Human body walking process centroid stability assessment method and system

Abstract

A method for evaluating the stability of the center of mass in a human body walking process belongs to the technical field of health evaluation, and comprises the following steps of: accurately dividing a swing phase in the human body walking process through a gait segmentation technology by utilizing posture data acquired by inertial measurement units worn on two feet; carrying out deep analysis on attitude data acquired by an inertial measurement unit at a posterior lumbosacrum in the swing phase to obtain a mass center stability characteristic index in the swing phase; performing data processing on the characteristic indexes of the stability of the centroid in the swing phase, substituting the processed indexes into a proposed dynamic centroid stability index expression, and calculating a dynamic centroid stability index; the mass center stability of the human body in the walking process is evaluated according to the dynamic mass center stability index. The invention further provides a system for evaluating the mass center stability in the human body walking process. The system comprises an acquisition module, a transmission module and a data processing module. According to the evaluation method and system provided by the invention, the convenience and flexibility of mass center stability evaluation can be improved, and a new characteristic index is provided for fall risk and balance capability evaluation.

Description

Technical Field

[0001] The invention belongs to the technical field of health assessment and relates to a method for assessing the stability of the center of mass by utilizing the kinematic information of the center of mass during human walking. Background Art

[0002] Center of mass stability refers to the ability of an object to maintain a balanced center of mass position when it is moving or stationary, which is particularly important for humans. It can reflect a person's balance ability in dynamic activities such as walking and running, and the quality of balance ability is directly related to the risk of falling. By evaluating the center of mass stability, the risk of falling can be indirectly predicted and understood. With the increase of age, the various functions of the human body decline, resulting in a decrease in the balance ability of the elderly, an increase in the risk of falling, and a decrease in the quality of life. Therefore, evaluating and calculating the center of mass stability is crucial for predicting the risk of falling, which can help individuals detect balance disorders early, develop personalized rehabilitation plans, and monitor the efficacy. In addition, the evaluation of center of mass stability also plays an important role in sports training and medical research. Many sports such as gymnastics and skating require athletes to have good balance ability, which is good center of mass stability. Paying attention to center of mass stability can not only significantly reduce fall accidents, improve quality of life and safety, but also provide important value for patients' rehabilitation.

[0003] The analysis of center of mass stability during the swing phase is of great significance. During the swing phase, when one leg leaves the ground, the position and motion of the center of mass of the human body are critical to overall balance and stability. Analysis of center of mass stability during the swing phase helps to understand how the human body maintains balance under dynamic conditions and how it adjusts its gait to adapt to different environmental conditions.

[0004] At present, the technologies for evaluating the stability of the center of mass during human walking mainly include pressure sensor technology, video analysis technology, machine learning technology and inertial sensor technology. Pressure sensor technology detects the pressure distribution of the foot by installing sensors on the insole or the ground to analyze the gait and the stability of the center of pressure of the sole. However, it can only analyze the motion stability of the lower limbs in the standing phase of human walking, and cannot analyze the motion stability of the lower limbs in the swing phase. In addition, it is limited by the venue and sensor distribution and cannot fully reflect the changes in the center of mass of the whole body. Video analysis technology captures images through cameras and extracts the positions of key points of the body to provide intuitive center of mass trajectories, but it is affected by lighting and camera angles, and data processing is complex. Machine learning technology improves the accuracy of evaluation by training models with large amounts of data, but it requires a lot of data and computing resources. Summary of the invention

[0005] In order to overcome the shortcomings of the existing technology, the present invention provides a method and system for evaluating the center of mass stability during human walking, aiming to improve the convenience and flexibility of center of mass stability evaluation and provide new characteristic indicators for fall risk and balance ability evaluation; the scheme uses posture data collected by inertial measurement units worn on both feet to accurately divide the swing phase of human walking through gait segmentation technology; then, the posture data collected by the inertial measurement unit at the posterior lumbar sacrum in the swing phase is deeply analyzed to obtain the center of mass stability characteristic index in the swing phase, and the dynamic center of mass stability index is calculated, and the dynamic center of mass stability index is used to evaluate the center of mass stability during human walking.

[0006] The technical solution adopted by the present invention to solve the technical problem is:

[0007] A method for evaluating the stability of the center of mass of a human body during walking comprises the following steps:

[0008] Step 1: record the subject information, wear an inertial measurement unit on the subject, and establish communication between the inertial measurement unit and the computer host software;

[0009] Step 2, start the test, and let the subject implement the dynamic walking test paradigm according to the walking test trajectory, during which the wearable inertial measurement unit is used to collect human posture data;

[0010] Step 3, ending the test, transmitting the attitude data collected by the three inertial measurement units in step 2 to the computer by wireless communication;

[0011] Step 4, gait phase division: perform data analysis on the posture data collected by the inertial measurement unit worn on both feet in step 3 to divide the gait phase;

[0012] Step 5, based on the gait phase obtained in step 4 and the posture data collected by the inertial measurement unit at the posterior lumbar sacrum, an in-depth analysis is performed to obtain a characteristic index of the center of mass stability in the swing phase;

[0013] Step 6, calculation of dynamic center of mass stability index: data processing is performed on the center of mass stability characteristic index in the swing phase obtained in step 5, and then it is brought into the established S-type utility function of each index to obtain the utility value corresponding to each index parameter, and finally the utility value corresponding to each index parameter is brought into the established dynamic center of mass stability index expression to obtain the dynamic center of mass stability index;

[0014] Step 7, center of mass stability assessment: The center of mass stability of the human body during walking is assessed according to the dynamic center of mass stability index obtained in step 6.

[0015] Furthermore, in step 5, the characteristic indexes of the center of mass stability in the swing phase include the center of mass swing trajectory length, the center of mass average swing rate, the center of mass offset area, the center of mass average swing amplitude and the center of mass swing symmetry, which are calculated as follows:

[0016] (5.1) Center of mass swing trajectory length: The total motion trajectory length of the human body's center of mass during the experiment. The position change of the human body's center of mass is measured by the inertial measurement unit at the sacrum, and the motion trajectory of the center of mass is calculated. First, the initial position of the human body's center of mass is determined, including the coordinates in the three directions of the x, y and z axes. Then, the acceleration of the human body's center of mass in these three directions is measured, and the acceleration is integrated to obtain the position change of the center of mass, which is added to the initial position to obtain the position of the center of mass at a certain moment. Finally, during the entire test time, the Euclidean distance of the center of mass in three-dimensional space is calculated, and these distances are accumulated to obtain the center of mass swing trajectory length L, which is calculated as follows:

[0017]

[0018] Among them, r 0 =[r x0 r y0 r z0 ] represents the initial position of the human body's center of mass in the three directions of the x, y and z axes; a=[a x a y a z ] represents the acceleration of the human body's center of mass in the three directions of x, y and z axes, and the position change of the center of mass is obtained by integrating it; T is the total test time; It represents the Euclidean distance of the center of mass of the human body in three-dimensional space, and L is the length of the center of mass swing trajectory;

[0019] (5.2) Average center of mass swing rate: The average value of the center of mass swing trajectory length per unit time in the total test time. The average speed is obtained by dividing the total length of the center of mass swing trajectory by the total test time. The average center of mass swing rate v avg The calculation formula is as follows:

[0020]

[0021] Among them, L t is the length of the center of mass swing trajectory per unit time, and T is the total test time;

[0022] (5.3) Center of mass offset area: The integral of the change in the projection area of ​​the waist sensor parallel to the back plane of the human body at each moment over time. The center of mass offset area A cnv The calculation formula is as follows:

[0023]

[0024] Among them, (x i ,y i ,z i ) is the coordinate of the waist sensor at the current moment, (x i+1 ,y i+1 ,z i+1 ) is the coordinate of the waist sensor at the next moment;

[0025] (5.4) Average center of mass swing amplitude: The average distance from the center of mass swing trajectory to the center of mass swing. First, the center of mass position of each sampling point is summed and divided by the number of sampling points to obtain the coordinates of the center of mass swing. Then, the distance from each sampling point to the center of mass swing is calculated, and the average of these distances is calculated. The center of mass swing C cen It is the average value of the centroid position, and the calculation formula is as follows:

[0026]

[0027] Among them, (r x,i ,r y,i ,r z,i ) is the centroid position of the ith data sample, N is the number of sampling points,

[0028] Average swing amplitude of center of mass A avg :

[0029]

[0030] Among them, C cen,x , C cen,y and C cen,z They represent the x-coordinate, y-coordinate and z-coordinate of the center of mass swing respectively;

[0031] (5.5) Center of mass swing symmetry: The ratio of the center of mass swing trajectory of the left foot to the right foot when walking. The calculation formula of the center of mass swing symmetry CR is as follows:

[0032]

[0033] Among them, L lc and L rc They respectively represent the length of the center of mass swing trajectory when the left foot walks and the length of the center of mass swing trajectory when the right foot walks.

[0034] Furthermore, in step 6, the dynamic center of mass stability index expression is established as follows:

[0035] 6.1, Establishing the data set: Implement steps 1-3 on the subjects, collect the human posture data of each subject, and have professional doctors score the subjects using the Timed Up and Go Test TUG scale;

[0036] 6.2, Data processing and correlation analysis: Steps 4 and 5 are performed on the multiple groups of human posture data obtained in 6.1, and gait phase division is performed to obtain multiple groups of center of mass stability characteristic indicators in the swing phase. The obtained multiple groups of center of mass stability characteristic indicators in the swing phase and the TUG scale scores obtained in 6.1 are normalized using the Min-Max normalization method to obtain the normalized value X of the center of mass stability characteristic indicators in each swing phase and the TUG scale score. norm , so that they have the same dimension and numerical range; the Spearman correlation coefficient analysis method was used for the normalized multiple groups of TUG scale scores and multiple groups of swing phase center of mass stability characteristic indicators. The analysis results showed that the five normalized swing phase center of mass stability characteristic indicators were significantly positively correlated with the TUG scale scores;

[0037] 6.3, define the S-type utility function: According to 6.2, the five normalized swing phase center of mass stability characteristic indicators are all positive indicators, and the data distribution changes rapidly in the middle range and slowly at both ends. Therefore, the Logistic function is selected as the S-type utility function of each indicator, and its expression is:

[0038]

[0039] Among them, U n (X norm ) is the utility value corresponding to the parameter of each indicator, ranging from 0 to 1; k n is the slope parameter of each indicator curve, which controls the steepness of the curve; the larger the slope, the steeper the curve and the faster the utility value changes; 0,n is the midpoint position parameter of each index curve, which determines the translation position of the curve;

[0040] 6.4, define the loss function and optimize the S-type utility function: In order to optimize the k of the S-type utility function, the characteristic index of the center of mass stability in each swing phase n and x 0,n , define the mean square error MSE as the loss function to measure the deviation between the utility value predicted by the model and the actual observed value. By minimizing this loss function, adjust the parameters of the utility function to improve its prediction accuracy; its expression is:

[0041]

[0042] Among them, L n (θ n ) is the loss function of each indicator; X norm,i is the value of the i-th sample after normalization of each index; θ n =(k n ,x 0,n) are the model parameters of each indicator; N is the number of samples; Y n,i is the actual utility value of the ith sample of each indicator, which is the normalized TUG scale score in 6.2;

[0043] 6.5, Update k and x using gradient descent 0 Parameters: Calculate the loss function of each indicator for k and x 0 Gradient and It means that the loss function varies with k and x 0 The direction and rate of change, set the learning rate a, and update the parameters according to the gradient descent formula: Among them, k t and x 0 (t) is the parameter value of the tth iteration; update k and x corresponding to each indicator 0 Parameters until each loss function converges, the convergence condition is: ΔL=|L(θ t+1 )-L(θ t )|<∈, where ∈ is the set threshold;

[0044] 6.6, establish the dynamic center of mass stability index expression: First, calculate the normalized coefficient of variation CV of the center of mass stability characteristic index in each swing phase obtained in 6.2 norm,n , and then the power average synthesis method is used to establish the dynamic center of mass stability index expression, which is shown as follows:

[0045]

[0046] Among them, U(X norm,n ) is the utility value corresponding to the nth indicator input parameter value, m is the power, and DCMSI is the dynamic center of mass stability index;

[0047] 6.7. Find the optimal power m of the dynamic center of mass stability index expression in 6.6: The initial range of the power is determined to be within [-2,3]; within this range, the power m is modeled with a step size of 0.1 to obtain the Spearman correlation coefficient between the dynamic center of mass stability index generated by the dynamic center of mass stability index expressions of different powers and the normalized TUG scale score in 6.2, and the power corresponding to the maximum correlation is selected as the optimal power m, which is substituted into the dynamic center of mass stability index expression in 6.6 to obtain the optimal power dynamic center of mass stability index expression.

[0048] Furthermore, in step 2, the walking test trajectory includes a human standing position mark, a 3-meter to 5-meter straight walking dotted line mark and a turnaround mark; when implementing the human balance ability detection paradigm in step 2, the subject first stands on the human standing position mark, and when the subject hears the instruction, the dynamic walking test paradigm begins.

[0049] The dynamic walking test paradigm is a 3-meter to 5-meter straight-line return walking test, which is used to evaluate the center of mass stability of the human body during walking.

[0050] Preferably, in step 2, the posture data includes at least three-axis acceleration and angular velocity.

[0051] Preferably, the three-axis acceleration includes left foot movement acceleration, right foot movement acceleration and lumbar sacrum movement acceleration; the angular velocity includes left foot movement angular velocity, right foot movement angular velocity and lumbar sacrum movement angular velocity.

[0052] Furthermore, in step 4, the gait phase is calculated as follows: first, the three-axis acceleration and angular velocity data collected from the sensors at the left foot, right foot and lumbar sacrum are processed by a first-order Butterworth low-pass filter to remove noise and smooth the data; by analyzing the characteristic changes of the pitch angle in the angular velocity data, the key events in the gait cycle are identified: heel strike (HS) and toe off (TO); the above two gait events mark the transition of the gait from one stage to another; heel strike marks the beginning of the stance phase, and toe off marks the entry of the leg into the swing phase. Based on the identified gait events, the gait cycle is divided into different phases.

[0053] In step 7, the center of mass stability assessment method is: the obtained dynamic center of mass stability index is divided into center of mass stability intervals using the quartile analysis method, and the center of mass stability is graded: below the second quartile is a low risk of center of mass stability, between the second and third quartiles is a medium risk of center of mass stability, and greater than the third quartile is a high risk of center of mass stability.

[0054] A system for evaluating the stability of the center of mass of a human body during walking is provided, and a method for evaluating the stability of the center of mass of a human body during walking is implemented. The system comprises a collection module, a transmission module and a data processing module, wherein:

[0055] The acquisition module includes a walking test track and an inertial measurement unit, wherein the walking test track is used to implement a dynamic walking test paradigm, and the inertial measurement unit is used to acquire posture data;

[0056] The transmission module includes a wireless communication module, which is used to transmit the posture data received from the acquisition module to the data processing module in a wireless communication manner;

[0057] The data processing module includes a computer, on which upper computer software is installed, which is used to calculate the characteristic index of center of mass stability in the swing phase based on posture data, and obtain a dynamic center of mass stability index based on the analysis of the characteristic index of center of mass stability in the swing phase to evaluate the center of mass stability during human walking.

[0058] The technical concept of the present invention is: using inertial sensor technology, mainly using three inertial measurement units for collection and analysis. In contrast, inertial sensor technology shows significant advantages: the inertial measurement unit is not restricted by the site environment, is easy to wear and carry, does not affect natural walking, and has the characteristics of low cost and small size. The measurement results of the inertial measurement unit are crucial to understanding human motion patterns, gait analysis and center of mass stability evaluation, especially the evaluation of the center of mass stability of the lower limb swing phase during human walking. The present invention uses the data collected by the two-foot IMU to divide the gait phase, takes the position of the waist IMU as the position of the human center of mass, and analyzes the collected data to evaluate the stability of the human center of mass. Although there are problems of integral error and noise, the accuracy can be significantly improved by data fusion and filter design. In summary, the inertial measurement unit provides an efficient, accurate and user-friendly solution in the evaluation of center of mass stability, overcomes the limitations of the prior art, and provides a portable and effective technical means for the evaluation of center of mass stability during human walking.

[0059] The inertial measurement unit of the present invention is worn on the left foot instep, the right foot instep and the posterior lumbar sacrum. The inertial measurement unit at the posterior lumbar sacrum is regarded as the center of mass position of the human body. The posture data including triaxial acceleration and angular velocity are collected. The inertial measurement unit transmits the collected data to the computer host software through wireless communication for data visualization, gait phase division, calculation of characteristic index of center of mass stability in swing phase, establishment of dynamic center of mass stability index expression and calculation of dynamic center of mass stability index. When implementing the dynamic walking test paradigm, the subject stands naturally on the human standing position mark of the walking test track, with both feet as wide as shoulders, and performs a 3-meter to 5-meter straight return walking test. After the test, the computer host software divides the gait phase based on the posture data collected by the inertial measurement unit, calculates the characteristic index of center of mass stability in swing phase, calculates the dynamic center of mass stability index using the dynamic center of mass stability index expression, and evaluates the center of mass stability of the human body during walking. The dynamic center of mass stability index characterizes the stability of the center of mass during human walking. The higher the dynamic center of mass stability index, the more unstable the center of mass of the human body is during walking, and the more likely it is to fall.

[0060] Compared with the prior art, the beneficial effects of the present invention are mainly manifested in that: the present invention designs a wearable human walking center of mass stability evaluation method and system for the stability of the human walking center of mass. The method uses the posture data collected by the inertial measurement unit worn on the instep of both feet, and accurately divides the swing phase of the human walking process through the gait segmentation technology. Subsequently, the posture data collected by the inertial measurement unit at the posterior lumbar sacrum in the swing phase is deeply analyzed, and 5 characteristic indices of the center of mass stability in the swing phase are proposed and obtained. The characteristic indices of the center of mass stability in the swing phase are input into the proposed dynamic center of mass stability index expression after data processing, and the dynamic center of mass stability index (Dynamic Center of Mass Stability Index, referred to as DCMSI) is calculated. Finally, the DCMSI is used to evaluate the center of mass stability during human walking. DCMSI can quantitatively reflect the stability of the center of mass during human walking. The larger the DCMSI value, the more unstable the human walking is, which provides a new quantitative index for gait analysis and walking stability evaluation. The system only includes three inertial measurement units, a walking test track and a computer. It has the advantages of small footprint, easy to carry and transport, simple and convenient to wear, and high cost performance. It is very suitable for the evaluation and comparison of the center of mass stability of the human body during walking in various places, and provides a reliable tool for the study of the center of mass stability of the human body during walking. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a schematic diagram of the posture data collection process;

[0062] Figure 2 Schematic diagram of the wearing position of the inertial measurement unit, where (a) is the back and (b) is the front;

[0063] Figure 3 This is a schematic diagram of the walking test trajectory;

[0064] Figure 4 It is a schematic diagram of the gait phase of a human body when walking;

[0065] Figure 5 This is a schematic diagram of the actual application process of the human body balance ability detection method. DETAILED DESCRIPTION

[0066] The present invention will be further described below in conjunction with the accompanying drawings.

[0067] Example 1

[0068] Reference Figure 1 to Figure 5 , a method for evaluating the center of mass stability of a human body during walking. This method uses the posture data collected by the inertial measurement unit worn on both feet and the gait segmentation technology to accurately divide the swing phase of the human body during walking. The swing phase diagram is shown in Figure 4Then, the posture data collected by the inertial measurement unit at the posterior lumbar sacrum during the swing phase is deeply analyzed to obtain the characteristic index of center of mass stability during the swing phase. Finally, the dynamic center of mass stability index is calculated to evaluate the center of mass stability of the human body during walking. Figure 5 The method for evaluating the stability of the center of mass of a human body during walking described in this embodiment includes the following steps:

[0069] Step 1: Inform the test subjects of the test process and record the test subject information. Figure 2 Make the subjects wear the inertial measurement unit and establish communication between the inertial measurement unit and the computer host software;

[0070] The human posture collection wearing scheme is: an inertial measurement unit is worn on the left and right insteps and the posterior lumbar sacrum of the subject, namely the first inertial measurement unit 3 and the second inertial measurement unit 4. The inertial measurement units are worn on the left and right insteps to divide the gait phase of the human body during walking. The third inertial measurement unit 2 worn on the posterior lumbar sacrum is used to characterize the movement of the human body's center of mass. The back of the third inertial measurement unit 2 is pasted with a Velcro hook surface, and is tied with a Velcro elastic elastic band, wherein the waist is a snap-on Velcro elastic elastic band 1, which can be matched with the Velcro fur surface on the back of the third inertial measurement unit 2 for easy fixation.

[0071] Step 2: Start the test and let the subjects perform a dynamic walking test paradigm according to the set walking test trajectory, including a 3-meter straight walking test and a curved walking test. During this period, the three inertial measurement units worn by the subjects are used to collect human posture data;

[0072] Step 3, ending the test, transmitting the attitude data collected by the three inertial measurement units in step 2 to the computer by wireless communication;

[0073] Step 4, gait phase division: perform data analysis on the posture data collected by the inertial measurement unit worn on both feet in step 3 to divide the gait phase;

[0074] The gait phases are divided as follows:

[0075] 4.1, performing a first-order Butterworth low-pass filter on the triaxial acceleration and angular velocity data collected from the sensors at the left foot, right foot and posterior lumbar sacrum in step 3 to remove noise and smooth the data;

[0076] 4.2, analyze the characteristic changes of pitch angle in the angular velocity data after low-pass filtering, and identify the key events in the gait cycle: Heel Strike (HS) and Toe Off (TO). The above two gait events mark the transition from one stage to another. Heel strike marks the beginning of the stance phase of one leg, and toe off marks the entry of the leg into the swing phase;

[0077] 4.3, based on the gait events identified in 4.2, the gait cycle is divided into different phases.

[0078] Step 5, calculation of characteristic indexes of center of mass stability in the swing phase: Based on the gait phase obtained in step 4 and the posture data collected by the inertial measurement unit at the posterior lumbar sacrum, in-depth analysis is performed to obtain characteristic indexes of center of mass stability in the swing phase, including center of mass swing trajectory length, center of mass average swing rate, center of mass offset area, center of mass average swing amplitude, and center of mass swing symmetry. The calculation method is as follows:

[0079] (5.1) Center of mass swing trajectory length: The total motion trajectory length of the human body's center of mass during the experiment. The position change of the human body's center of mass is measured by the inertial measurement unit at the sacrum, and the motion trajectory of the center of mass is calculated. First, the initial position of the human body's center of mass is determined, including the coordinates in the three directions of the x, y and z axes. Then, the acceleration of the human body's center of mass in these three directions is measured, and the acceleration is integrated to obtain the position change of the center of mass, which is added to the initial position to obtain the position of the center of mass at a certain moment. Finally, during the entire test time, the Euclidean distance of the center of mass in three-dimensional space is calculated, and these distances are accumulated to obtain the center of mass swing trajectory length L, which is calculated as follows:

[0080]

[0081] Among them, r 0 =[r x0 r y0 r z0 ] represents the initial position of the human body's center of mass in the three directions of the x, y and z axes; a=[a x a y a z ] represents the acceleration of the human body's center of mass in the three directions of x, y and z axes, and the position change of the center of mass is obtained by integrating it; T is the total test time; It represents the Euclidean distance of the center of mass of the human body in three-dimensional space, and L is the length of the center of mass swing trajectory;

[0082] (5.2) Average center of mass swing rate: The average value of the center of mass swing trajectory length per unit time in the total test time. The average speed is obtained by dividing the total length of the center of mass swing trajectory by the total test time. The average center of mass swing rate vavg The calculation formula is as follows:

[0083]

[0084] Among them, L t is the length of the center of mass swing trajectory per unit time, and T is the total test time;

[0085] (5.3) Center of mass offset area: The integral of the change in the projection area of ​​the waist sensor parallel to the back plane of the human body at each moment over time. Center of mass offset area A cnv The calculation formula is as follows:

[0086]

[0087] Among them, (x i ,y i ,z i ) is the coordinate of the waist sensor at the current moment, (x i+1 ,y i+1 ,z i+1 ) is the coordinate of the waist sensor at the next moment;

[0088] (5.4) Average center of mass swing amplitude: The average distance from the center of mass swing trajectory to the center of mass swing. First, sum the center of mass positions of each sampling point and divide by the number of sampling points to get the coordinates of the center of mass swing. Then, calculate the distance from each sampling point to the center of mass swing and find the average of these distances. The center of mass swing C cen It is the average value of the centroid position, and the calculation formula is as follows:

[0089]

[0090] Among them, (r x,i ,r y,i ,r z,i ) is the centroid position of the i-th data sample, and N is the number of sampling points;

[0091] Average swing amplitude of center of mass A avg :

[0092]

[0093] Among them, C cen,x , C cen,y and C cen,z They represent the x-coordinate, y-coordinate and z-coordinate of the center of mass swing respectively.

[0094] (5.5) Center of mass swing symmetry: The ratio of the center of mass swing trajectory of the left foot to the right foot when walking. The calculation formula of the center of mass swing symmetry CR is as follows:

[0095]

[0096] Among them, L lc and L rc They represent the length of the center of mass swing trajectory when walking with the left foot and the length of the center of mass swing trajectory when walking with the right foot respectively;

[0097] Step 6, calculation of dynamic center of mass stability index: perform data processing in 6.2 on the center of mass stability characteristic index in the swing phase obtained in step 5, and then bring it into the established S-type utility function of each index in 6.3 to obtain the utility value corresponding to each index parameter, and finally bring the utility value corresponding to each index parameter into the dynamic center of mass stability index expression established in 6.6 and 6.7 to obtain the dynamic center of mass stability index. The process of establishing the dynamic center of mass stability index expression is as follows:

[0098] 6.1, Establishing a data set: Implement steps 1-3 on a large number of subjects, collect the human posture data of each subject, and have professional doctors score the subjects using the Timed Up and Go Test TUG scale;

[0099] 6.2, Data processing and correlation analysis: Steps 4 and 5 are performed on the multiple groups of human posture data obtained in 6.1, and gait phase division is performed to obtain multiple groups of center of mass stability characteristic indicators in the swing phase. The obtained multiple groups of center of mass stability characteristic indicators in the swing phase and the TUG scale scores obtained in 6.1 are normalized using the Min-Max normalization method to obtain the normalized value X of the center of mass stability characteristic indicators in each swing phase and the TUG scale score. norm , so that they have the same dimension and numerical range for comparison and analysis; the Spearman correlation coefficient analysis method was used for the normalized multiple groups of TUG scale scores and multiple groups of swing phase center of mass stability characteristic indices. The analysis results showed that the five normalized swing phase center of mass stability characteristic indices were significantly positively correlated with the TUG scale scores;

[0100] 6.3, define the S-type utility function: According to 6.2, the five normalized swing phase center of mass stability characteristic indicators are all positive indicators, and the data distribution changes rapidly in the middle range and slowly at both ends. Therefore, the Logistic function is selected as the S-type utility function of each indicator, and its expression is:

[0101]

[0102] Among them, U n (X norm ) is the utility value corresponding to the parameter of each indicator, ranging from 0 to 1; k n is the slope parameter of each indicator curve, which controls the steepness of the curve; the larger the slope, the steeper the curve and the faster the utility value changes;0,n is the midpoint position parameter of each index curve, which determines the translation position of the curve;

[0103] 6.4, define the loss function and optimize the S-type utility function: In order to optimize the k of the S-type utility function, the characteristic index of the center of mass stability in each swing phase n and x 0,n , define the mean square error MSE as the loss function to measure the deviation between the utility value predicted by the model and the actual observed value. By minimizing this loss function, adjust the parameters of the utility function to improve its prediction accuracy; its expression is:

[0104]

[0105] Among them, L n (θ n ) is the loss function of each indicator; X norm,i is the value of the i-th sample after normalization of each index; θ n =(k n ,x 0,n ) are the model parameters of each indicator; x is the number of samples; Y n,i is the actual utility value of the ith sample of each indicator, which is the normalized TUG scale score in 6.2;

[0106] 6.5, Update k and x using gradient descent 0 Parameters: Calculate the loss function of each indicator for k and x 0 Gradient and It means that the loss function varies with k and x 0 The direction and rate of change, select learning rate α = 0.001, and update the parameters according to the gradient descent formula: Among them, k t and x 0 (t) is the parameter value of the tth iteration. Update k and x corresponding to each indicator 0 Parameters until each loss function converges, the convergence condition is: ΔL=|L(θ t+1 )-L(θ t )|<∈, where ∈ is the set threshold, which is 10 -6 ;

[0107] 6.6, establish the dynamic center of mass stability index expression: First, calculate the normalized coefficient of variation CV of the center of mass stability characteristic index in each swing phase obtained in 6.2 norm,n Then the power average synthesis method is used to establish the dynamic center of mass stability index expression, which is shown as follows:

[0108]

[0109] Among them, U(X norm,n ) is the utility value corresponding to the nth indicator input parameter value, m is the power, and DCMSI is the dynamic center of mass stability index;

[0110] 6.7. Find the optimal power m of the dynamic center of mass stability index expression in 6.6: The key task of establishing the dynamic center of mass stability index expression is to determine the optimal power. If the power is too large or too small, the power average synthesis model will lose its practical significance. Therefore, the rough range of the power is determined within [-2,3]. Within this range, the power m is modeled with a step size of 0.1 to obtain the Spearman correlation coefficient between the dynamic center of mass stability index generated by the dynamic center of mass stability index expressions of different powers and the normalized TUG scale score in 6.2. The power corresponding to the maximum correlation is selected as the optimal power m, and is substituted into the dynamic center of mass stability index expression in 6.6 to obtain the optimal power dynamic center of mass stability index expression.

[0111] Step 7, center of mass stability assessment: The center of mass stability of the human body during walking is assessed according to the dynamic center of mass stability index obtained in step 6. The assessment method is as follows:

[0112] The dynamic center of mass stability index is divided into center of mass stability intervals using the quartile analysis method, and the center of mass stability is graded: those below the second quartile are low-risk, those between the second and third quartiles are medium-risk, and those greater than the third quartile are high-risk.

[0113] The solution of this embodiment uses the posture data collected by the inertial measurement unit worn on both feet, and accurately divides the swing phase of the human body during walking through the gait segmentation technology. Subsequently, the posture data collected by the waist inertial measurement unit in the swing phase is deeply analyzed to obtain the center of mass stability characteristic index in the swing phase. The center of mass stability characteristic index in the swing phase is input into the proposed dynamic center of mass stability index expression after data processing, and the dynamic center of mass stability index (Dynamic Center of Mass Stability Index, referred to as DCMSI) is calculated. DCMSI can quantitatively reflect the stability of the center of mass during human walking. The larger the DCMSI value, the more unstable the human body is.

[0114] Example 2

[0115] Reference Figure 1 to Figure 3A system for evaluating the stability of the center of mass during human walking includes an acquisition module, a transmission module and a data processing module, wherein the acquisition module includes a walking test track and three inertial measurement units, the walking test track is used to implement a dynamic walking test paradigm, and the inertial measurement unit is used to collect posture data. The transmission module includes a wireless communication module, which is used to transmit the posture data received from the acquisition module to the data processing module in a wireless communication manner. The data processing module includes a computer, and a host computer software is installed on the computer, which is used to calculate the dynamic center of mass stability index based on the posture data.

[0116] Specifically, refer to Figure 1 The inertial measurement unit is used to collect attitude data, including three-axis acceleration and angular velocity data. The inertial measurement unit sends the collected attitude data to the host computer software through wireless communication. The host computer software is installed on a computer and is used to receive the attitude data transmitted by the inertial measurement unit, and store all data and calculate characteristic indicators after the test is completed.

[0117] Reference Figure 2 , the human posture collection wearing scheme is: wear an inertial measurement unit on the left and right insteps and the posterior lumbar sacrum of the subject, namely the first inertial measurement unit 3 and the second inertial measurement unit 4. The inertial measurement units are worn on the left and right insteps to divide the gait phase of the human body during walking. The third inertial measurement unit 2 worn on the posterior lumbar sacrum is used to characterize the movement of the human body's center of mass. The back of the third inertial measurement unit 2 is pasted with a Velcro hook surface, and is tied with a Velcro elastic elastic band, wherein the waist is a buckle-type Velcro elastic elastic band 1, which can be matched with the Velcro fur surface on the back of the third inertial measurement unit 2 for easy fixation.

[0118] Reference Figure 3 The walking test track includes indicator marks, including a human standing position mark, with the standing position mark as the starting point. Further, a 3-meter to 5-meter straight walking dotted line mark and a return mark are set in front of the standing position mark. During the experiment, the subject is required to stand naturally at the human standing position mark position on the walking test track, with both feet as wide as shoulders as possible.

[0119] The contents described in the embodiments of this specification are merely enumerations of implementation forms of the inventive concept and are for illustrative purposes only. The protection scope of the present invention should not be considered to be limited to the specific forms described in this embodiment, and the protection scope of the present invention also extends to equivalent technical means that can be thought of by ordinary technicians in this field based on the inventive concept.

Claims

1. A method for evaluating the stability of the center of mass of a human body during walking, characterized in that: The method comprises the following steps: Step 1: record the subject information, wear an inertial measurement unit on the subject, and establish communication between the inertial measurement unit and the computer host software; Step 2, start the test, and let the subject implement the dynamic walking test paradigm according to the walking test trajectory, during which the wearable inertial measurement unit is used to collect human posture data; Step 3, ending the test, transmitting the attitude data collected by the three inertial measurement units in step 2 to the computer by wireless communication; Step 4, gait phase division: perform data analysis on the posture data collected by the inertial measurement unit worn on both feet in step 3 to divide the gait phase; Step 5, based on the gait phase obtained in step 4 and the posture data collected by the inertial measurement unit at the posterior lumbar sacrum, an in-depth analysis is performed to obtain a characteristic index of the center of mass stability in the swing phase; Step 6, calculation of dynamic center of mass stability index: data processing is performed on the center of mass stability characteristic index in the swing phase obtained in step 5, and then it is brought into the established S-type utility function of each index to obtain the utility value corresponding to each index parameter, and finally the utility value corresponding to each index parameter is brought into the established dynamic center of mass stability index expression to obtain the dynamic center of mass stability index; Step 7, center of mass stability assessment: The center of mass stability of the human body during walking is assessed according to the dynamic center of mass stability index obtained in step 6.

2. A method for evaluating the stability of the center of mass of a human body during walking as claimed in claim 1, characterized in that: In step 5, the characteristic indexes of the center of mass stability in the swing phase include the center of mass swing trajectory length, the average center of mass swing rate, the center of mass offset area, the average center of mass swing amplitude and the center of mass swing symmetry, which are calculated as follows: (5.1) Center of mass swing trajectory length: The total motion trajectory length of the human body's center of mass during the experiment. The position change of the human body's center of mass is measured by the inertial measurement unit at the sacrum, and the motion trajectory of the center of mass is calculated. First, the initial position of the human body's center of mass is determined, including the coordinates in the three directions of the x, y and z axes. Then, the acceleration of the human body's center of mass in these three directions is measured, and the acceleration is integrated to obtain the position change of the center of mass, which is added to the initial position to obtain the position of the center of mass at a certain moment. Finally, during the entire test time, the Euclidean distance of the center of mass in three-dimensional space is calculated, and these distances are accumulated to obtain the center of mass swing trajectory length L, which is calculated as follows: Where r0 = [r x0 r y0 r z0 ] represents the initial position of the human body's center of mass in the three directions of the x, y and z axes; a=[a x a y a z ] represents the acceleration of the human body's center of mass in the three directions of x, y and z axes, and the position change of the center of mass is obtained by integrating it; T is the total test time; It represents the Euclidean distance of the center of mass of the human body in three-dimensional space, and L is the length of the center of mass swing trajectory; (5.2) Average center of mass swing rate: The average value of the center of mass swing trajectory length per unit time in the total test time. The average speed is obtained by dividing the total length of the center of mass swing trajectory by the total test time. The average center of mass swing rate v avg The calculation formula is as follows: Among them, L t is the length of the center of mass swing trajectory per unit time, and T is the total test time; (5.3) Center of mass offset area: The integral of the change in the projection area of ​​the waist sensor parallel to the back plane of the human body at each moment over time. The center of mass offset area A cnv The calculation formula is as follows: Among them, (x i ,y i , z i ) is the coordinate of the waist sensor at the current moment, (x i+1 ,y i+1 , z i+1 ) is the coordinate of the waist sensor at the next moment; (5.4) Average center of mass swing amplitude: the average distance from the center of mass swing trajectory to the center of mass swing. First, sum the center of mass positions of each sampling point and divide by the number of sampling points to obtain the coordinates of the center of mass swing. Then, calculate the distance from each sampling point to the center of mass swing and find the average of these distances. Center of mass swing C cen It is the average value of the centroid position, and the calculation formula is as follows: Among them, (r x,i , r y,i , r z,i ) is the centroid position of the ith data sample, N is the number of sampling points, Average swing amplitude of center of mass A avg : Among them, C cen，x , C cen，y and C cen，z They represent the x-coordinate, y-coordinate and z-coordinate of the center of mass swing respectively; (5.5) Center of mass swing symmetry: The ratio of the center of mass swing trajectories of the left foot and the right foot when walking. The calculation formula of the center of mass swing symmetry CR is as follows: Among them, L lc and L rc They respectively represent the length of the center of mass swing trajectory when the left foot walks and the length of the center of mass swing trajectory when the right foot walks.

3. A method for evaluating the stability of the center of mass of a human body during walking as claimed in claim 2, characterized in that: In step 6, the dynamic center of mass stability index expression is established as follows: 6.1, Establishing the data set: Implement steps 1-3 on the subjects, collect the human posture data of each subject, and have professional doctors score the subjects using the Timed Up and Go Test TUG scale; 6.2, Data processing and correlation analysis: Steps 4 and 5 are performed on the multiple groups of human posture data obtained in 6.1, and gait phase division is performed to obtain multiple groups of center of mass stability characteristic indicators in the swing phase. The obtained multiple groups of center of mass stability characteristic indicators in the swing phase and the TUG scale scores obtained in 6.1 are normalized using the Min-Max normalization method to obtain the normalized value X of the center of mass stability characteristic indicators in each swing phase and the TUG scale score. norm , so that they have the same dimension and numerical range; the Spearman correlation coefficient analysis method was used for the normalized multiple groups of TUG scale scores and multiple groups of swing phase center of mass stability characteristic indicators. The analysis results showed that the five normalized swing phase center of mass stability characteristic indicators were significantly positively correlated with the TUG scale scores; 6.3, define the S-type utility function: According to 6.2, the five normalized swing phase center of mass stability characteristic indicators are all positive indicators, and the data distribution changes rapidly in the middle range and slowly at both ends. Therefore, the Logistic function is selected as the S-type utility function of each indicator, and its expression is: Among them, U n (X norm ) is the utility value corresponding to the parameter of each indicator, ranging from 0 to 1; k n is the slope parameter of each indicator curve, which controls the steepness of the curve; the larger the slope, the steeper the curve and the faster the utility value changes; 0,n is the midpoint position parameter of each index curve, which determines the translation position of the curve; 6.4, define the loss function and optimize the S-type utility function: In order to optimize the k of the S-type utility function, the characteristic index of the center of mass stability in each swing phase n and x 0，n , define the mean square error MSE as the loss function to measure the deviation between the utility value predicted by the model and the actual observed value. By minimizing this loss function, adjust the parameters of the utility function to improve its prediction accuracy; its expression is: Among them, L n (θ n ) is the loss function of each indicator; X norm,i is the value of the i-th sample after normalization of each index; θ n =(k n , x 0，n ) are the model parameters of each indicator; N is the number of samples; Y n，i is the actual utility value of the ith sample of each indicator, which is the normalized TUG scale score in 6.2; 6.5, Update the parameters k and x0 using the gradient descent method: Calculate the gradient of the loss function of each indicator with respect to k and x0 and Indicates the direction and rate at which the loss function changes with k and x0, sets the learning rate α, and updates the parameters according to the gradient descent formula: Among them, k t and x0(t) are the parameter values ​​of the tth iteration; update the k and x0 parameters corresponding to each indicator until each loss function converges, and the convergence condition is: ΔL=|L(θ t+1 )-L(θ t )|<∈, where ∈ is the set threshold; 6.6, establish the dynamic center of mass stability index expression: First, calculate the normalized coefficient of variation CV of the center of mass stability characteristic index in each swing phase obtained in 6.2 norm,n , and then the power average synthesis method is used to establish the dynamic center of mass stability index expression, which is shown as follows: Among them, U(X norm，n ) is the utility value corresponding to the nth indicator input parameter value, m is the power, and DCMSI is the dynamic center of mass stability index; 6.

7. Find the optimal power m of the dynamic center of mass stability index expression in 6.6: The initial range of the power is determined to be within [-2,3]; within this range, the power m is modeled with a step size of 0.1 to obtain the Spearman correlation coefficient between the dynamic center of mass stability index generated by the dynamic center of mass stability index expressions of different powers and the normalized TUG scale score in 6.2, and the power corresponding to the maximum correlation is selected as the optimal power m, which is substituted into the dynamic center of mass stability index expression in 6.6 to obtain the optimal power dynamic center of mass stability index expression.

4. A method for evaluating the stability of the center of mass of a human body during walking as claimed in any one of claims 1 to 3, characterized in that: In step 2, the walking test trajectory includes a human standing position mark, a 3-meter to 5-meter straight walking dotted line mark, and a turnaround mark; when implementing the human balance ability detection paradigm described in step 2, the subject first stands on the human standing position mark, and when the subject hears the instruction, the dynamic walking test paradigm begins.

5. A method for evaluating the stability of the center of mass of a human body during walking as claimed in claim 4, characterized in that: The dynamic walking test paradigm is a 3-meter to 5-meter straight-line return walking test, which is used to evaluate the center of mass stability of the human body during walking.

6. A method for evaluating the stability of the center of mass of a human body during walking as claimed in claim 4, characterized in that: In step 2, the posture data at least includes three-axis acceleration and angular velocity.

7. A method for evaluating the stability of the center of mass of a human body during walking as claimed in claim 6, characterized in that: The three-axis acceleration includes the left foot movement acceleration, the right foot movement acceleration and the lumbar sacrum movement acceleration; the angular velocity includes the left foot movement angular velocity, the right foot movement angular velocity and the lumbar sacrum movement angular velocity.

8. A method for evaluating the stability of the center of mass of a human body during walking as claimed in any one of claims 1 to 3, characterized in that: In step 4, the gait phase is calculated as follows: first, the three-axis acceleration and angular velocity data collected from the sensors at the left foot, the right foot and the lumbar sacrum are subjected to a first-order Butterworth low-pass filter to remove noise and smooth the data; by analyzing the characteristic changes of the pitch angle in the angular velocity data, the key events in the gait cycle are identified: heel touch HS and toe lift TO; the above two gait events mark the transition of the gait from one stage to another; the heel touch marks the beginning of the stance phase, and the toe lift marks the entry of the leg into the swing phase; based on the identified gait events, the gait cycle is divided into different phases.

9. A method for evaluating the stability of the center of mass of a human body during walking as claimed in any one of claims 1 to 3, characterized in that: In step 7, the center of mass stability assessment method is: the obtained dynamic center of mass stability index is divided into center of mass stability intervals using the quartile analysis method, and the center of mass stability is graded: below the second quartile is a low risk of center of mass stability, between the second and third quartiles is a medium risk of center of mass stability, and greater than the third quartile is a high risk of center of mass stability.

10. A system for implementing the method for evaluating the center of mass stability of a human body during walking as claimed in claim 1, characterized in that: The system includes a collection module, a transmission module and a data processing module, wherein: The acquisition module includes a walking test track and an inertial measurement unit, wherein the walking test track is used to implement a dynamic walking test paradigm, and the inertial measurement unit is used to acquire posture data; The transmission module includes a wireless communication module, which is used to transmit the posture data received from the acquisition module to the data processing module in a wireless communication manner; The data processing module includes a computer, on which upper computer software is installed, which is used to calculate the characteristic index of center of mass stability in the swing phase based on posture data, and obtain a dynamic center of mass stability index based on the analysis of the characteristic index of center of mass stability in the swing phase to evaluate the center of mass stability during human walking.

Images