continuous vector space, enabling the model to better learn patterns and correlations in the time dimension. A convolutional layer is added before the encoder layer to capture local patterns in the channel dimension, thereby better understanding the correlations between channels and improving modeling capabilities. To enhance the analysis of inter-task influence processes, a context encoder layer is added after the encoder layer, considering the mutual influences and contextual information between different tasks, providing more accurate regression predictions.

[0073] To facilitate understanding of the task rating model, this embodiment of the invention provides a specific structure for the task rating model. See also... Figure 6The diagram shows the structure of a task rating model. The task rating model includes an input embedding layer, a feature extraction unit, and a decoder layer connected in sequence. The feature extraction unit includes a convolutional layer, a first encoder layer, a context encoder layer, and a fusion layer. The context encoder layer includes multiple second encoder layers.

[0074] exist Figure 6 Based on this, this embodiment of the invention provides an implementation method for rating a subject's hand movements based on subject data using a task rating model, thereby obtaining task rating results. The subject data cube M can be input into a trained CTFCE network model to obtain the subject's motor ability and cognitive level ratings (i.e., task rating results). See steps 1 to 3 below for details:

[0075] Step 1: Linearly embed the stress data cube into the input embedding layer to map the stress data cube to a two-dimensional vector corresponding to each time dimension.

[0076] Before inputting into the embedding layer, data preprocessing is performed. This includes operations such as data normalization, smoothing, and missing value imputation to ensure data quality and consistency.

[0077] Furthermore, fully connected layers in the input embedding layer are used to map the pressure data acquired by each pressure sensor to a low-dimensional continuous vector. Each pressure sensor acts as an input channel, and multiple fully connected layers are used to process multiple pressure sensors.

[0078] Furthermore, the low-dimensional continuous vector output by the fully connected layer is linearly embedded. The linear embedding process is as follows: (1) Input data shape: (time steps, number of sensors, number of tasks), where the time steps are the time dimension, the number of sensors are the sensor dimension, and the number of tasks are the task dimension; (2) Flatten the input data into a two-dimensional tensor with the shape of (time steps, number of sensors * number of tasks); (3) Use a fully connected layer to map the two-dimensional tensor to a lower-dimensional continuous vector representation; (4) Output shape: (time steps, embedding dimension), where (time steps, embedding dimension) is the two-dimensional vector corresponding to each of the aforementioned time dimensions.

[0079] Step 2: Using the feature extraction unit, features are extracted from the two-dimensional vector corresponding to each time dimension to obtain the target feature vector corresponding to the pressure data cube. See steps 2.1 to 2.4 below for details:

[0080] Step 2.1: Perform convolution operations on the two-dimensional vectors corresponding to each time dimension through convolutional layers to extract the initial feature vectors corresponding to each time dimension.

[0081] In practical implementation, the shape and number of adaptive convolutional kernels are defined and can be selected according to task requirements and data characteristics. The output of the embedding layer is used as the input tensor, with a shape of (time steps, embedding dimension). One-dimensional convolutional layers (Conv1D) are used to perform convolution operations between sensors, setting the size and number of convolutional kernels, as well as the parameters of other convolutional layers. Different activation functions, pooling layers, or other post-processing of convolutional operations can be selected to further extract features and reduce dimensionality. The output shape of the convolutional operation is (time steps, number of convolutional kernels). This (time steps, number of convolutional kernels) is also the initial feature vector corresponding to each time dimension.

[0082] Step 2.2: The initial feature vector is mapped and transformed through the first encoder layer to extract the high-dimensional feature vector corresponding to each time dimension.

[0083] In practice, the output of the convolution operation is used as the input data. Linear layers (fully connected layers) or other types of layers are used to map and transform the data to extract higher-level feature representations (i.e., high-dimensional feature vectors). Appropriate activation functions, regularization methods, etc., are selected as needed.

[0084] Step 2.3: Capture long-range dependencies and contextual information in the high-dimensional feature vectors through each second encoder layer in the context encoder layer to extract the context feature vector corresponding to each time dimension.

[0085] In the specific implementation, the output of the first encoding layer is used as the input data. The context encoder layer adopts structures such as Transformer to model the context and interact with the data. It contains multiple second encoder layers, each containing modules such as multi-head self-attention mechanisms and feedforward neural networks to capture long-range dependencies and contextual information in the input data. The output of the context encoder layer can be adjusted in dimension through pooling operations (such as average pooling or max pooling) or other methods to obtain the final representation vector or feature sequence (i.e., the context feature vector).

[0086] Step 2.4: Through the Fusion layer, obtain time features, metadata features, and external environment features, and based on the time features, metadata features, and external environment features, perform feature fusion on the context feature vector corresponding to each time dimension to obtain the target feature vector corresponding to the stress data cube.

[0087] In practical applications, the Fusion layer also needs to acquire additional feature sources, including: time features (hours, minutes, seasons, weekdays / weekends, etc.), metadata features (sensor location, installation angle, manufacturer information, etc.), external environmental features (weather conditions, temperature, humidity, air pressure, etc.), and other features (historical data of the sensor, data from other related tasks, etc.).

[0088] The output of the context encoder layer is concatenated or combined with representations from other feature sources in an appropriate manner to form a comprehensive feature representation. Different methods are used for feature fusion, such as weighted fusion, concatenated fusion, and element-wise operations. If weighted fusion is used, trainable weights can be learned to adaptively adjust the contributions of different features. If concatenated fusion is used, representations of different features can be concatenated along the feature dimension. The output of the fusion layer is the fused feature representation (i.e., the target feature vector), which incorporates information from the context encoder layer and other feature sources.

[0089] Step 3: Determine the task rating result based on the target feature vector through the decoder layer; the task rating result includes the rating results corresponding to multiple rating indicators.

[0090] In one example, the decoder layer design is as follows: determine the structure and parameter settings of the decoder, and choose to use a recurrent neural network (such as LSTM, GRU) or an attention mechanism, etc.

[0091] In one example, the decoder input is prepared as follows: the output of the Fusion layer is used as the initial input to the decoder. The output of the Fusion layer is mapped and transformed through a linear layer (fully connected layer) or other means to match the hidden state dimension of the decoder.

[0092] In one example, the decoder operates as follows: It generates regression results stepwise using an autoregressive approach, meaning one regression result is generated at each time step. At each time step, the decoder generates a regression result based on the current input and the hidden state, and uses it as the input for the next time step. Different decoder strategies can be chosen, such as using an attention mechanism to weight the input sequence to better focus on important time steps.

[0093] In one example, the output preparation is as follows: Based on the decoder's output, some follow-up processing is performed, such as applying activation functions and adjusting dimensions. The final output is the regression value (i.e., the task rating result).

[0094] For further details, please continue to see Figure 6 , Figure 6The diagram also illustrates how a radar chart of rating results can be generated based on the rating results for each rating indicator in the task rating results. The rating results can be presented in the form of levels. For example, the radar chart can clearly and intuitively display the rating results for each rating indicator of the subject, such as a significantly higher level for the subject's excessive positive feedback and working memory refresh ability, and significantly lower levels for attention concentration and inhibitory control ability.

[0095] The dual-task assessment of hand movement and cognitive control abilities in this embodiment of the invention establishes a relationship model between the paradigm test data matrix and hand movement and refined cognitive abilities through a CTFCE model. Traditional dual-task paradigms, such as answering questions while moving, may carry certain risks, such as excessive attention to cues while neglecting the direction of movement. This paradigm displays the cues on a screen, allowing the subject to simultaneously observe both movement and cues. Traditional paradigms struggle with refining attention levels; this model, through data regression using a CTFCE network, can refine and quantify attention types. This model places significant emphasis on changes in interdimensional data (reflecting cognitive performance under different time intervals and time pressures), which can also reflect abilities such as concentration, working memory refresh capacity, and cognitive flexibility; therefore, this assessment process is effective.

[0096] Optionally, the aforementioned CTFCE model can also be used to assess a person's cognitive abilities. Some studies or cases have quantified cognitive abilities numerically. For example, the Montreal Cognitive Assessment (MoCA) is an assessment tool used to screen for mild cognitive impairment. It can assess multiple cognitive domains, with a total score of 30 points.

[0097] In practical applications, the CTFCE model needs to be trained before it can be put into use in order to ensure that the output of the CTFCE model has high accuracy.

[0098] To facilitate understanding, this embodiment of the invention also provides an implementation method for training a CTFCE model. Specifically, before model training, a professional clinician or psychotherapist needs to complete a single-item quantitative rating for n (n>100) children. For each subject, a clinical paradigm scale or techniques such as EEG and EMG are used to assess cognitive control abilities, including attention span, hyperactivity, inhibitory control, cognitive flexibility, and working memory refresh rate. A 4- or 5-level rating (5-level rating: excellent, good, average, fair, poor) is used to score the subject. The child's hand motor ability is scored using a child development manual assessment. Then, this dual-task combined test is performed on these n subjects, resulting in n data matrices M. The ratings are digitized (e.g., 5-level rating scores are: excellent 90 points, good 70 points, average 50 points, fair 30 points, poor 10 points), and the model is trained with the data matrices to find the mapping relationship.

[0099] In summary, the rating method for children's cognitive and motor dual tasks provided in this embodiment of the invention has at least the following characteristics:

[0100] (1) In this embodiment of the invention, a dual-task assessment method for motion cognition based on a piezoelectric touchscreen is designed. The touchscreen detects children's hand movements, with grasping as the primary form of hand movement, ensuring safety and comfort. This method is simple and easy to implement, reducing errors caused by children's emotions and fatigue due to complex steps. The excellent mechanical properties and high force-voltage responsiveness of the piezoelectric touchscreen provide a guarantee of reliability and measurement accuracy for subsequent dual-task quantification data processing.

[0101] (2) In the dual-task evaluation process, it is necessary to conduct a comprehensive analysis of the task data in terms of time dimension, local patterns, and inter-task influence. Existing deep learning algorithms do not fully consider these three types of issues. The CTFCE model proposed in this embodiment of the invention takes into account the analysis and processing of time dimension, local patterns, and inter-task influence. First, the time series data is mapped to a continuous vector space, enabling the model to better learn the patterns and correlations in the time dimension; a convolutional layer is added before the encoder layer of the model, which can capture local patterns in the channel dimension, thereby better understanding the correlation between channels and improving the modeling ability; to strengthen the analysis of the inter-task influence process, a context encoder layer is added after the encoder layer, which considers the mutual influence and context information between different tasks, and provides more accurate regression prediction.

[0102] Based on the foregoing embodiments, this invention provides a rating device for children's cognitive and motor dual tasks, see [link to documentation]. Figure 7 The diagram shows a structural schematic of a rating device for children's cognitive and motor dual tasks. The device mainly includes the following parts:

[0103] The data acquisition module 702 is used to acquire subject data collected by the piezoelectric touch panel; wherein, the subject data is collected when the subject's hand touches the piezoelectric touch panel and the subject performs at least one hand movement sub-task according to the task prompts of the child's cognitive and motor dual tasks.

[0104] The task rating module 704 is used to rate the hand movements of subjects based on subject data using a pre-trained task rating model, and obtain task rating results.

[0105] This invention provides a rating device for children's cognitive and motor dual tasks, and designs a rating method for cognitive and motor dual tasks based on a piezoelectric touch panel. The piezoelectric touch panel detects children's hand movements, with grasping as the main form of children's hand movements, ensuring safety and comfort. In addition, this method is simple and easy to implement, which can reduce the error problems caused by children's emotions and fatigue due to complex steps. Furthermore, the good mechanical properties and high force-voltage responsiveness of the piezoelectric touch panel provide a guarantee of reliability and measurement accuracy for subsequent dual-task quantification data processing, thereby making the task rating results output by the task rating model highly reliable and accurate.

[0106] In one embodiment, the piezoelectric touch panel is configured with a pressure sensor array; the data acquisition module 702 is further configured to:

[0107] Pressure data was acquired from each pressure sensor in the pressure sensor array during the subject's performance of each hand movement sub-task.

[0108] Based on the time dimension, task dimension, and sensor dimension, the pressure data collected by each pressure sensor is constructed into a pressure data cube; where each data point in the pressure data cube is used to characterize the pressure data collected by the current pressure sensor when the subject performs the current hand movement sub-task at the current time.

[0109] The stress data cube was used as subject data.

[0110] In one implementation, the task rating model includes an input embedding layer, a feature extraction unit, and a decoder layer connected in sequence; the task rating module 704 is further configured to:

[0111] By using the input embedding layer, the stress data cube is linearly embedded to map the stress data cube to a two-dimensional vector corresponding to each time dimension.

[0112] The feature extraction unit extracts features from the two-dimensional vector corresponding to each time dimension to obtain the target feature vector corresponding to the pressure data cube.

[0113] The task rating result is determined based on the target feature vector through the decoder layer; the task rating result includes the rating results corresponding to multiple rating indicators.

[0114] In one implementation, the feature extraction unit includes a convolutional layer, a first encoder layer, a context encoder layer, and a fusion layer; the task rating module 704 is further configured to:

[0115] By using convolutional layers, convolution operations are performed on the two-dimensional vectors corresponding to each time dimension to extract the initial feature vectors corresponding to each time dimension;

[0116] The initial feature vector is mapped and transformed through the first encoder layer to extract the high-dimensional feature vector corresponding to each time dimension.

[0117] By capturing long-range dependencies and contextual information in high-dimensional feature vectors through each second encoder layer in the context encoder layer, the context feature vector corresponding to each time dimension is extracted.

[0118] Through the Fusion layer, time features, metadata features, and external environment features are obtained. Based on the time features, metadata features, and external environment features, the context feature vectors corresponding to each time dimension are fused to obtain the target feature vector corresponding to the stress data cube.

[0119] In one implementation, the rating indicators include hand motor skills, attention span, hyperactivity, inhibitory control, cognitive flexibility, and working memory refresh rate.

[0120] In one implementation, a radar chart generation module is further included, for:

[0121] A radar chart of rating results is generated based on the rating results for each rating indicator in the task rating results.

[0122] In one implementation, the task prompt includes: a color touch sequence and a color ring, wherein the descriptive text and display color corresponding to each color in the color touch sequence are the same or different, and the color ring includes multiple ring-shaped graphics, each ring-shaped graphic corresponding to a color;

[0123] The cognitive and motor dual task for children involves participants changing the opening and closing degree of an approximate ring formed by their fingertips and the lower edge of their palms in accordance with the color touch sequence, in order to match the ring-shaped pattern corresponding to one or more colors in the color touch sequence.

[0124] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0125] This invention provides an electronic device, specifically, the electronic device includes a processor and a storage device; the storage device stores a computer program, and the computer program, when run by the processor, executes the method described in any of the above embodiments.

[0126] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. The electronic device 100 includes: a processor 80, a memory 81, a bus 82, and a communication interface 83. The processor 80, the communication interface 83, and the memory 81 are connected through the bus 82. The processor 80 is used to execute executable modules, such as computer programs, stored in the memory 81.

[0127] The memory 81 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 83 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0128] Bus 82 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 8 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0129] The memory 81 is used to store programs. After receiving an execution instruction, the processor 80 executes the program. The method executed by the device for defining the flow process disclosed in any of the foregoing embodiments of the present invention can be applied to the processor 80 or implemented by the processor 80.

[0130] The processor 80 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 80 or by software instructions. The processor 80 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 81. The processor 80 reads the information in memory 81 and, in conjunction with its hardware, completes the steps of the above method.

[0131] The computer program product of the readable storage medium provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For specific implementation, please refer to the foregoing method embodiments, which will not be repeated here.

[0132] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0133] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A rating method for children's cognitive and motor dual tasks, characterized in that, include: Acquire subject data collected by a piezoelectric touch panel; wherein, the subject data is collected when the subject's hand touches the piezoelectric touch panel and the subject performs at least one hand movement sub-task according to the task prompts of a child's cognitive and motor dual task; the piezoelectric touch panel consists of five layers, including: a glass substrate as a protective layer, a patterned standard indium tin oxide electrode, a PVDF film, a continuous ITO electrode, and a PET layer; The task rating model, which is obtained through pre-training, is used to rate the hand movements of the subject based on the subject data, and the task rating result is obtained. The task prompt includes: a color touch sequence and a color ring. In the color touch sequence, the description text and display color corresponding to each color may be the same or different. The color ring includes multiple ring-shaped graphics with different radii, each of which is nested with another ring-shaped graphic, and each ring-shaped graphic corresponds to a color. The child's cognitive and motor dual task includes: the subject changes the opening and closing degree of an approximate ring formed by the fingertips and the lower edge of the palm according to the color touch sequence, so as to match the ring-shaped pattern corresponding to one or more colors in the color touch sequence.

2. The rating method for children's cognitive and motor dual tasks according to claim 1, characterized in that, The piezoelectric touch panel is equipped with a pressure sensor array; Acquire subject data collected by the piezoelectric touch panel, including: During the process of the subject performing each of the hand movement sub-tasks, pressure data collected by each pressure sensor in the pressure sensor array is acquired; Based on the time dimension, task dimension, and sensor dimension, the pressure data collected by each pressure sensor is constructed into a pressure data cube; wherein, each data in the pressure data cube is used to characterize the pressure data collected by the current pressure sensor when the subject performs the current hand movement sub-task at the current time. The pressure data cube is used as subject data.

3. The rating method for children's cognitive and motor dual tasks according to claim 2, characterized in that, The task rating model comprises an input embedding layer, a feature extraction unit, and a decoder layer connected in sequence. Using the pre-trained task rating model, the model rates the subject's hand movements based on the subject's data, yielding a task rating result, including: The pressure data cube is linearly embedded through the input embedding layer to map the pressure data cube to a two-dimensional vector corresponding to each time dimension. The feature extraction unit extracts features from the two-dimensional vector corresponding to each time dimension to obtain the target feature vector corresponding to the pressure data cube. The task rating result is determined based on the target feature vector through the decoder layer; wherein, the task rating result includes rating results corresponding to multiple rating indicators.

4. The rating method for children's cognitive and motor dual tasks according to claim 3, characterized in that, The feature extraction unit includes a convolutional layer, a first encoder layer, a context encoder layer, and a fusion layer. Through this feature extraction unit, features are extracted from the two-dimensional vector corresponding to each time dimension to obtain the target feature vector corresponding to the stress data cube. The convolutional layer performs a convolution operation on the two-dimensional vector corresponding to each time dimension to extract the initial feature vector corresponding to each time dimension. The initial feature vector is mapped and transformed through the first encoder layer to extract the high-dimensional feature vector corresponding to each time dimension. By capturing long-range dependencies and contextual information in the high-dimensional feature vector through each second encoder layer in the context encoder layer, the context feature vector corresponding to each time dimension is extracted. Through the Fusion layer, time features, metadata features, and external environment features are obtained. Based on the time features, metadata features, and external environment features, feature fusion is performed on the context feature vector corresponding to each time dimension to obtain the target feature vector corresponding to the stress data cube.

5. The method of claim 3, wherein the child cognitive and motor dual task rating is based on a cognitive and motor dual task rating scale. The rating indicators include hand motor skills, attention span, hyperactivity, inhibitory control, cognitive flexibility, and working memory refresh rate.

6. The method of claim 3, wherein the child cognitive and motor dual task rating is based on a cognitive and motor dual task rating scale. The method further includes: A rating result radar chart is generated based on the rating result corresponding to each rating indicator in the task rating result.

7. A device for rating child cognitive and motor dual tasking, characterized in that, The rating method for implementing the dual cognitive and motor tasks for children as described in claim 1 includes: The data acquisition module is used to acquire subject data collected by the piezoelectric touch panel; wherein, the subject data is collected when the subject's hand touches the piezoelectric touch panel and the subject performs at least one hand movement sub-task according to the task prompts of the child's cognitive and motor dual tasks; The task rating module is used to rate the subject's hand movements based on the subject's data using a pre-trained task rating model, and obtain the task rating result.

8. An electronic device, comprising: The method includes a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement the method of any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions that, when invoked and executed by a processor, cause the processor to perform the method according to any one of claims 1 to 6.

Images