Method and device for evaluating dyspraxia, and storage medium

Abstract

The application provides an evaluation method and device for movement coordination disorder and a storage medium, and comprises the following steps: obtaining original signals of a test user within a preset time length, wherein the original signals are acceleration original waveform signals collected through a preset sampling frequency; determining a plurality of target amplitude sequences meeting a preset standard based on a preset sliding window and the original signals; calculating sample entropy features, fractal dimension features and line energy features corresponding to each target amplitude sequence; and determining a movement coordination disorder evaluation result of the test user according to the sample entropy features, the fractal dimension features, the line energy features and a preset movement coordination disorder identification model. The application realizes full-automatic screening of movement coordination disorder and improves the accuracy of evaluation.

Description

Technical Field

[0001] This application relates to the field of data processing technology, specifically to an assessment method, device, and storage medium for motor coordination disorders. Background Technology

[0002] Developmental coordination disorder (DCD) is a neurodevelopmental disorder characterized by significant impairment in motor coordination, severely impacting children's daily activities, academic performance, and long-term physical and mental health. Current assessment methods for DCD primarily utilize standardized behavioral scales, such as the Childhood Motor Assessment Kit-2 (MABC-2) and the Large Muscle Group Development Test-3 (TGMD-3). These assessment methods require professionals to observe children performing designated movements in fixed, controlled environments such as hospitals or laboratories, and then manually score them. The scoring results are highly dependent on the assessor's personal experience, and for children with mild symptoms or those in a borderline state, inconsistencies in scoring among assessors are common. Furthermore, the lack of a unified and objective physical evaluation standard, coupled with the fact that unfamiliar testing environments can induce anxiety in children, leading to performances inconsistent with their actual abilities, and the short, single-session testing model cannot accurately reflect the child's true motor state in real-life situations, resulting in inaccurate assessment results. Summary of the Invention

[0003] This application provides a method, apparatus, and storage medium for assessing motor coordination disorders. By calculating sample entropy characteristics, fractal dimension characteristics, and linear energy characteristics based on the original signal, the accuracy of the assessment for motor coordination disorders is improved.

[0004] In a first aspect, embodiments of this application provide a method for assessing motor coordination disorders, comprising: Acquire the raw signal of the subject within a preset time period, wherein the raw signal is the original acceleration waveform signal obtained by sampling at a preset sampling frequency; Based on a preset sliding window and the original signal, multiple target amplitude sequences that meet preset criteria are determined; Calculate the sample entropy feature, fractal dimension feature, and linear energy feature corresponding to each of the target amplitude sequences; The assessment results of the subject's motor coordination disorder are determined based on the sample entropy features, the fractal dimension features, the linear energy features, and the preset motor coordination disorder identification model.

[0005] Secondly, embodiments of this application provide an assessment device for motor coordination disorders, comprising: The first acquisition unit is used to acquire the original signal of the test user within a preset time period, wherein the original signal is the original acceleration waveform signal acquired by a preset sampling frequency. The second acquisition unit is used to determine multiple target amplitude sequences that meet preset criteria based on a preset sliding window and the original signal; The calculation unit is used to calculate the sample entropy features, fractal dimension features, and linear energy features corresponding to each of the target amplitude sequences; The determining unit is used to determine the assessment result of the subject's motor coordination disorder based on the sample entropy feature, the fractal dimension feature, the linear energy feature, and the preset motor coordination disorder identification model.

[0006] Thirdly, embodiments of this application provide a terminal device, the terminal device including at least one processor, a communication interface and a memory, the communication interface being used to send and / or receive data, the memory being used to store a computer program, and the at least one processor being used to call the computer program stored in the memory to implement any of the methods of the first aspect of this application.

[0007] Fourthly, embodiments of this application provide an electronic device including a processor and a memory, the memory being used to store computer program code, the computer program code including computer instructions, wherein, when the processor executes the computer instructions, the electronic device executes the instructions as in any of the methods of the first aspect of this application.

[0008] Fifthly, embodiments of this application provide a computer-readable storage medium storing a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform some or all of the steps described in any method of the first aspect of this application.

[0009] Sixthly, embodiments of this application provide a computer program operable to cause a computer to perform some or all of the steps described in any method of the first aspect of this application. This computer program may be a software installation package.

[0010] As can be seen, in this embodiment, the original signal of the test user within a preset time period is acquired. The original signal is the original acceleration waveform signal obtained by sampling at a preset frequency. Based on a preset sliding window and the original signal, multiple target amplitude sequences that meet preset criteria are determined. The sample entropy features, fractal dimension features, and linear energy features corresponding to each target amplitude sequence are calculated. The assessment result of the test user's motor coordination disorder is determined based on the sample entropy features, fractal dimension features, linear energy features, and a preset motor coordination disorder identification model. This application extracts target amplitude sequences from the original signal and realizes fully automatic screening of motor coordination disorders based on the sample entropy features, fractal dimension features, and linear energy features calculated from the target amplitude sequences, thereby improving the accuracy of the assessment. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A schematic diagram of an evaluation system provided for an embodiment of this application; Figure 2 A flowchart illustrating an assessment method for motor coordination disorder provided in an embodiment of this application; Figure 3 A schematic diagram illustrating an application scenario provided in an embodiment of this application; Figure 4 A flowchart illustrating the method for determining an amplitude sequence provided in an embodiment of this application; Figure 5 A flowchart illustrating the method for determining sample entropy features provided in an embodiment of this application; Figure 6 A flowchart illustrating the method for determining fractal dimension features provided in an embodiment of this application; Figure 7 A functional unit block diagram of an assessment device for motor coordination disorder provided in an embodiment of this application; Figure 8 This is a schematic diagram of the structure of a server provided in an embodiment of this application. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0014] The terms "second," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such processes, methods, products, or apparatus.

[0015] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0016] Please combine Figure 1 , Figure 1 A schematic diagram of an evaluation system provided for an embodiment of this application, such as... Figure 1 As shown, the evaluation system includes a server and an accelerometer, which are communicatively connected to facilitate data exchange. The server in the evaluation system is used to implement any of the following methods: it can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms.

[0017] Please refer to the following. Figure 2 , Figure 2 A flowchart illustrating an assessment method for motor coordination disorders provided in this application embodiment is shown below. Figure 2 As shown in the embodiment of this application, a method for assessing motor coordination disorder includes: S201, Obtain the raw signal of the subject within a preset time period.

[0018] The original signal is the original acceleration waveform signal acquired through a preset sampling frequency. Specifically, the accelerometer is fixed to a specific body part of the test user to collect data. Please refer to... Figure 3 , Figure 3 This is a schematic diagram illustrating an application scenario provided in an embodiment of this application, such as... Figure 3 As shown, the specific body part can be the right hip joint, near the anterior superior iliac spine. The target accelerometer 300 is fixed to this specific body part of the test user, especially when the test user is a child, to avoid interference and improve the accuracy of the collected data. In this example, the accelerometer can be a triaxial accelerometer. After collecting data within a preset duration at a preset frequency, the accelerometer outputs the raw acceleration waveform signal without hardware filtering, thus preserving the high-frequency physical waveform including neuromuscular fine-tuning characteristics. The preset sampling frequency can be 100Hz; however, it is understood that the preset sampling frequency can be set according to needs and is not limited here. The preset duration is set to a continuous number of days, such as seven days, to cover the test user's daily life trajectory and improve the accuracy of the detection results. It is understood that the preset duration can be set according to actual needs, such as more than seven days; the specific duration is not limited here.

[0019] S202, based on a preset sliding window and the original signal, determine multiple target amplitude sequences that meet preset criteria.

[0020] After acquiring the raw signal, multiple target amplitude sequences that meet preset criteria are determined based on a pre-defined sliding window and the raw signal. These preset criteria include intensity, period, and steady-state criteria. By selecting multiple target amplitude sequences using these preset criteria, environmental noise and interference from non-target behaviors are reduced, thereby improving the accuracy of the evaluation results.

[0021] S203, calculate the sample entropy feature, fractal dimension feature and linear energy feature corresponding to each of the target amplitude sequences.

[0022] The calculation of sample entropy, fractal dimension, and linear energy features for each target amplitude sequence involves several steps. Sample entropy (SampEn) measures the irregularity of human temporal movements. Specifically, it quantifies the self-similarity of gait temporal trajectories by calculating the probability of recurring patterns in the target amplitude sequence. If the central nervous system's recruitment and control mechanism for motor units is impaired in the subject, the disorder of their gait temporal trajectory will increase, and the corresponding sample entropy feature value will be higher.

[0023] Higuchi Fractal Dimension (HFD) is used to characterize the geometric complexity of action sequences. Self-similarity features at different scales can be calculated using specific algorithms.

[0024] Linear energy characteristics can be used to assess the average explosive power and energy level of macroscopic movements such as limb coordination. Specifically, linear energy characteristics can include root mean square (RMS) amplitude and Euclidean Norm Minus One (ENMO) signal vector amplitude.

[0025] S204, determine the assessment result of the subject's motor coordination disorder based on the sample entropy feature, the fractal dimension feature, the linear energy feature, and the preset motor coordination disorder identification model.

[0026] Specifically, the assessment results of the participants' motor coordination disorders are determined based on sample entropy characteristics, fractal dimension characteristics, linear energy characteristics, and a pre-defined motor coordination disorder identification model. The motor coordination disorder identification model can accept multiple feature inputs and output a predicted probability value of the participants' disease.

[0027] As can be seen, in this example, by collecting the original signals of the test users within a preset time period, extracting the target amplitude sequence from the original signals, and calculating the sample entropy characteristics and fractal dimension characteristics, a fully automatic and objective early screening for children's motor coordination disorders is achieved. This avoids the impaired central control ability of test users with DCD being masked by the body's compensatory mechanisms, thereby improving the accuracy of the assessment.

[0028] In one possible example, the preset criteria include an intensity criterion, a periodicity criterion, and a steady-state criterion. Determining multiple target amplitude sequences that satisfy the preset criteria based on a preset sliding window and the original signal includes: determining a first amplitude sequence based on a preset vector net amplitude calculation formula and the original signal; scanning the first amplitude sequence according to the sliding window with a preset window length and a preset step size to obtain multiple second amplitude sequences; determining whether each second amplitude sequence satisfies the preset criteria, including: if the mean of the second amplitude sequence falls within a preset intensity threshold range, then the second amplitude sequence is determined to satisfy the intensity criterion; if the periodicity characteristic of the second amplitude sequence satisfies a preset periodicity characteristic, then the second amplitude sequence is determined to satisfy the periodicity criterion; calculating the ratio between the standard deviation and the mean of the second amplitude sequence to obtain the coefficient of variation corresponding to the second amplitude sequence; if the coefficient of variation is lower than a preset steady-state threshold, then the second amplitude sequence is determined to satisfy the steady-state criterion; and determining the second amplitude sequences among the various second amplitude sequences that satisfy the intensity criterion, the periodicity criterion, and the steady-state criterion as the multiple target amplitude sequences.

[0029] For specific examples, please refer to Figure 4 , Figure 4 A flowchart illustrating the method for determining an amplitude sequence provided in this application embodiment is shown below. Figure 4 As shown, determining the first amplitude sequence based on the preset vector net amplitude calculation formula and the original signal includes: S401, convert the original signal into a time-series signal in units of gravitational acceleration.

[0030] The raw signals collected in real time by the accelerometer include data from the X-axis, Y-axis, and Z-axis of the test user. The corresponding calibration conversion coefficients are found according to the acceleration range that the accelerometer can detect. The raw signals are then converted into acceleration time-series signals with gravitational acceleration g as the unit of measurement and arranged in the order of the collection time of each sampling point.

[0031] S402, bandpass filtering is performed on the timing signal according to the preset low-frequency cutoff threshold and high-frequency cutoff threshold to obtain the denoised timing signal.

[0032] Specifically, based on preset low-frequency and high-frequency cutoff thresholds, the timing signal undergoes bandpass filtering, retaining signals between these thresholds. By filtering out signals below the low-frequency cutoff threshold, low-frequency drift caused by changes in gravity direction can be eliminated; by filtering out signals above the high-frequency cutoff threshold, extremely high-frequency noise caused by environmental vibrations can be eliminated, thus obtaining a denoised timing signal.

[0033] S403, based on the preset vector net amplitude calculation formula, calculate the vector net amplitude of each signal in the denoised time-series signal, remove the static gravity component, and obtain the first amplitude sequence composed of the vector net amplitudes after removing the static gravity component.

[0034] The calculation involves determining the net vector magnitude of each signal in the denoised time-series signal. The formula for calculating the net vector magnitude is as follows: ; in, Let x be the net vector magnitude corresponding to the t-th signal within the denoised time-series signal. t Let y be the X-axis value corresponding to the t-th signal. t Let z be the Y-axis value corresponding to the t-th signal. tLet be the Z-axis value corresponding to the t-th signal. The X-axis, Y-axis, and Z-axis values ​​of each signal are squared and summed. The square root of this sum is then taken to obtain the vector amplitude. One is subtracted from this vector amplitude (removing the static component of Earth's gravity, 1g), and then compared with 0. The maximum of the two values ​​is taken to obtain the net vector amplitude. An amplitude sequence is formed from the net vector amplitudes after removing the static gravity component.

[0035] It is evident that by processing the original signal and removing the static component of Earth's gravity, dynamic acceleration data related to human movement can be preserved, reducing various noise interferences in natural life scenarios and improving the efficiency of subsequent data processing.

[0036] The preset standards can include intensity standards, periodic standards, and steady-state standards. A first amplitude sequence is scanned using a sliding window with a preset window length and preset step size to obtain multiple second amplitude sequences. In specific applications, the preset window length can be set to 6,000 sampling points corresponding to 60 seconds, and the preset step size can be set to 1 second. The first amplitude sequence of the test user is scanned using the sliding window. Furthermore, there should be no non-wearing markers or data loss points within the sliding window. If non-wearing markers or data loss points exist, the second amplitude sequence corresponding to the sliding window is discarded. This ensures the physical temporal continuity of the 6,000 high-frequency sampling points. Next, it can be determined whether each second amplitude sequence meets the preset standards. Specifically, the mean of the second amplitude sequence is calculated. If the mean of the second amplitude sequence falls within the preset intensity threshold range, then the second amplitude sequence is considered to meet the intensity standard. This preset intensity threshold range can be between 0.05g and 0.4g to exclude signal interference caused by prolonged sitting or vigorous running and jumping by the test user.

[0037] If the periodicity of the second amplitude sequence meets a preset periodicity characteristic, then the second amplitude sequence is considered to meet the periodicity standard. Specifically, an autocorrelation operation is performed on the second amplitude sequence based on the autocorrelation function (ACF) to obtain an autocorrelation sequence. The autocorrelation sequence can include multiple autocorrelation function values, each corresponding to a specific time delay. The autocorrelation sequence characterizes the similarity between the second amplitude sequence and itself at various time delays. The position of the main peak and the corresponding peak value are obtained within a preset time delay range. This main peak position is used to characterize the lag time of the action signal repetition. Next, it is determined whether the peak value is greater than or equal to a preset peak value, which can be set to 0.6. The size of the peak value reflects the regularity of the gait rhythm under neural control; a larger peak value indicates more continuous movement. The ratio of the preset sampling frequency to the number of lag sampling points corresponding to the peak position is determined as the walking frequency corresponding to the second amplitude sequence. Then, it is determined whether the walking frequency belongs to a preset step frequency range, which can be set between 1.5Hz and 2.5Hz. Finally, when the main peak value is greater than or equal to the preset main peak value and the walking frequency belongs to the preset step frequency range, the second amplitude sequence is determined to meet the periodic standard.

[0038] Specifically, the ratio between the standard deviation and the mean of the second amplitude sequence is calculated, and this ratio is set as the coefficient of variation (CV) to obtain the coefficient of variation corresponding to the second amplitude sequence. The formula for calculating the coefficient of variation is: the coefficient of variation equals the standard deviation of the second amplitude sequence divided by the mean of the second amplitude sequence. After obtaining the coefficient of variation, if it is lower than a preset steady-state threshold, the second amplitude sequence is considered to meet the steady-state criteria. If it is equal to or higher than the preset steady-state threshold, a sudden action switch or pause is determined within the second amplitude sequence. The preset steady-state threshold can be set to 15%. Finally, the second amplitude sequences that meet the intensity, period, and steady-state criteria are identified as multiple target amplitude sequences.

[0039] As can be seen in this example, by using multi-dimensional quantification standards, the target amplitude sequence can be extracted from the original data, making the evaluation results based on the target amplitude sequence more accurate.

[0040] In one possible example, after calculating the vector net amplitude of each signal in the denoised time-series signal based on a preset vector net amplitude calculation formula, and removing the static gravity component to obtain the first amplitude sequence composed of the vector net amplitudes after removing the static gravity component, the method further includes: detecting the first amplitude sequence using a sliding time window. Specifically, for example, the length of the sliding time window is 60 seconds. The standard deviation of the vector net amplitude within the sliding time window is calculated. If the peak values ​​of the standard deviation are all less than 0.05g, the time period corresponding to the sliding time window is determined to be a non-wearing period of sleep or a sedentary period. The vector net amplitude corresponding to the sliding time window is then removed, thereby eliminating data interference from sleep and other time-based data acquisition and improving the efficiency of subsequent data processing.

[0041] In one possible example, calculating the sample entropy feature, fractal dimension feature, and linear energy feature corresponding to each of the target amplitude sequences includes: for each of the target amplitude sequences, performing the following steps: obtaining the net amplitude of each target vector in the target amplitude sequence; squaring each net amplitude of the target vector to obtain multiple squared values; calculating the arithmetic mean of the multiple squared values; performing a square root operation on the arithmetic mean to obtain the root mean square value corresponding to the target amplitude sequence; determining the linear energy feature based on the root mean square value; determining the sample entropy feature based on the target amplitude sequence; and determining the fractal dimension feature based on the target amplitude sequence.

[0042] In a specific example, for each target amplitude sequence: First, the sample entropy feature is determined based on the target amplitude sequence. The sample entropy feature can quantify the chaos of the temporal trajectory corresponding to the target amplitude sequence, thus reflecting the state of the neural regulation mechanism at a microscopic level. Furthermore, the fractal dimension feature is determined based on the target amplitude sequence. The fractal dimension feature can represent the geometric complexity of the action sequence at different time scales.

[0043] Specifically, linear energy characteristics can be determined based on the target amplitude sequence to assess the explosive power and energy level of macroscopic movements. Specifically, the net amplitude of each target vector in the target amplitude sequence is obtained, and each net amplitude of the target vector is squared to obtain multiple squared values. The arithmetic mean of these multiple squared values ​​is calculated, and the square root of the arithmetic mean is taken to obtain the root mean square value corresponding to the target amplitude sequence. Linear energy characteristics are then determined based on the root mean square value.

[0044] As can be seen, in this example, by using the target amplitude sequence as a unified data basis for calculation, sample entropy features, fractal dimension features, and linear energy features can be extracted under the same physical benchmark, providing data support for subsequent evaluation and improving the accuracy and reliability of the evaluation.

[0045] In one possible example, please see Figure 5 , Figure 5 A flowchart illustrating the method for determining sample entropy features provided in this application embodiment is shown in the figure. The step of determining sample entropy features based on the target amplitude sequence includes: S501, obtain multiple first template vectors from the target amplitude sequence according to a first preset length.

[0046] First, multiple first template vectors are obtained from the target amplitude sequence according to a first preset length. Specifically, the first preset length can be set to 2, and the first template vectors are continuously extracted point by point from the target amplitude sequence based on this length to obtain multiple first template vectors.

[0047] S502, calculate the first Chebyshev distance between any two first template vectors among the plurality of first template vectors, and count the first pairing values ​​where the first Chebyshev distance is less than the preset similarity tolerance.

[0048] The first Chebyshev distance is calculated by subtracting the corresponding elements of the two first template vectors and taking the absolute value of each subtraction. The largest absolute value among all these values ​​is taken as the Chebyshev distance between the two sets of template vectors. The preset similarity tolerance can be set to twice the standard deviation of the target amplitude sequence, i.e., r = 0.2SD, where r is the preset similarity tolerance and SD is the standard deviation of the target amplitude sequence. The second number of pairings between any two vectors in the multiple first template vectors is counted, and the first pairing value where the first Chebyshev distance is less than the preset similarity tolerance is also counted.

[0049] S503, obtain multiple second template vectors from the target amplitude sequence according to a second preset length, wherein the second preset length is greater than the first preset length.

[0050] Specifically, multiple second template vectors are obtained from the target amplitude sequence according to a second preset length, where the second preset length is greater than the first preset length. For example, the second preset length can be the first preset length plus one.

[0051] S504, calculate the second Chebyshev distance between any two second template vectors among the plurality of second template vectors, and count the second pairing values ​​where the second Chebyshev distance is less than the preset similarity tolerance.

[0052] Next, the second Chebyshev distance between any two second template vectors in the plurality of second template vectors is calculated, the third number of pairings of any two vectors in the plurality of second template vectors is counted, and the second pairing value where the second Chebyshev distance is less than the preset similarity tolerance is counted.

[0053] S505, determine the sample entropy feature based on the second pairing value and the first pairing value.

[0054] Specifically, the process involves calculating a first ratio between the first paired value and the second quantity, setting this ratio as the first probability. Then, a second ratio between the second paired value and the third quantity is calculated, and this ratio is set as the second probability. After obtaining the second probability and the first probability, the ratio between the second probability and the first probability is calculated. Finally, the sample entropy feature can be obtained based on this ratio. The specific formula for calculating the sample entropy feature is as follows: ; Where S is the sample entropy feature, A is the second probability, and B is the first probability.

[0055] As can be seen, in this example, the chaos and unpredictability of the subject's action trajectory are determined by the sample entropy feature, which improves the accuracy of subsequent evaluation.

[0056] In one possible example, please see Figure 6 , Figure 6 A flowchart illustrating the method for determining fractal dimension features provided in this application embodiment is shown below. Figure 6 As shown, determining the fractal dimension features based on the target amplitude sequence includes: S601, according to each of the preset multiple sampling intervals, the vector net amplitude is extracted sequentially from different starting positions of the target amplitude sequence to obtain multiple sub-sequences corresponding to each sampling interval.

[0057] The self-similarity features of acceleration trajectories at different scales can be calculated using the Higuchi Fractal Dimension (HFD) algorithm, thus characterizing the geometric complexity of the action sequence. Multiple sampling intervals are pre-set. Based on each of these pre-set intervals, the corresponding net vector amplitude is sequentially extracted from different starting positions of the target amplitude sequence, obtaining multiple sub-sequences for each sampling interval. For example, if the total number of net vector amplitudes in the target amplitude sequence is N, multiple sub-sequences are constructed based on each sampling interval k, where k is less than N and k and N are positive integers. The starting sampling point corresponding to sampling interval k is m, resulting in the following sub-sequences: Here, m can be set to 1, 2, 3, ..., k. For example, if m=1 and k is 3, then 1, 4, 7, ... will be extracted sequentially from the target amplitude sequence. The net amplitude vectors form a subsequence, which is... Different values ​​of m result in different subsequences, thus obtaining multiple subsequences corresponding to each sampling interval.

[0058] S602, calculate the curve length corresponding to each of the subsequences.

[0059] The process involves calculating the curve length for each subsequence, and then normalizing the curve length of each subsequence based on the total net amplitude of the target amplitude sequence and the sampling interval corresponding to each subsequence. The normalized curve length is calculated as follows: the absolute value of the difference between the net amplitudes of all adjacent vectors within the subsequence is calculated and summed. This sum is then divided by the corresponding sampling interval k and multiplied by the normalization correction coefficient to obtain the normalized curve length of the subsequence. The specific formula is as follows: ; in, Let be the curve length of the subsequence with sampling starting point m, where m is the sampling starting point and i is the summation variable, where i = 1, 2, 3, ... ,in If the value is a decimal, it is rounded down. k is the sampling interval, and N is the total net amplitude of the target amplitude sequence. For normalization correction coefficients, Characterization represents the pair The result of the operation is rounded down. If the result is a decimal, the largest integer less than that decimal is taken; if the result is an integer, the original value is retained.

[0060] S603, calculate the average curve length of all subsequences corresponding to each sampling interval to obtain the average curve length corresponding to each sampling interval.

[0061] The average curve length of all subsequences corresponding to each sampling interval is calculated using the following formula: ; in, The average curve length, is the equivalent curve length with m as the starting sampling point within multiple subsequences, and k is the sampling interval.

[0062] S604, calculate the first logarithm of the length of each of the average curves and the second logarithm of the reciprocal of the corresponding sampling interval.

[0063] Specifically, the first logarithm of the average curve length corresponding to each sampling interval is calculated, along with the second logarithm of the reciprocal of the corresponding sampling interval. That is, the natural logarithm of each average curve length is calculated. Set the natural logarithm as the first logarithm value, and calculate the natural logarithm of the reciprocal of the sampling interval corresponding to the length of each average curve. Set the natural logarithm as the second logarithm value.

[0064] S605, Perform linear regression fitting on the first logarithmic value and the second logarithmic value to obtain the regression line.

[0065] In this process, linear regression fitting, such as least squares, is performed on each first logarithmic value and its corresponding second logarithmic value to obtain the regression line.

[0066] S606, Determine the fractal dimension feature based on the slope of the regression line.

[0067] Finally, since the average curve length and the sampling interval satisfy a power-law relationship, the fractal dimension feature is determined based on the slope of the regression line.

[0068] As can be seen, in this example, by quantifying the acceleration trajectory of the test user, the microscopic disorder characteristics of the test user's gait can be accurately reflected, thereby improving the accuracy of recognition.

[0069] In one possible example, determining the motion coordination disorder assessment result of the subject user based on the sample entropy feature, the fractal dimension feature, the linear energy feature, and a preset motion coordination disorder identification model includes: normalizing the sample entropy feature, the fractal dimension feature, and the linear energy feature corresponding to each of the target amplitude sequences to obtain a feature vector corresponding to each of the target amplitude sequences; inputting each of the feature vectors into the preset motion coordination disorder identification model to obtain a reference risk score corresponding to each of the target amplitude sequences; counting the first number of target amplitude sequences in each of the reference risk scores that exceed a preset score threshold; calculating the proportion of the first number to the total number of all target amplitude sequences; and comparing the proportion with a preset judgment threshold to obtain the motion coordination disorder assessment result of the subject user.

[0070] In a specific example, the Z-score normalization method is used to normalize the sample entropy features, fractal dimension features, and linear energy features. This normalization process involves constructing a first feature set based on the sample entropy features corresponding to the target amplitude sequence, a second feature set based on the fractal dimension features corresponding to the target amplitude sequence, and a third feature set based on the linear energy features corresponding to the target amplitude sequence. The baseline mean and standard deviation for each feature set are calculated. For each feature within each feature set, the corresponding baseline mean is subtracted from the feature, and then divided by the corresponding baseline standard deviation to obtain the normalized dimensionless features of each feature. The specific formula is as follows: ; in, Let X be the dimensionless feature corresponding to the feature in the nth feature set, where n = 1, 2, or 3. n For features within the nth feature set, The benchmark mean corresponding to the nth feature set; Let be the benchmark standard deviation corresponding to the nth feature set.

[0071] The dimensionless eigenvalues ​​corresponding to the same target amplitude sequence are concatenated into vectors to construct the eigenvector of the corresponding target amplitude sequence.

[0072] Then, each feature vector can be input into a preset motion coordination obstacle recognition model to obtain a reference risk score corresponding to each target amplitude sequence.

[0073] The motor coordination disorder identification model can employ a multilayer perceptron (MLP) neural network as the core classifier and use the measurement results of the Movement Assessment Battery for Children (MABC-2) clinical scale as the model's supervised labels for training. Specifically, during training, the percentile scores of the participating children in the clinical scale test are obtained. Samples with a percentile below or equal to 15% are labeled as positive samples with a risk of motor coordination disorder, i.e., labeled 1. Samples with a percentile above 15% are labeled as negative samples with a healthy label, i.e., labeled 0, where 15% is the 15th percentile of the clinical scale output results. This drives the model to learn the deep mapping relationship between sample entropy features, fractal dimension features, linear energy features, and clinical diagnostic logic.

[0074] The motion coordination disorder recognition model consists of an input layer, a hidden layer, and an output layer. The input layer receives feature vectors, which include sample entropy features, fractal dimension features, and linear energy features.

[0075] The hidden layer uses the ReLU activation function for non-linear feature extraction. The activation function formula is: ; Where H is the hidden layer output vector, W1 is the hidden layer weight matrix, h is the input feature vector, and b1 is the bias term. The hidden layer can map the input feature vector to a high-dimensional sparse space through nonlinear transformation, thereby accurately extracting the potential abnormal patterns in the micro-waveform of the action.

[0076] In the output layer, the model can use the Sigmoid activation function to map the feature output of the hidden layer to a disease probability prediction value between zero and one, thus obtaining the reference risk score corresponding to each target amplitude sequence. The calculation formula is as follows: ; in, For the reference risk score, e is a constant, W2 is the output layer weight matrix, H is the hidden layer output vector, and b2 is the output layer bias. After obtaining the reference risk score corresponding to each target amplitude sequence, the first number of target amplitude sequences exceeding the preset score threshold in each reference risk score is counted, and the total number of all target amplitude sequences is counted. The proportion of the first number to the total number is calculated, and the proportion is compared with the preset judgment threshold to obtain the assessment result of the subject's motor coordination disorder. Specifically, the age of the subject is obtained, and multiple healthy users of the same age are found based on the age. The average of the reference proportion values ​​corresponding to the multiple healthy users is calculated, and the judgment threshold is determined based on this average value. If the proportion value of the subject is greater than or equal to the preset judgment threshold, a motor coordination disorder assessment result is generated, which includes a motor quality distribution map and a score. The score is the score corresponding to the proportion value found from the preset score set, which includes different proportion values ​​and different scores. If the proportion value is less than the preset judgment threshold, a healthy motor coordination disorder assessment result is generated for the subject.

[0077] As can be seen, in this example, the judgment is based on the reference risk score of all target amplitude sequences within a preset time period, which effectively avoids instantaneous noise interference and improves the accuracy and reliability of the assessment results of motor coordination disorder.

[0078] In one possible example, the training process of the motion coordination obstacle recognition model includes: employing a binary cross-entropy loss and introducing an L2 regularization term to prevent overfitting. The specific loss function is calculated as follows: ; Where Loss is the loss function, and P is the total number of training samples. The true label of the i-th sample is either 0 or 1. Let λ be the reference risk score output by the model for the i-th sample, and λ be the L2 regularization coefficient used to control the regularization strength, for example, set to 0.001. This represents the sum of squares of all parameters in the model's weight matrix.

[0079] The Adam optimizer (Adaptive Moment Estimation) is used to dynamically update the model's weight matrix W and bias b based on the gradient of the loss function through the backpropagation algorithm until the loss function converges or the preset maximum number of iterations is reached.

[0080] This application embodiment can divide the electronic device into functional units according to the above method example. For example, each function can be divided into a separate functional unit, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0081] When dividing functional modules according to their respective functions, the following is combined with... Figure 7 The assessment device for motor coordination disorder in the embodiments of this application will be described in detail. Figure 7 A functional unit block diagram of an assessment device for motor coordination disorder provided in this application embodiment. An assessment device for motor coordination disorder includes: The first acquisition unit 701 is used to acquire the original signal of the test user within a preset time period, wherein the original signal is the original acceleration waveform signal obtained by acquiring a preset sampling frequency. The second acquisition unit 702 is used to determine multiple target amplitude sequences that meet preset criteria based on a preset sliding window and the original signal; The calculation unit 703 is used to calculate the sample entropy features, fractal dimension features, and linear energy features corresponding to each of the target amplitude sequences; The determining unit 704 is used to determine the assessment result of the subject's motion coordination disorder based on the sample entropy feature, the fractal dimension feature, the linear energy feature and the preset motion coordination disorder identification model.

[0082] In one possible example, the preset criteria include an intensity criterion, a periodic criterion, and a steady-state criterion. The second acquisition unit 702 is further configured to: determine a first amplitude sequence based on a preset vector net amplitude calculation formula and the original signal; scan the first amplitude sequence according to a sliding window with a preset window length and a preset step size to obtain a plurality of second amplitude sequences; and determine whether each of the second amplitude sequences meets the preset criteria, including: if the mean of the second amplitude sequence is within a preset intensity threshold range, then the second amplitude sequence is determined to meet the intensity criterion; if the periodic feature of the second amplitude sequence meets a preset periodic feature, then the second amplitude sequence is determined to meet the periodic criterion; calculate the ratio between the standard deviation and the mean of the second amplitude sequence to obtain the coefficient of variation corresponding to the second amplitude sequence; if the coefficient of variation is lower than a preset steady-state threshold, then the second amplitude sequence is determined to meet the steady-state criterion; and determine the second amplitude sequences that meet the intensity criterion, the periodic criterion, and the steady-state criterion among the various second amplitude sequences as the plurality of target amplitude sequences.

[0083] In one possible example, the second acquisition unit 702 is further configured to: convert the original signal into a time-series signal in units of gravitational acceleration; and perform bandpass filtering on the time-series signal according to preset low-frequency cutoff thresholds and high-frequency cutoff thresholds to obtain a denoised time-series signal; and calculate the vector net amplitude of each signal in the denoised time-series signal based on the preset vector net amplitude calculation formula, and remove the static gravity component to obtain the first amplitude sequence composed of the vector net amplitudes after removing the static gravity component.

[0084] In one possible example, the second acquisition unit 702 is further configured to: perform autocorrelation operation on the second amplitude sequence based on an autocorrelation function to obtain an autocorrelation sequence, the autocorrelation sequence including multiple autocorrelation function values, each autocorrelation function value corresponding to a time delay, the autocorrelation sequence characterizing the similarity between the second amplitude sequence and itself at each time delay; determine the main peak position and the main peak value corresponding to the main peak position within a preset time delay range of the autocorrelation sequence; determine whether the main peak value is greater than or equal to a preset main peak value; determine the walking frequency corresponding to the second amplitude sequence based on the main peak position and the preset sampling frequency; determine whether the walking frequency belongs to a preset step frequency range; and determine that the second amplitude sequence satisfies the periodicity standard when the main peak value is greater than or equal to the preset main peak value and the walking frequency belongs to the preset step frequency range.

[0085] In one possible example, the calculation unit 703 is further configured to: for each of the target amplitude sequences, perform the following steps: obtaining the net amplitude of each target vector in the target amplitude sequence; squaring each of the net amplitudes of the target vectors to obtain multiple squared values; calculating the arithmetic mean of the multiple squared values; taking the square root of the arithmetic mean to obtain the root mean square value corresponding to the target amplitude sequence; determining the linear energy feature based on the root mean square value; determining the sample entropy feature based on the target amplitude sequence; and determining the fractal dimension feature based on the target amplitude sequence.

[0086] In one possible example, the calculation unit 703 is further configured to: obtain a plurality of first template vectors from the target amplitude sequence according to a first preset length; calculate a first Chebyshev distance between any two first template vectors among the plurality of first template vectors, and count a first pairing value where the first Chebyshev distance is less than a preset similarity tolerance; obtain a plurality of second template vectors from the target amplitude sequence according to a second preset length, wherein the second preset length is greater than the first preset length; calculate a second Chebyshev distance between any two second template vectors among the plurality of second template vectors, and count a second pairing value where the second Chebyshev distance is less than the preset similarity tolerance; and determine the sample entropy feature based on the second pairing value and the first pairing value.

[0087] The calculation unit 703 is further configured to: sequentially extract vector net amplitudes from different starting positions of the target amplitude sequence according to each of the preset multiple sampling intervals, to obtain multiple sub-sequences corresponding to each sampling interval; calculate the curve length corresponding to each sub-sequence; calculate the average curve length of all sub-sequences corresponding to each sampling interval to obtain the average curve length corresponding to each sampling interval; calculate the first logarithm of each average curve length and the second logarithm of the corresponding sampling interval; perform linear regression fitting on the first logarithm and the second logarithm to obtain a regression line; and determine the fractal dimension feature based on the slope of the regression line.

[0088] In one possible example, the determining unit 704 is further configured to: normalize the sample entropy feature, fractal dimension feature, and linear energy feature corresponding to each of the target amplitude sequences to obtain feature vectors corresponding to each of the target amplitude sequences; input each of the feature vectors into a preset motion coordination disorder recognition model to obtain a reference risk score corresponding to each of the target amplitude sequences; count the first number of target amplitude sequences in each of the reference risk scores that exceeds a preset score threshold; calculate the proportion of the first number to the total number of all target amplitude sequences; and compare the proportion with a preset judgment threshold to obtain the motion coordination disorder assessment result of the test user.

[0089] Please see Figure 8 , Figure 8 This is a schematic diagram of the structure of a server provided in an embodiment of this application. Figure 8As shown, the server includes a processor 801, a memory 803, a communication module 802, and a program 804. The processor 801 is communicatively connected to the memory 803 and the communication module 802 via an internal communication bus. It is understood that the number of processors 801 can be set according to actual needs, for example, one, two, or more; the specific number of processors 801 is not limited here. The number of programs 804 can also be set according to actual needs, for example, one, two, or more; the specific number of programs 804 is not limited here.

[0090] The program 804 is stored in the memory 803 and executed by the processor 801. The program 804 includes instructions for performing any step in the above method embodiments.

[0091] The processor 801 may be, for example, a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, units, and circuits described in conjunction with the disclosure of this application. The processor 801 may also be a combination that implements computing functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc. The communication unit may be a communication module 802, a transceiver, a transceiver circuit, etc., and the storage unit may be a memory 803.

[0092] The memory 803 can be volatile memory or non-volatile memory, or it can include both. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct RAM (DR RAM).

[0093] This application also provides an electronic device, including a processor and a memory, wherein the memory is used to store computer program code, the computer program code including computer instructions, and the electronic device performs the method described above when the processor executes the computer instructions.

[0094] This application also provides a computer-readable storage medium storing a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method described above.

[0095] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0096] This application also provides a computer storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods described in the above method embodiments, wherein the computer includes an electronic device.

[0097] This application also provides a computer program product, which includes a computer program operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments.

[0098] The computer program product may be a software installation package, and the aforementioned computer includes electronic devices.

[0099] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0100] In the several embodiments provided in this application, it should be understood that the disclosed methods, apparatuses, and systems can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for example, the division of units is merely a logical functional division, and other division methods may exist in actual implementation; for example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0101] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0102] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can be physically comprised separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware or in the form of hardware plus software functional units.

[0103] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute some steps of the methods described in the various embodiments of the present 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.

[0104] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can easily conceive of variations or substitutions without departing from the spirit and scope of the present invention, and various modifications and alterations can be made, including combinations of the different functions and implementation steps described above, as well as software and hardware implementation methods, all of which are within the protection scope of the present invention.

Claims

1. A method for assessing motor coordination disorder, characterized in that, include: Acquire the raw signal of the subject within a preset time period, wherein the raw signal is the original acceleration waveform signal obtained by sampling at a preset sampling frequency; Based on a preset sliding window and the original signal, multiple target amplitude sequences that meet preset criteria are determined; Calculate the sample entropy feature, fractal dimension feature, and linear energy feature corresponding to each of the target amplitude sequences; The assessment results of the subject's motor coordination disorder are determined based on the sample entropy features, the fractal dimension features, the linear energy features, and the preset motor coordination disorder identification model.

2. The method according to claim 1, characterized in that, The preset criteria include intensity criteria, periodic criteria, and steady-state criteria. The process of determining multiple target amplitude sequences that satisfy the preset criteria based on a preset sliding window and the original signal includes: The first amplitude sequence is determined based on the preset vector net amplitude calculation formula and the original signal; The first amplitude sequence is scanned by the sliding window with a preset window length and a preset step size to obtain multiple second amplitude sequences; Determining whether each of the second amplitude sequences meets the preset criteria includes: If the mean of the second amplitude sequence falls within the preset intensity threshold range, then the second amplitude sequence is determined to meet the intensity standard. If the periodicity of the second amplitude sequence satisfies the preset periodicity, then the second amplitude sequence is judged to meet the periodicity standard. Calculate the ratio between the standard deviation and the mean of the second amplitude sequence to obtain the coefficient of variation corresponding to the second amplitude sequence; If the coefficient of variation is lower than a preset steady-state threshold, then the second amplitude sequence is determined to meet the steady-state criterion. The second amplitude sequences that satisfy the intensity standard, the period standard, and the steady-state standard in each of the second amplitude sequences are determined as the plurality of target amplitude sequences.

3. The method according to claim 2, characterized in that, The determination of the first amplitude sequence based on the preset vector net amplitude calculation formula and the original signal includes: The original signal is converted into a time-series signal in units of gravitational acceleration; Based on preset low-frequency cutoff thresholds and high-frequency cutoff thresholds, the timing signal is subjected to bandpass filtering to obtain a denoised timing signal. Based on the preset vector net amplitude calculation formula, the vector net amplitude of each signal in the denoised time-series signal is calculated, and the static gravity component is removed to obtain the first amplitude sequence composed of the vector net amplitudes after removing the static gravity component.

4. The method according to claim 2, characterized in that, The step of determining that the second amplitude sequence meets the periodic standard if the periodic characteristics of the second amplitude sequence satisfy the preset periodic characteristics includes: Autocorrelation operation is performed on the second amplitude sequence based on the autocorrelation function to obtain an autocorrelation sequence. The autocorrelation sequence includes multiple autocorrelation function values, and each autocorrelation function value corresponds to a time delay. The autocorrelation sequence characterizes the similarity between the second amplitude sequence and itself at each time delay. Within a preset time delay range of the autocorrelation sequence, the position of the main peak and the corresponding peak value are determined. Determine whether the main peak value is greater than or equal to the preset main peak value; Based on the main peak position and the preset sampling frequency, the walking frequency corresponding to the second amplitude sequence is determined; Determine whether the walking frequency falls within a preset step frequency range; When the main peak value is greater than or equal to the preset main peak value, and the walking frequency belongs to the preset step frequency range, the second amplitude sequence is determined to meet the period standard.

5. The method according to any one of claims 1-4, characterized in that, The calculation of the sample entropy feature, fractal dimension feature, and linear energy feature corresponding to each of the target amplitude sequences includes: For each of the target amplitude sequences, the following steps are performed: Obtain the net magnitude of each target vector in the target magnitude sequence; Squaring the net magnitude of each of the target vectors yields multiple squared values. Calculate the arithmetic mean of the plurality of squared values; The square root of the arithmetic mean is taken to obtain the root mean square value corresponding to the target amplitude sequence. The linear energy characteristics are determined based on the root mean square value; Determine the sample entropy characteristics based on the target amplitude sequence; The fractal dimension features are determined based on the target amplitude sequence.

6. The method according to claim 5, characterized in that, The step of determining the sample entropy features based on the target amplitude sequence includes: Multiple first template vectors are obtained from the target amplitude sequence according to a first preset length; Calculate the first Chebyshev distance between any two first template vectors among the plurality of first template vectors, and count the first pairing values ​​where the first Chebyshev distance is less than the preset similarity tolerance; Multiple second template vectors are obtained from the target amplitude sequence according to a second preset length, wherein the second preset length is greater than the first preset length; Calculate the second Chebyshev distance between any two second template vectors among the plurality of second template vectors, and count the second pairing values ​​where the second Chebyshev distance is less than the preset similarity tolerance; The sample entropy feature is determined based on the second pairing value and the first pairing value.

7. The method according to claim 5, characterized in that, Determining the fractal dimension features based on the target amplitude sequence includes: According to the preset multiple sampling intervals, the vector net amplitude is extracted sequentially from different starting positions of the target amplitude sequence to obtain multiple sub-sequences corresponding to each sampling interval; Calculate the curve length corresponding to each of the subsequences; Calculate the average curve length of all subsequences corresponding to each sampling interval to obtain the average curve length corresponding to each sampling interval; Calculate the first logarithm of the length of each of the average curves, and the second logarithm of the reciprocal of the corresponding sampling interval; A linear regression is performed on the first logarithmic value and the second logarithmic value to obtain a regression line; The fractal dimension feature is determined based on the slope of the regression line.

8. The method according to any one of claims 1-4, characterized in that, The step of determining the assessment result of the subject's motor coordination disorder based on the sample entropy feature, the fractal dimension feature, the linear energy feature, and the preset motor coordination disorder recognition model includes: The sample entropy feature, fractal dimension feature, and linear energy feature corresponding to each of the target amplitude sequences are normalized to obtain the feature vector corresponding to each of the target amplitude sequences. Each of the aforementioned feature vectors is input into the preset motion coordination obstacle recognition model to obtain a reference risk score corresponding to each of the target amplitude sequences; Count the first number of target amplitude sequences that exceed a preset scoring threshold in each of the aforementioned reference risk scores; Calculate the proportion of the first quantity to the total quantity of all the target amplitude sequences; The ratio value is compared with a preset judgment threshold to obtain the assessment result of the subject's motor coordination disorder.

9. An assessment device for motor coordination disorder, characterized in that, include: The first acquisition unit is used to acquire the original signal of the test user within a preset time period, wherein the original signal is the original acceleration waveform signal obtained by acquiring a preset sampling frequency. The second acquisition unit is used to determine multiple target amplitude sequences that meet preset criteria based on a preset sliding window and the original signal. The calculation unit is used to calculate the sample entropy features, fractal dimension features, and linear energy features corresponding to each of the target amplitude sequences; The determining unit is used to determine the assessment result of the subject's motor coordination disorder based on the sample entropy feature, the fractal dimension feature, the linear energy feature, and the preset motor coordination disorder identification model.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described in any one of claims 1-8.

Images