A multi-dimensional ankle stability training evaluation method and system

Abstract

The application relates to the technical field of rehabilitation evaluation, and discloses a multi-dimensional ankle joint stability training evaluation method and system. The method comprises the following steps: establishing a sensor coordinate conversion matrix through standard mass block calibration, collecting six-dimensional biomechanical data of a subject, calculating a force-angle coupling response coefficient according to eight-sector division to determine a weak direction, and combining time-frequency domain feature extraction and direction imbalance index calculation to obtain a comprehensive stability score. The application improves the multi-dimensional quantitative precision of ankle joint stability evaluation and the pertinence of individualized training scheme formulation.

Description

Technical Field

[0001] This application relates to the field of rehabilitation assessment technology, and in particular to a multidimensional ankle joint stability training assessment method and system. Background Technology

[0002] The ankle joint, as a key hub in the lower limb kinetic chain, directly impacts standing balance, gait control, and the risk of sports injuries. Traditional ankle stability assessment methods primarily rely on clinical scale scores and simple single-leg standing tests. In recent years, with the development of biomechanical measurement technology, force plate-based quantitative assessment methods have been increasingly applied in clinical and sports rehabilitation fields. Existing ankle stability assessment devices typically employ fixed force plates combined with pressure sensors to evaluate overall balance ability by measuring parameters such as trajectory length, swing area, and average velocity of the pressure center trajectory during static standing or dynamic balance. Some advanced systems incorporate unstable support surface designs, such as swing platforms or air cushion devices, to increase the difficulty of balance challenges and test the subject's dynamic stability, using force sensors to collect and analyze plantar pressure distribution data. These methods, to a certain extent, achieve objective quantitative assessment of ankle stability, providing data support for clinical diagnosis and rehabilitation training.

[0003] However, existing ankle stability assessment methods have significant technical limitations. First, the spatial-temporal synchronization of sensor systems is insufficient. In existing devices, force and angle sensors are typically designed and calibrated independently, lacking a unified spatial coordinate system and a precise timestamp alignment mechanism. This leads to an ambiguous correspondence between pressure and angle data in the spatiotemporal dimensions, failing to accurately capture the dynamic coupling relationship between the movement of the pressure center and the tilt of the support surface. This directly affects the accuracy of analyzing real-time ankle control strategies. Second, the comprehensiveness of assessment dimensions is severely lacking. Most systems only focus on the two-dimensional movement trajectory of the pressure center in the horizontal plane, failing to effectively integrate multi-dimensional biomechanical parameters such as pedal inclination angle, plantar pressure distribution symmetry, and vertical load changes at the center of gravity. This results in assessment results that only reflect planar balance ability and cannot comprehensively characterize the ankle's stability control mechanism in three-dimensional space. Third, directional diagnostic capabilities are lacking. Existing methods mainly judge stability levels through simple statistics of the pressure center offset amplitude, lacking quantitative analysis of differences in stability ability in different directions. They cannot identify weaknesses in specific directions such as anterior-posterior, lateral, and oblique, making it difficult to provide targeted guidance for the development of personalized training programs. Summary of the Invention

[0004] This application provides a multi-dimensional ankle joint stability training assessment method and system, which solves the problems of inaccurate correspondence of multi-sensor data, weak directionality and lack of diagnostic ability in the prior art by establishing a sensor space-time synchronous calibration mechanism and a force-angle spatiotemporal coupling analysis model, thereby improving the multi-dimensional quantitative accuracy of ankle joint stability assessment and the pertinence of personalized training program formulation.

[0005] Firstly, this application provides a multi-dimensional ankle joint stability training assessment method, which includes:

[0006] Step S1: Place the standard mass block at the preset calibration point of the multi-axis pedal device, collect the output value of the force sensor array and the reading of the angle sensor, and establish the pressure center coordinate transformation matrix and the tilt angle correction function;

[0007] Step S2: The subject stands on the multi-axis pedal device. The output value of the force sensor array and the reading of the angle sensor are processed synchronously by the pressure center coordinate transformation matrix and the tilt angle correction function to generate a six-dimensional raw data matrix containing the pressure center displacement sequence and the pedal tilt angle sequence.

[0008] Step S3: Divide the pressure center displacement sequence and the pedal tilt angle sequence in the six-dimensional original data matrix into eight sector directions, calculate the force-angle coupling response coefficient of each sector, and determine the main weak direction from the three sectors with the largest force-angle coupling response coefficient;

[0009] Step S4: Extract time-frequency domain features from the six-dimensional original data matrix to obtain the trajectory length and frequency domain power ratio. Combine the directional imbalance index corresponding to the main weak direction with the comprehensive stability score calculated according to the hierarchical scoring function.

[0010] Secondly, this application provides a multi-dimensional ankle joint stability training and assessment system, the multi-dimensional ankle joint stability training and assessment system comprising:

[0011] The data acquisition module is used to place a standard mass block at the preset calibration point of the multi-axis pedal device, acquire the output values ​​of the force sensor array and the readings of the angle sensor, and establish the pressure center coordinate transformation matrix and the tilt angle correction function.

[0012] The processing module is used to process the output values ​​of the force sensor array and the readings of the angle sensor simultaneously by the pressure center coordinate transformation matrix and the tilt angle correction function when the subject stands on the multi-axis pedal device, thereby generating a six-dimensional raw data matrix containing the pressure center displacement sequence and the pedal tilt angle sequence.

[0013] The partitioning module is used to divide the pressure center displacement sequence and the pedal tilt angle sequence in the six-dimensional original data matrix into eight sector directions, calculate the force-angle coupling response coefficient of each sector, and determine the main weak direction by the three sectors with the largest force-angle coupling response coefficient.

[0014] The extraction module is used to extract time-frequency domain features from the six-dimensional original data matrix to obtain the trajectory length and frequency domain power ratio, and calculate the comprehensive stability score according to the hierarchical scoring function in combination with the directional imbalance index corresponding to the main weak direction.

[0015] The technical solution provided in this application fundamentally solves the technical deficiency in existing technologies where the output values ​​of force sensor arrays and the readings of angle sensors lack a unified correspondence in the spatial-temporal dimensions by implementing a standard mass block calibration procedure on a multi-axis pedal device and establishing a pressure center coordinate transformation matrix and an inclination correction function. This calibration mechanism incorporates dispersed sensor measurement data into a unified spatial rectangular coordinate system, establishes a precise linear mapping from the measurement unit output value to the pressure center coordinates through a torque balance equation, and constructs a systematic error correction function by fitting the theoretical inclination angle with the actual reading. This ensures that the subsequent real-time acquisition of multi-dimensional biomechanical signals possesses strict spatiotemporal synchronization and measurement accuracy. Based on this, during the subject's standing test, the sensor data is synchronously processed through the transformation matrix and correction function to generate a six-dimensional raw data matrix containing a pressure center displacement sequence and a pedal inclination angle sequence. This data matrix not only covers multiple biomechanical dimensions such as planar displacement, spatial inclination angle, vertical pressure, and bilateral symmetry, but more importantly, it ensures the precise correspondence of data in each dimension at each sampling moment, laying a reliable data foundation for subsequent force-angle coupling relationship analysis. This application innovatively divides the pressure center displacement sequence and pedal tilt angle sequence in the six-dimensional original data matrix into eight sectors. By calculating the force-angle coupling response coefficient of each sector, a quantitative correlation model is established to determine the degree of pedal tilt in a specific direction when the pressure center points to that direction. The physical meaning of this model clearly reflects the ankle joint's anti-overturning ability in different directions. The diagnostic method of determining the main weak direction by the three sectors with the largest coupling response coefficients breaks through the limitation of existing technologies that only evaluate the overall situation based on the pressure center offset amplitude and cannot identify directional defects. This provides a precise diagnostic basis for targeted strengthening of weak directions in personalized rehabilitation training. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments 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 based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of one embodiment of the multidimensional ankle joint stability training and assessment method in this application.

[0018] Figure 2 This is a schematic diagram illustrating the temporal changes of each dimension of the six-dimensional original data matrix in the embodiments of this application;

[0019] Figure 3 This is a schematic diagram showing the force-angle cross-correlation function curves and response time delay characteristics of different sectors in the embodiments of this application;

[0020] Figure 4 This is a schematic diagram of the structure of the multi-dimensional ankle joint stability training and assessment device in the embodiments of this application;

[0021] Figure descriptions: 401, square anti-slip pedal; 402, front and rear tilting axis; 403, left and right swinging axis; 404, force sensor array; 405, angle sensor; 406, metal support base; 407, shaft damping structure; 408, preset calibration point; 409, data transmission interface; 410, power interface. Detailed Implementation

[0022] This application provides a multi-dimensional ankle joint stability training and assessment method and system. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data used can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0023] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the multidimensional ankle joint stability training and assessment method in this application includes:

[0024] Step S1: Place the standard mass block at the preset calibration point 408 of the multi-axis pedal device, collect the output value of the force sensor array 404 and the reading of the angle sensor 405, and establish the pressure center coordinate transformation matrix and tilt angle correction function;

[0025] Specifically, a precise mapping relationship between sensor data and physical quantities is established through a calibration procedure. The multi-axis pedal device includes a front-to-back tilt axis 402 and a left-to-right swing axis 403. The intersection of the two axes serves as the origin of the spatial coordinate system. The force sensor array 404 is arranged in a 20-point measurement matrix according to the heel area, arch area, first metatarsal area, fifth metatarsal area, and toe area. When a standard mass block is placed on the pedal surface at 9 calibration points distributed in a 3×3 grid, there is a torque balance relationship between the output value of each force sensor and the position of the pressure center. A transformation matrix from output value to coordinate is established through matrix operations. The deviation between the reading of the angle sensor 405 and the actual tilt angle is eliminated by a correction function to eliminate systematic errors.

[0026] Step S2: The subject stands on the multi-axis pedal device. The output value of the force sensor array 404 and the reading of the angle sensor 405 are processed synchronously through the pressure center coordinate transformation matrix and the tilt angle correction function to generate a six-dimensional raw data matrix containing the pressure center displacement sequence and the pedal tilt angle sequence.

[0027] Specifically, the subject stands behind the pedal, and the force sensor array 404 continuously outputs 20 pressure signals at a frequency of 200 Hz, while the angle sensor 405 simultaneously outputs tilt angle signals along the front-to-back and left-to-right axes. The output values ​​of the 20 force sensors at each sampling moment are substituted into the transformation matrix established in step S1, and the X and Y coordinates of the pressure center at that moment are calculated through weighted summation, forming a pressure center displacement sequence. The raw readings of the angle sensor 405 are processed by a correction function to obtain the true tilt angle value, forming the pedal tilt angle sequence. Simultaneously, the sum of the output values ​​of the 20 force sensors is calculated to obtain the total plantar pressure sequence, and the difference between the outputs of the 10 sensors on the left and 10 sensors on the right, divided by the sum, yields the pressure distribution asymmetry index sequence. The six-dimensional data sequences each contain 12,000 sampling points during the 60-second test, combined to form a six-dimensional raw data matrix.

[0028] Step S3: Divide the pressure center displacement sequence and pedal tilt angle sequence in the six-dimensional original data matrix into eight sectors, calculate the force-angle coupling response coefficient of each sector, and determine the main weak direction from the three sectors with the largest force-angle coupling response coefficient;

[0029] Specifically, the tread plate is divided into eight sectors at 45-degree intervals centered on the origin, with each sector corresponding to a principal directional angle. The X and Y coordinates of the pressure center displacement sequence are used to calculate the pressure center direction angle using the arctangent function. The number of points out of 12,000 sampling points whose direction angles fall within the angle range of each sector is counted, and divided by the total number of points to obtain the time percentage. The forward and backward tilt angles and left and right tilt angles of the tread plate tilt angle sequence are also obtained by using the arctangent function to obtain the tilt direction angle. When the tilt direction angle points to a certain sector, the tilt angle amplitude at that moment is accumulated to the cumulative angle amplitude of that sector. The force-angle coupling response coefficient is defined as the cumulative angle amplitude divided by the time percentage and then divided by the test duration. Physically, it represents the average angular velocity of the tread plate tilting in a certain direction when the pressure center remains in that direction per unit time. The larger the value, the weaker the anti-overturning ability in that direction. The coupling response coefficients of the eight sectors are arranged in descending order of value, and the center direction angles of the first three sectors are taken as the main weak directions. The standard deviation of the coupling response coefficients of the eight sectors is divided by the mean to obtain the directional imbalance index.

[0030] Step S4: Extract time-frequency domain features from the six-dimensional original data matrix to obtain the trajectory length and frequency domain power ratio. Combine the directional imbalance index corresponding to the main weak directions and calculate the comprehensive stability score according to the hierarchical scoring function.

[0031] Specifically, the instantaneous displacement increment is obtained by taking the square root of the sum of the squared differences in the X and Y coordinates between adjacent sampling points in the pressure center displacement sequence. The trajectory length is obtained by accumulating 11,999 instantaneous displacement increments. The X and Y coordinate sequences of the pressure center displacement sequence are subjected to Fast Fourier Transform to obtain the frequency domain power spectrum. The frequency range of 0 to 10 Hz is divided into a low-frequency band (0.1 to 0.5 Hz), a mid-frequency band (0.5 to 2 Hz), and a high-frequency band (2 to 10 Hz). The power integral value of each frequency band is divided by the total power to obtain the frequency domain power ratio. The static control score is calculated using a piecewise function based on the area of ​​the 95% confidence ellipse of the pressure center. The dynamic adjustment score is calculated using a piecewise function based on the average velocity obtained by dividing the trajectory length by the test duration. The sensory integration score is calculated based on the deviation of the low-frequency power ratio from the ideal value of 0.5. The comprehensive stability score is calculated by weighting the static control score, dynamic adjustment score, and sensory integration score with weights of 0.2, 0.2, and 0.15, respectively, and then subtracting the directional imbalance index multiplied by a penalty coefficient of 50 to obtain the directional penalty factor. The score range is 0 to 100 points, which realizes the quantitative grading assessment of ankle joint stability.

[0032] In this application, a comprehensive quantitative evaluation system from the underlying control mechanism to overall performance is constructed by extracting time-frequency domain features from a six-dimensional original data matrix to obtain multi-level parameters such as trajectory length and frequency domain power ratio. This is combined with a technical scheme that calculates the comprehensive stability score using a hierarchical scoring function based on the directional imbalance index corresponding to the main weak directions. The time-frequency domain feature extraction calculates the trajectory length, reflecting the dynamic regulation activity, by accumulating the Euclidean distance between adjacent sampling points. Principal component analysis extracts the elliptical swing area, reflecting the static control boundary. Fast Fourier Transform converts the time-domain signal into a frequency-domain power spectrum and divides it into low-frequency, mid-frequency, and high-frequency bands to calculate the power ratio. This frequency-domain analysis technique can distinguish the relative contributions of slow regulation dominated by the visual-vestibular system and fast feedback dominated by proprioception in balance control, revealing the efficiency level of sensory system integration and significantly improving the interpretability of the evaluation results for neural control strategies. The stratified scoring function calculates static control score, dynamic adjustment score, and sensory integration score based on parameters such as swing area, trajectory length, and frequency domain power ratio. The scores of each sub-item are weighted and summed, and then the directional penalty factor calculated by the directional imbalance index is subtracted to obtain a comprehensive stability score ranging from 0 to 100. This scoring system not only achieves quantitative grading of ankle joint stability, but also intuitively identifies the specific ability deficiencies of the subjects through radar charts of multiple sub-item scores. Incorporating the directional imbalance index as a penalty factor into the comprehensive score highlights the negative impact of directional stability defects on overall stability, guiding clinicians and rehabilitation therapists to focus on targeted training of directional weaknesses, thereby improving the practical guiding value of the assessment results for clinical decision-making and training program development.

[0033] In one specific embodiment, step S1 includes:

[0034] A spatial rectangular coordinate system is established with the intersection of the front and rear tilting axis 402 and the left and right swinging axis 403 of the multi-axis pedal device as the origin, and the planar coordinate position of the measuring unit in each force sensor array 404 is determined.

[0035] Standard mass blocks are placed sequentially on nine preset calibration points 408 on the upper surface of the multi-axis pedal device, and the output values ​​of each measurement unit and the tilt angle reading of the angle sensor 405 are recorded simultaneously to obtain the calibration dataset.

[0036] Based on the output values ​​of each measurement unit and their corresponding planar coordinate positions in the calibration dataset, a linear mapping relationship between the output values ​​of the measurement units and the coordinates of the pressure center is established through the torque balance equation, and a pressure center coordinate transformation matrix is ​​generated.

[0037] The tilt angle readings in the calibration dataset are fitted with the theoretical tilt angles generated by the standard mass block at each calibration point to construct a tilt angle error correction function as the tilt angle correction function.

[0038] Specifically, the spatial rectangular coordinate system is established with the intersection of the two axes as the origin. The X-axis extends horizontally along the left-right swing direction, the Y-axis extends horizontally along the front-back tilt direction, and the Z-axis is vertically upward. The 20 measuring units in the force sensor array 404 are distributed in five areas: the heel area, the arch area, the first metatarsal area, the fifth metatarsal area, and the toe area. The coordinate position of each measuring unit on the pedal plane is determined by measuring its X-axis and Y-axis distances relative to the origin. Nine preset calibration points 408 are evenly distributed on the upper surface of the pedal in a 3x3 grid, with an adjacent point spacing of 150 mm. When a standard mass block with a mass of 50 kg is placed sequentially at each calibration point, the 20 measuring units synchronously output force value signals, and the angle sensors 405 of the front-back tilt axis 402 and the left-right swing axis 403 synchronously output tilt angle readings. Each calibration point corresponds to a set of data containing 20 force values ​​and 2 tilt angles. The nine calibration points obtain a total of 180 force value data and 18 tilt angle data, which constitute the calibration dataset.

[0039] The torque balance equation is based on the principle of statics. The X-coordinate of the pressure center is equal to the sum of the products of the output values ​​of each measuring unit and their Y-coordinate positions, divided by the sum of the output values ​​of all measuring units. The Y-coordinate of the pressure center is equal to the sum of the products of the output values ​​of each measuring unit and their X-coordinate positions, divided by the sum of the output values ​​of all measuring units. A linear equation system is established using nine sets of known pressure center positions and corresponding measuring unit output values ​​from the calibration dataset, and the coefficients of the transformation matrix are obtained by solving. The tilt correction function is constructed by using the actual readings of the angle sensor 405 from the calibration dataset as the independent variable and the theoretical tilt angle generated by the standard mass block at each calibration point as the dependent variable. The theoretical tilt angle is calculated by dividing the torque of the mass block position relative to the origin by the product of the mass block's weight and the distance from the intersection of the two axes to the pedal surface. A polynomial fitting method is used to establish a mapping function from the reading values ​​to the true tilt angle values, eliminating the nonlinear error and zero-point drift error of the sensor.

[0040] In one specific embodiment, step S2 includes:

[0041] The subject stood on the multi-axis pedal device, and the output value of the force sensor array 404 and the reading of the angle sensor 405 were synchronously acquired at a sampling frequency of 200 Hz to obtain the raw signal sequence of 60 seconds of test duration.

[0042] Substitute the output value of the force sensor array 404 in the original signal sequence into the pressure center coordinate transformation matrix to calculate the pressure center displacement sequence, including the X-direction displacement sequence and the Y-direction displacement sequence.

[0043] Substitute the readings of angle sensor 405 in the original signal sequence into the tilt angle correction function to calculate the pedal tilt angle sequence, including the front and rear tilt angle sequence and the left and right tilt angle sequence.

[0044] The total plantar pressure sequence is obtained by summing the output values ​​of the force sensor array 404. The pressure distribution asymmetry index sequence is obtained by calculating the ratio of the pressure difference between the left and right sensor groups to the total pressure. The pressure center displacement sequence, pedal inclination sequence, total plantar pressure sequence and pressure distribution asymmetry index sequence are combined to form a six-dimensional original data matrix.

[0045] Specifically, a 200 Hz sampling frequency means acquiring 200 data points per second. A 60-second test yields 12,000 data points. At each time point, the output values ​​of 20 force sensors and the readings of 2 angle sensors 405 are recorded simultaneously. The raw signal sequence contains 12,000 sets of force values ​​and 12,000 sets of tilt angle data. The pressure center coordinate transformation matrix converts the output values ​​of the 20 force sensors at each time point into the X and Y coordinates of the pressure center at that moment. The X-direction displacement sequence consists of 12,000 X-coordinate values ​​arranged in chronological order, and the Y-direction displacement sequence consists of 12,000 Y-coordinate values ​​arranged in chronological order, reflecting the motion trajectory of the pressure center on the pedal plane. The tilt angle correction function converts the raw readings of the angle sensors 405 at each time point into corrected true tilt angle values. The forward and backward tilt angle sequence consists of 12,000 corrected tilt angle values ​​of the forward and backward tilt axes 402, and the left and right tilt angle sequence consists of 12,000 corrected tilt angle values ​​of the left and right swing axes 403, reflecting the dynamic tilt changes of the pedal in the two axes.

[0046] The total plantar pressure sequence is obtained by summing the output values ​​of 20 force sensors at each time point. The 12,000 total pressure values ​​are arranged chronologically to form a sequence reflecting the temporal variation characteristics of the vertical load on the subject's center of gravity. The pressure distribution asymmetry index is calculated by dividing the 20 force sensors into two groups based on their left and right positions. The left sensor group contains 10 measurement units located to the left of the pedal centerline, and the right sensor group contains 10 measurement units located to the right of the pedal centerline. At each time point, the sum of the output values ​​of the left and right groups is calculated separately. The difference between the two is divided by the sum of the two values ​​to obtain the asymmetry index at that time. The 12,000 asymmetry index values ​​constitute the pressure distribution asymmetry index sequence. The six-dimensional raw data matrix arranges the data in six dimensions row by row. The first row is the X-direction displacement sequence, the second row is the Y-direction displacement sequence, the third row is the fore-and-aft tilt angle sequence, the fourth row is the left-and-right tilt angle sequence, the fifth row is the total plantar pressure sequence, and the sixth row is the pressure distribution asymmetry index sequence. Each dimension contains 12,000 sampling points, and the matrix size is 6 rows and 12,000 columns.

[0047] Figure 2 This is a schematic diagram illustrating the temporal changes of each dimension of the six-dimensional original data matrix in the embodiments of this application; Figure 2This study showcases the raw data sequence changes across six dimensions of the subjects during a 60-second test. From top to bottom, the sequence includes: the X-direction displacement sequence of the pressure center, exhibiting periodic oscillations with an amplitude of approximately ±30 mm; the Y-direction displacement sequence of the pressure center, with an oscillation amplitude of approximately ±25 mm; the fore-and-aft inclination sequence of the pedal, fluctuating within ±10 degrees; the lateral inclination sequence of the pedal, with an inclination amplitude slightly smaller than that in the fore-and-aft direction; the total plantar pressure sequence, with baseline pressure maintained at approximately 500 Newtons and accompanied by periodic fluctuations; and the pressure distribution asymmetry index sequence, with values ​​varying between ±0.2, reflecting the dynamic equilibrium of bilateral pressure distribution. All six dimensions of data were recorded simultaneously at a sampling frequency of 200 Hz. The fluctuation frequency, amplitude characteristics, and temporal correspondences of the time-series curves provide a data foundation for subsequent multi-level feature extraction and weak point diagnosis.

[0048] In one specific embodiment, step S3 divides the pressure center displacement sequence and pedal tilt angle sequence in the six-dimensional original data matrix into eight sector directions, specifically:

[0049] Centered on the origin of the spatial rectangular coordinate system, the tread plate is divided into eight sectors at 45-degree intervals: front, front right, right, rear right, rear, rear left, left, and front left. The central direction angles of each sector are 0 degrees, 45 degrees, 90 degrees, 135 degrees, 180 degrees, 225 degrees, 270 degrees, and 315 degrees, respectively.

[0050] Perform arctangent operation on the X-direction displacement sequence and Y-direction displacement sequence in the pressure center displacement sequence to obtain the pressure center direction angle sequence, and count the time proportion of the pressure center direction angle entering each sector angle interval at each sampling time.

[0051] Perform arctangent calculation on the front-to-back tilt angle sequence and the left-to-right tilt angle sequence in the pedal tilt angle sequence to obtain the pedal tilt direction angle sequence. When the pedal tilt direction angle points to a certain sector, accumulate the tilt angle amplitude at the corresponding moment to obtain the cumulative angle amplitude of the corresponding sector.

[0052] Specifically, the 360-degree circle is divided into eight sectors at 45-degree intervals. The angle range of the directly front sector is -22.5 degrees to +22.5 degrees; the right-front sector is 22.5 degrees to 67.5 degrees; the right-direction sector is 67.5 degrees to 112.5 degrees; the right-rear sector is 112.5 degrees to 157.5 degrees; and the rear sector is 157.5 degrees to 202.5 degrees. The equivalent ranges are -157.5 degrees to -112.5 degrees. The left rear sector ranges from 202.5 degrees to 247.5 degrees, or equivalently from -112.5 degrees to -67.5 degrees. The directly left sector ranges from 247.5 degrees to 292.5 degrees, or equivalently from -67.5 degrees to -22.5 degrees. The left front sector ranges from 292.5 degrees to 337.5 degrees, or equivalently from -22.5 degrees to -337.5 degrees. The pressure center direction angle is obtained by performing arctangent calculations on the X-direction and Y-direction displacement sequences. The arctangent function calculates the angle corresponding to the quotient of the Y-direction displacement value divided by the X-direction displacement value. 12,000 pressure center direction angle values ​​are calculated from 12,000 sampling times, forming a direction angle sequence. The number of sampling points falling within each sector's angle interval is counted by traversing this sequence. The time percentage of a certain sector is equal to the number of sampling points in that sector divided by the total number of sampling points (12,000).

[0053] The calculation of the pedal tilt direction angle uses the forward and backward tilt angle sequence as the Y-axis component and the left and right tilt angle sequence as the X-axis component. The angle corresponding to the quotient of the forward / backward tilt angle value divided by the left / right tilt angle value is calculated using the arctangent function, resulting in a tilt direction angle sequence of 12,000 pedal tilt direction angle values. The calculation of the cumulative angle amplitude iterates through 12,000 sampling times, determining whether the tilt direction angle at each time falls within the angle range of a certain sector. When the tilt direction angle points to that sector, the tilt angle amplitude at that time is calculated. The tilt angle amplitude is defined as the square root of the sum of the squares of the forward / backward tilt angle value and the squares of the left / right tilt angle value. The cumulative angle amplitude of all tilt angle amplitudes pointing to that sector is accumulated to obtain the cumulative angle amplitude of that sector. Eight cumulative angle amplitude values ​​are calculated for each of the eight sectors. The larger the value, the more severe the cumulative tilt of the pedal in that direction.

[0054] In one specific embodiment, step S3 calculates the force-angle coupling response coefficients of each sector, specifically as follows:

[0055] Divide the cumulative angular amplitude of each sector by the product of the time proportion of the corresponding sector and the test duration to obtain the force-angle coupling response coefficient of each sector.

[0056] Extract the time intervals pointing to each sector from the pressure center displacement sequence, and calculate the radial displacement sequence of the pressure center and the tilt projection sequence of the pedal in the corresponding direction;

[0057] Cross-correlation function calculations were performed on the radial displacement sequence and tilt projection sequence of the pressure center to obtain cross-correlation coefficient curves with a time delay range of -200 milliseconds to +200 milliseconds. The time delay parameters corresponding to the peak values ​​of the cross-correlation coefficient curves were extracted as the response time delay characteristics of each sector.

[0058] Specifically, the force-angle coupling response coefficient is calculated by using the cumulative angle amplitude of a sector as the numerator, multiplying the time percentage of that sector by the test duration of 60 seconds as the denominator, and dividing the two to obtain the coupling response coefficient of that sector, expressed in degrees per second. Multiplying the time percentage by the test duration yields the cumulative time the pressure center actually remains in that sector. Dividing the cumulative angle amplitude by the cumulative time reflects the average angular velocity of the pedal tilting in that direction per unit time when the pressure center remains in that direction; a larger value indicates weaker anti-overturning control capability in that direction. The extraction of the radial displacement sequence of the pressure center involves traversing 12,000 sampling times, selecting the time points when the pressure center's directional angle enters the angle interval of a certain sector, extracting the corresponding X-direction and Y-direction displacement values ​​at these time points, and calculating the square root of the sum of their squares to obtain the radial distance of the pressure center relative to the origin. These radial distance values ​​are arranged in chronological order to constitute the radial displacement sequence of the pressure center for that sector.

[0059] The tilt projection sequence is calculated by projecting the forward and backward tilt angle sequences and the left and right tilt angle sequences onto the center direction angle of the sector. The projection value is equal to the left and right tilt angle value multiplied by the cosine of the center direction angle of the sector, plus the forward and backward tilt angle value multiplied by the sine of the center direction angle of the sector. The projection values ​​corresponding to the time points when the pressure center points to the sector are extracted and arranged in chronological order to form the tilt projection sequence. The cross-correlation function is calculated by using the radial displacement sequence of the pressure center as the first signal and the tilt projection sequence as the second signal. The time delay parameter traverses from -200 ms to +200 ms with a sampling period of 5 ms. For each time delay value, the second signal is shifted by the corresponding time, and then multiplied point by point with the first signal and summed. After normalization, the cross-correlation coefficient corresponding to that time delay is obtained. The cross-correlation coefficients corresponding to all time delay values ​​form a cross-correlation coefficient curve. The time delay parameter corresponding to the peak value of the curve represents the time delay or lead relationship between the pedal tilt and the pressure center movement. A positive time delay indicates that the tilt lags behind the pressure center movement, reflecting passive imbalance, while a negative time delay indicates that the tilt moves before the pressure center, reflecting active predictive control.

[0060] Figure 3 This is a schematic diagram showing the force-angle cross-correlation function curves and response time delay characteristics of different sectors in the embodiments of this application; Figure 3The cross-correlation function curves between the radial displacement of the pressure center and the projection of the pedal tilt angle are displayed for three typical sectors. The right rear sector (solid line) is the primary weak direction, with a peak cross-correlation coefficient of 0.65 occurring at a time lag of +25 milliseconds. The positive time lag indicates that the pedal tilt lags behind the movement of the pressure center, reflecting the passive imbalance characteristics and slow neuromuscular response in this direction. The left rear sector (dashed line) is the secondary weak direction, with a peak of 0.70 occurring at a time lag of +18 milliseconds. It also exhibits a passive control mode but with a slightly faster reaction speed than the right rear sector. The front sector (dotted line) is the normal control direction, with a peak of 0.85 occurring at a time lag of -5 milliseconds. The negative time lag indicates that the pedal tilt moves before the pressure center, reflecting good active predictive control capability in this direction. The differences in the peak heights of the three curves quantify the coordination level in different directions. The sign and absolute value of the time lag parameter reveal the differences in the neural feedback speed and control strategies of the ankle joint in each direction, providing a quantitative basis for the spatiotemporal coupling relationship for accurate diagnosis of weak directions.

[0061] In one specific embodiment, step S3 determines the main weak direction based on the three sectors with the largest force-angle coupling response coefficients, specifically as follows:

[0062] The mean value of the force-angle coupling response coefficients of the eight sectors is calculated as the normalized reference value. The normalized coupling coefficients are obtained by dividing the force-angle coupling response coefficients of each sector by the normalized reference value.

[0063] Extract the absolute values ​​of the peak values ​​of the cross-correlation coefficient curves and the response time delay characteristics of each sector, and calculate the comprehensive weakness score of each sector according to the weighted formula, with weight coefficients of 0.5, 0.3, and 0.2 respectively;

[0064] The comprehensive weakness scores of the eight sectors are sorted in descending order. The center direction angles corresponding to the three sectors with the highest scores are selected as the main weak directions. The ratio of the standard deviation to the mean of the comprehensive weakness scores of the eight sectors is calculated to obtain the directional imbalance index.

[0065] Specifically, the normalized baseline value is obtained by summing the force-angle coupling response coefficients of the eight sectors and dividing by 8 to get the mean. The force-angle coupling response coefficient of each sector is then divided by this mean to obtain the normalized coupling coefficient. Normalization eliminates differences in absolute values ​​among different subjects or under different testing conditions, making the relative strength relationships between sectors clearer. The weighted formula for the comprehensive weakness score contains three components: the first component is the normalized coupling coefficient multiplied by a weighting factor of 0.5; the second component is 1 minus the peak value of the cross-correlation coefficient curve multiplied by a weighting factor of 0.3; and the third component is the absolute value of the response lag characteristic divided by 200 milliseconds and multiplied by a weighting factor of 0.2. The sum of these three components yields the comprehensive weakness score for that sector. A smaller peak value in the cross-correlation coefficient indicates poorer coordination between pressure center movement and pedal tilt; subtracting the peak value from 1 converts it into a penalty. A larger absolute value in the response lag indicates a slower neuromuscular response; normalization by dividing by 200 milliseconds ensures its value range is consistent with the other components.

[0066] The eight sectors are sorted in descending order of their overall weakness scores, with the sector with the highest score corresponding to the weakest direction in ankle joint stability. The center angles of the top three sectors are selected as the primary weak directions, indicating the specific areas requiring focused training. The directional imbalance index is calculated by first calculating the standard deviation of the overall weakness scores for the eight sectors, reflecting the dispersion of scores across sectors. Then, the mean of the eight scores is calculated. The coefficient of variation, obtained by dividing the standard deviation by the mean, is the directional imbalance index. A larger index value indicates a more significant difference in stability between different directions, suggesting a clear directional deficiency; a smaller value indicates more balanced stability across directions.

[0067] In one specific embodiment, step S4 includes:

[0068] The Euclidean distance between adjacent sampling points in the pressure center displacement sequence of the six-dimensional original data matrix is ​​accumulated to obtain the trajectory length. Principal component analysis is performed on the pressure center displacement sequence to extract the major and minor axes of the ellipse containing 95% of the data points. The area of ​​the ellipse is calculated as the swing area parameter.

[0069] Fast Fourier transforms were performed on the X-direction displacement sequence and Y-direction displacement sequence of the pressure center displacement sequence to obtain the power spectral density function. The frequency range was divided into a low-frequency band of 0.1 to 0.5 Hz, a mid-frequency band of 0.5 to 2 Hz, and a high-frequency band of 2 to 10 Hz. The ratio of the power in each frequency band to the total power was calculated to obtain the power ratio in the frequency domain.

[0070] The static control score and dynamic adjustment score are calculated separately according to the swing area parameter and trajectory length using a piecewise function. The sensory integration score is calculated according to the power proportion of the low-frequency band in the frequency domain.

[0071] Multiply the directional imbalance index by a penalty coefficient of 50 to obtain the directional penalty factor. Then, add the static control score, dynamic adjustment score, and sensory integration score using a weighted summation formula, and subtract the directional penalty factor to obtain the comprehensive stability score.

[0072] Specifically, the trajectory length is calculated by traversing 12,000 sampling points. The squared difference between the X-direction and Y-direction displacement values ​​of two adjacent sampling points is calculated. The square root of the sum of these two squared differences yields the Euclidean distance for that segment. The sum of the Euclidean distances for 11,999 segments gives the total trajectory length of the pressure center's movement. Principal component analysis treats the X-direction and Y-direction displacement sequences as a two-dimensional dataset. The covariance matrix is ​​calculated, and the eigenvalues ​​and eigenvectors are solved. The eigenvectors are oriented along the principal axes of the ellipse. After sorting by eigenvalue size, larger eigenvalues ​​correspond to the major axis, and smaller eigenvalues ​​correspond to the minor axis. The length of the interval containing 95% of the data points is calculated after projection along the principal axes and taken as the length of the major and minor axes of the ellipse. The area of ​​the ellipse is equal to pi multiplied by the length of the major axis multiplied by the length of the minor axis. The Fast Fourier Transform (FFT) transforms the X-direction and Y-direction displacement sequences from the time domain to the frequency domain, obtaining the amplitude of each frequency component. The square of the amplitude is divided by the number of sampling points to obtain the power spectral density function. The frequency range from 0 to 10 Hz is divided into three frequency bands, and the power spectral density function in each frequency band is integrated to obtain the total power of that frequency band. The total power of each frequency band is divided by the sum of the total power of the three frequency bands to obtain the frequency domain power ratio.

[0073] The piecewise function for static control scoring sets the score as follows: 100 points for an area of ​​oscillation less than 300 square millimeters; 100 points minus 0.15 multiplied by the difference between the oscillation area and 300 for an area of ​​300 to 600 square millimeters; and 0 points minus 0.05 multiplied by the difference between the oscillation area and 600 for an area of ​​oscillation greater than 600 square millimeters. The minimum score is 0 points. The piecewise function for dynamic adjustment scoring first calculates the trajectory length divided by 60 seconds to obtain the average speed. 100 points are awarded when the average speed is less than 400 millimeters per second; 100 points minus 0.125 multiplied by the difference between the average speed and 400 for an average speed between 400 and 800 millimeters per second; and 0 points minus 0.1 multiplied by the difference between the average speed and 800 for an average speed greater than 800 millimeters per second. The sensory integration score is calculated based on the deviation of the low-frequency power ratio from the ideal value of 0.5. The score equals 100 minus 200 multiplied by the absolute value of the difference between the low-frequency power ratio and 0.5. The weighted summation formula for the comprehensive stability score multiplies the static control score by 0.2, the dynamic adjustment score by 0.2, and the sensory integration score by 0.15. After adding the three weighted values, the directional imbalance index multiplied by 50 is subtracted to obtain the directional penalty factor. The final score range is limited to between 0 and 100.

[0074] The above describes the multi-dimensional ankle joint stability training and assessment method in the embodiments of this application. The following describes the multi-dimensional ankle joint stability training and assessment system in the embodiments of this application. One embodiment of the multi-dimensional ankle joint stability training and assessment system in the embodiments of this application includes:

[0075] The acquisition module is used to place a standard mass block at the preset calibration point 408 of the multi-axis pedal device, acquire the output value of the force sensor array 404 and the reading of the angle sensor 405, and establish the pressure center coordinate transformation matrix and tilt angle correction function.

[0076] The processing module is used to process the output value of the force sensor array 404 and the reading of the angle sensor 405 simultaneously with the pressure center coordinate transformation matrix and the tilt angle correction function when the subject stands on the multi-axis pedal device, so as to generate a six-dimensional raw data matrix containing the pressure center displacement sequence and the pedal tilt angle sequence.

[0077] The partitioning module is used to divide the pressure center displacement sequence and the pedal tilt angle sequence in the six-dimensional original data matrix into eight sector directions, calculate the force-angle coupling response coefficient of each sector, and determine the main weak direction by the three sectors with the largest force-angle coupling response coefficient.

[0078] The extraction module is used to extract time-frequency domain features from the six-dimensional original data matrix to obtain the trajectory length and frequency domain power ratio, and calculate the comprehensive stability score according to the hierarchical scoring function in combination with the directional imbalance index corresponding to the main weak direction.

[0079] The multidimensional ankle stability training and assessment system described above in the embodiments of this application has been presented. Please refer to [link / reference]. Figure 4 The following describes the multi-dimensional ankle joint stability training and assessment device in the embodiments of this application:

[0080] The device includes a square anti-slip pedal 401, a front-to-back tilt axis 402, a left-to-right swing axis 403, a force sensor array 404, an angle sensor 405, a metal support base 406, a shaft damping structure 407, a preset calibration point 408, a data transmission interface 409, and a power interface 410. The origin of the spatial rectangular coordinate system is set at the intersection of the front-to-back tilt axis 402 and the left-to-right swing axis 403. The square anti-slip pedal 401 provides a basic support surface for the subject's standing test and prevents slippage during the test to ensure safety. The front-to-back tilt axis 402 and the left-to-right swing axis 409... 3 provides the pedal with two-dimensional motion freedom in the forward and backward and left and right directions, which is the mechanical basis for the generation of the pedal tilt angle; the force sensor array 404 is distributed on the pedal surface according to the heel area, arch area, first metatarsal area, fifth metatarsal area, and toe area, serving as the core pressure acquisition unit, synchronously outputting pressure signals, and providing raw data for calculating the X / Y displacement sequence of the pressure center, the total plantar pressure sequence, and the pressure distribution asymmetry index sequence; the angle sensor 405 is embedded in the pedal edge and linked with the dual axes, accurately acquiring the axis rotation angle, providing data for generating the forward and backward tilt angle sequence and the left and right tilt angle sequence. Support; the metal support base 406 provides a stable load-bearing foundation for the entire device, adaptable to the load-bearing requirements of a 50 kg standard mass block calibration, preventing equipment deformation during calibration and testing; the shaft damping structure 407 is located at the connection point between the dual shafts and the base, allowing adjustment of the damping coefficient for pedal swing and tilt, simulating different balance challenge difficulties, and adapting to the assessment needs of different subjects such as rehabilitation patients and athletes; the preset calibration points 408 are distributed on the pedal surface in a 3×3 grid, serving as the calibration placement positions for the standard mass block, facilitating the collection of calibration data, establishment of the pressure center coordinate transformation matrix, and tilt angle correction function. The system provides a hardware benchmark; the origin of the spatial rectangular coordinate system serves as a unified coordinate benchmark and is the core reference point for calculating the pressure center coordinates, calibrating the tilt angle, and dividing the eight sectors; the data transmission interface 409 enables high-speed real-time transmission of 200 Hz synchronous sampling data between the force sensor array 404 and the angle sensor 405, providing a data path for subsequent data analysis such as generating a six-dimensional raw data matrix and extracting time-frequency domain features; the power interface 410 provides a stable and continuous power supply to the sensor array and data acquisition module, ensuring that the equipment is unpowered and data is not lost during calibration and the 60-second test.

[0081] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications 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.

Claims

1. A multi-dimensional ankle joint stability training assessment method, characterized in that, The method includes: Step S1: Place the standard mass block at the preset calibration point (408) of the multi-axis pedal device, collect the output value of the force sensor array (404) and the reading of the angle sensor (405), and establish the pressure center coordinate transformation matrix and tilt angle correction function; Step S2: The subject stands on the multi-axis pedal device. The output value of the force sensor array (404) and the reading of the angle sensor (405) are processed synchronously by the pressure center coordinate transformation matrix and the tilt angle correction function to generate a six-dimensional raw data matrix containing the pressure center displacement sequence and the pedal tilt angle sequence. Step S3: Divide the pressure center displacement sequence and the pedal tilt angle sequence in the six-dimensional original data matrix into eight sector directions, calculate the force-angle coupling response coefficient of each sector, and determine the main weak direction from the three sectors with the largest force-angle coupling response coefficient; Step S4: Extract time-frequency domain features from the six-dimensional original data matrix to obtain the trajectory length and frequency domain power ratio. Combine the directional imbalance index corresponding to the main weak direction with the comprehensive stability score calculated according to the hierarchical scoring function.

2. The multidimensional ankle joint stability training and assessment method according to claim 1, characterized in that, Step S1 includes: A spatial rectangular coordinate system is established with the intersection of the front and rear tilting axis (402) and the left and right swinging axis (403) of the multi-axis pedal device as the origin, and the planar coordinate position of the measuring unit in each force sensor array (404) is determined. The standard mass block is placed sequentially on nine preset calibration points (408) on the upper surface of the multi-axis pedal device, and the output value of each measurement unit and the tilt angle reading of the angle sensor (405) are recorded simultaneously to obtain the calibration dataset. Based on the output values ​​of each measurement unit and their corresponding planar coordinate positions in the calibration dataset, a linear mapping relationship between the output values ​​of the measurement units and the coordinates of the pressure center is established through the torque balance equation, thereby generating the pressure center coordinate transformation matrix. The tilt angle readings in the calibration dataset are fitted with the theoretical tilt angles generated by the standard mass block at each calibration point to construct a tilt angle error correction function as the tilt angle correction function.

3. The multidimensional ankle joint stability training and assessment method according to claim 2, characterized in that, Step S2 includes: The subject stands on the multi-axis pedal device, and the output value of the force sensor array (404) and the reading of the angle sensor (405) are synchronously acquired at a sampling frequency of 200 Hz to obtain the raw signal sequence of 60 seconds of test duration; Substitute the output value of the force sensor array (404) in the original signal sequence into the pressure center coordinate transformation matrix to calculate the pressure center displacement sequence, including the X-direction displacement sequence and the Y-direction displacement sequence; Substitute the angle sensor (405) readings from the original signal sequence into the tilt correction function to calculate the pedal tilt sequence, which includes the front and rear tilt sequence and the left and right tilt sequence; The total pressure sequence of the foot is obtained by summing the output values ​​of the force sensor array (404). The pressure difference between the left and right sensor groups is calculated as a ratio to the total pressure to obtain the pressure distribution asymmetry index sequence. The pressure center displacement sequence, the pedal tilt angle sequence, the total foot pressure sequence and the pressure distribution asymmetry index sequence are combined to form the six-dimensional original data matrix.

4. The multidimensional ankle joint stability training and assessment method according to claim 3, characterized in that, In step S3, the pressure center displacement sequence and the pedal tilt angle sequence in the six-dimensional original data matrix are divided into eight sector directions, specifically: Centered on the origin of the aforementioned spatial rectangular coordinate system, the tread plate is divided into eight sectors at 45-degree intervals: front, front right, front right, rear right, rear right, rear left, front left, and front left. The central direction angles of each sector are 0 degrees, 45 degrees, 90 degrees, 135 degrees, 180 degrees, 225 degrees, 270 degrees, and 315 degrees, respectively. Perform arctangent operation on the X-direction displacement sequence and Y-direction displacement sequence in the pressure center displacement sequence to obtain the pressure center direction angle sequence, and count the time percentage of the pressure center direction angle entering each sector angle interval at each sampling time. The front-to-back tilt angle sequence and the left-to-right tilt angle sequence in the pedal tilt angle sequence are subjected to arctangent operation to obtain the pedal tilt direction angle sequence. When the pedal tilt direction angle points to a certain sector, the tilt angle amplitude at the corresponding moment is accumulated to obtain the cumulative angle amplitude of the corresponding sector.

5. The multidimensional ankle joint stability training and assessment method according to claim 4, characterized in that, In step S3, the force-angle coupling response coefficients of each sector are calculated as follows: Divide the cumulative angular amplitude of each sector by the product of the time proportion of the corresponding sector and the test duration to obtain the force-angle coupling response coefficient of each sector. Extract the time periods pointing to each sector from the pressure center displacement sequence, and calculate the radial displacement sequence of the pressure center and the tilt angle projection sequence of the pedal in the corresponding direction; Cross-correlation function calculation is performed on the radial displacement sequence of the pressure center and the tilt projection sequence to obtain cross-correlation coefficient curves with a time delay range of -200 milliseconds to +200 milliseconds. The time delay parameter corresponding to the peak value of the cross-correlation coefficient curve is extracted as the response time delay feature of each sector.

6. The multidimensional ankle joint stability training and assessment method according to claim 5, characterized in that, In step S3, the main weak directions are determined by the three sectors with the largest force-angle coupling response coefficients, specifically as follows: The mean value of the force-angle coupling response coefficients of the eight sectors is calculated as the normalized reference value, and the normalized coupling coefficients are obtained by dividing the force-angle coupling response coefficients of each sector by the normalized reference value. Extract the peak value of the cross-correlation coefficient curve of each sector and the absolute value of the response time delay characteristic, and calculate the comprehensive weakness score of each sector according to the weighted formula, with weight coefficients of 0.5, 0.3 and 0.2 respectively; The comprehensive weakness scores of the eight sectors are sorted in descending order, and the center direction angles corresponding to the three sectors with the highest scores are selected as the main weak directions. The ratio of the standard deviation to the mean of the comprehensive weakness scores of the eight sectors is calculated to obtain the directional imbalance index.

7. The multidimensional ankle joint stability training and assessment method according to claim 1, characterized in that, Step S4 includes: The Euclidean distance between adjacent sampling points in the pressure center displacement sequence in the six-dimensional original data matrix is ​​accumulated to obtain the trajectory length. Principal component analysis is performed on the pressure center displacement sequence to extract the major and minor axes of an ellipse containing 95% of the data points, and the area of ​​the ellipse is calculated as the swing area parameter. Fast Fourier transform is performed on the X-direction displacement sequence and Y-direction displacement sequence of the pressure center displacement sequence to obtain the power spectral density function. The frequency range is divided into a low-frequency band of 0.1 to 0.5 Hz, a mid-frequency band of 0.5 to 2 Hz, and a high-frequency band of 2 to 10 Hz. The ratio of the power of each frequency band to the total power is calculated to obtain the frequency domain power ratio. The static control score and dynamic adjustment score are calculated respectively according to the swing area parameter and the trajectory length using a piecewise function. The sensory integration score is calculated according to the low-frequency power ratio in the frequency domain power ratio. The directional imbalance index is multiplied by a penalty coefficient of 50 to obtain a directional penalty factor. The static control score, the dynamic adjustment score, and the sensory integration score are weighted and summed according to a weighted summation formula, and then the directional penalty factor is subtracted to obtain the comprehensive stability score.

8. A multi-dimensional ankle joint stability training and assessment system, characterized in that, For implementing the multidimensional ankle stability training assessment method as described in any one of claims 1-7, the multidimensional ankle stability training assessment system comprises: The acquisition module is used to place a standard mass block at the preset calibration point (408) of the multi-axis pedal device, acquire the output value of the force sensor array (404) and the reading of the angle sensor (405), and establish the pressure center coordinate transformation matrix and the tilt angle correction function; The processing module is used to process the output value of the force sensor array (404) and the reading of the angle sensor (405) simultaneously by the pressure center coordinate transformation matrix and the tilt angle correction function when the subject stands on the multi-axis pedal device, so as to generate a six-dimensional raw data matrix containing the pressure center displacement sequence and the pedal tilt angle sequence. The partitioning module is used to divide the pressure center displacement sequence and the pedal tilt angle sequence in the six-dimensional original data matrix into eight sector directions, calculate the force-angle coupling response coefficient of each sector, and determine the main weak direction by the three sectors with the largest force-angle coupling response coefficient. The extraction module is used to extract time-frequency domain features from the six-dimensional original data matrix to obtain the trajectory length and frequency domain power ratio, and calculate the comprehensive stability score according to the hierarchical scoring function in combination with the directional imbalance index corresponding to the main weak direction.

9. The system according to claim 8, characterized in that, A standard mass block is placed at the preset calibration point (408) of the multi-axis pedal device. The output values ​​of the force sensor array (404) and the readings of the angle sensor (405) are collected to establish the pressure center coordinate transformation matrix and the tilt angle correction function, including: A spatial rectangular coordinate system is established with the intersection of the front and rear tilting axis (404) and the left and right swinging axis (403) of the multi-axis pedal device as the origin, and the planar coordinate position of the measuring unit in each force sensor array (404) is determined. The standard mass block is placed sequentially on nine preset calibration points (408) on the upper surface of the multi-axis pedal device, and the output value of each measurement unit and the tilt angle reading of the angle sensor (405) are recorded simultaneously to obtain the calibration dataset. Based on the output values ​​of each measurement unit and their corresponding planar coordinate positions in the calibration dataset, a linear mapping relationship between the output values ​​of the measurement units and the coordinates of the pressure center is established through the torque balance equation, thereby generating the pressure center coordinate transformation matrix. The tilt angle readings in the calibration dataset are fitted with the theoretical tilt angles generated by the standard mass block at each calibration point to construct a tilt angle error correction function as the tilt angle correction function.

10. The system according to claim 8, characterized in that, The subject stands on the multi-axis pedal device. The output values ​​of the force sensor array (404) and the readings of the angle sensor (405) are processed synchronously through the pressure center coordinate transformation matrix and the tilt angle correction function to generate a six-dimensional raw data matrix containing a pressure center displacement sequence and a pedal tilt angle sequence, including: The subject stands on the multi-axis pedal device, and the output value of the force sensor array (404) and the reading of the angle sensor (405) are synchronously acquired at a sampling frequency of 200 Hz to obtain the raw signal sequence of 60 seconds of test duration; Substitute the output value of the force sensor array (404) in the original signal sequence into the pressure center coordinate transformation matrix to calculate the pressure center displacement sequence, including the X-direction displacement sequence and the Y-direction displacement sequence; Substitute the angle sensor (405) readings from the original signal sequence into the tilt correction function to calculate the pedal tilt sequence, which includes the front and rear tilt sequence and the left and right tilt sequence; The total pressure sequence of the foot is obtained by summing the output values ​​of the force sensor array (404). The pressure difference between the left and right sensor groups is calculated as a ratio to the total pressure to obtain the pressure distribution asymmetry index sequence. The pressure center displacement sequence, the pedal tilt angle sequence, the total foot pressure sequence and the pressure distribution asymmetry index sequence are combined to form the six-dimensional original data matrix.

Images