Quantitative lower limb rehabilitation evaluation system and method

Abstract

The invention discloses a quantitative lower limb rehabilitation assessment system and method, relates to the technical field of lower limb rehabilitation assessment, and solves the technical problems that assessment of lower limb rehabilitation is not comprehensive enough and assessment results of lower limb rehabilitation are not accurate enough due to the fact that assessment of dynamic characteristics in walking is not considered in the prior art. The method comprises the following steps: acquiring walking data and motion data of a subject in a walking process in real time; gait data of the subject in the walking process are calculated based on the walking data; performing time domain and frequency domain analysis on the motion data to obtain motion signal data; evaluating and analyzing the gait data to obtain evaluation data; a rehabilitation score is obtained through a weighted summation method based on the gait data, the motion signal data and the assessment data, a lower limb rehabilitation assessment report is generated based on the rehabilitation score, and the technical problem that the assessment accuracy of lower limb rehabilitation is not high in the prior art is solved.

Description

Technical Field

[0001] The present invention belongs to the field of lower limb rehabilitation assessment, and specifically relates to a quantitative lower limb rehabilitation assessment system and method. Background Art

[0002] With the aging of the population, the demand for lower limb rehabilitation is increasing. There are many reasons for lower limb dysfunction, including sports injuries, nerve damage, and other diseases. Lower limb rehabilitation is an important part of recovery after many diseases and surgeries. Accurate lower limb function assessment helps to develop an effective rehabilitation plan, thereby improving the patient's quality of life. Therefore, the assessment of lower limb functional rehabilitation is of great significance in the elderly population and rehabilitation patients. Effective assessment of lower limb functional rehabilitation is of great significance to helping the elderly and the disabled.

[0003] The prior art (the invention application with the publication number CN117503117A) discloses a walking ability evaluation device, equipment and method, wherein the walking ability evaluation device comprises: a data input unit, an equipment configuration unit, a data display unit and a data analysis unit; the data input unit is used to obtain basic information of the subject, and input the basic information of the subject into a preset step length prediction model to output a predicted step length; the equipment configuration unit can output guidance information according to the predicted step length to guide the subject to perform a walking ability test, and receive gait information generated during the test; the application can quickly and accurately calculate and analyze the parameters such as the starting time, step length, support time, swing time, pressure center swing and bipedal difference of the subject, and realize the comprehensive evaluation of the symmetry, periodicity and stability of the walking process; the prior art mainly evaluates the symmetry, periodicity and stability of the walking process through gait information, and does not consider the evaluation of the dynamic characteristics of walking, such as the intensity, energy, stability, regularity, frequency, rhythm, energy concentration, complexity and coordination of the movement, resulting in an incomplete evaluation of lower limb rehabilitation, and thus making the evaluation result of lower limb rehabilitation inaccurate.

[0004] Therefore, the present invention proposes a quantitative lower limb rehabilitation assessment system and method to solve the above problems. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a quantitative lower limb rehabilitation assessment system and method, which is used to solve the technical problem that the prior art mainly evaluates the symmetry, periodicity and stability during walking through gait information, but does not consider the evaluation of the dynamic characteristics of walking, such as the intensity, energy, smoothness, regularity, frequency, rhythm, energy concentration, complexity and coordination of movement, resulting in an incomplete assessment of lower limb rehabilitation and thus an inaccurate assessment result of lower limb rehabilitation.

[0006] To achieve the above-mentioned object, a first aspect of the present invention provides a quantitative lower limb rehabilitation assessment system, comprising: a data acquisition module, a data analysis module and a lower limb rehabilitation assessment module;

[0007] Data acquisition module: used to collect the walking data and motion data of the subject in real time during walking;

[0008] Data analysis module: Calculate the gait data of the subject during walking based on the walking data; obtain motion signal data by performing time domain and frequency domain analysis on the motion data; and,

[0009] By evaluating and analyzing the gait data, evaluation data is obtained;

[0010] Lower limb rehabilitation assessment module: The rehabilitation score is obtained by weighted summation based on gait data, motion signal data and assessment data, and a lower limb rehabilitation assessment report is generated based on the rehabilitation score.

[0011] Preferably, the data analysis module is communicated and / or electrically connected to the data acquisition module and the lower limb rehabilitation assessment module respectively.

[0012] Preferably, the real-time collection of walking data and motion data of the subject during walking includes:

[0013] The walking data of the subject during walking is obtained through a flexible array pressure walkway; wherein the walking data includes: the time when the walking instruction is issued, the moment of lifting the foot, the moment of touching the ground, the moment of leaving the ground, the moment of landing, the position at the moment of landing, the position at the moment of leaving the ground, the plantar pressure center at the moment of landing, and the plantar pressure center at the moment of leaving the ground;

[0014] The motion data of the subjects during walking is collected in real time through the posture sensor; the motion data includes: hip joint activity Euler angle, knee joint activity Euler angle, ankle joint activity Euler angle, hip joint acceleration, knee joint acceleration, ankle joint acceleration, hip joint angular velocity, knee joint angular velocity, ankle joint angular velocity.

[0015] It should be noted that the flexible array pressure walkway is composed of a standing starting area, a walking test area, and an array-type acquisition control unit; flexible array-type pressure sensors are arranged in the standing starting area and the walking test area to sense the plantar pressure information of the subject during the standing-walking process; the array-type acquisition control unit is composed of a matrix scanning drive circuit, a scanning acquisition circuit, and a data communication circuit;

[0016] The posture sensors are tied to the subject's legs at the lower 1 / 3 of the left and right thighs, the upper 1 / 3 of the left and right tibias, and the second metatarsal, and are used to detect the joint activities of both lower limbs during the subject's standing and walking process; the posture sensor is a high-performance miniature wireless transmission posture sensor that integrates a three-axis gyroscope, a three-axis accelerometer, and a three-axis magnetometer, and can measure and record the subject's lower limb joint activity in real time during walking.

[0017] Preferably, the hip joint activity Euler angle is the angle between the thigh and the vertical axis in the sagittal plane, and the angle is defined as "positive" when flexing forward and as "negative" when extending backward; the knee joint activity Euler angle is the angle between the calf and the extension line of the thigh, and the angle value is defined to be always "positive"; the ankle joint activity Euler angle is the angle between the fifth toe of the foot and the normal line perpendicular to the calf, and the angle is defined as "positive" when dorsiflexing upward and as "negative" when plantar flexing downward.

[0018] Preferably, the step of calculating the gait data of the subject during walking based on the walking data includes:

[0019] Gait parameters and gait assessment parameters;

[0020] Gait parameters include:

[0021] Reaction time TA = foot lifting time T2 - walking command issuance time T1;

[0022] Action time TB = time of leaving the ground T4 - time of lifting the foot T2;

[0023] Gait cycle TF = ground contact time T7 - ​​ground contact time T3;

[0024] Pace S = stride W / (landing time T6 - ground departure time T4);

[0025] Step frequency F = 1 / gait period TF;

[0026] Gait assessment parameters include:

[0027] Starting time TC = ground-leave time T4 - walking command issuance time T1;

[0028] Support time TD = landing time T6 - ground contact time T5;

[0029] Swing time TE = landing time T5 - lift-off time T4;

[0030] Step length L = the position at the time of landing T6 - the position at the time of lifting off T4;

[0031] Stride W = plantar pressure center at landing time T6 - plantar pressure center at lift-off time T4;

[0032] Bipedal difference DX=left foot parameter / right foot parameter, wherein both left foot parameter and right foot parameter include: support time TD, swing time TE, step length L, and stride W.

[0033] By inputting the gait parameters into the pre-trained LSTM model, the feature vector corresponding to the gait parameters is obtained.

[0034] It should be noted that the gait cycle refers to the time required for a complete gait cycle, that is, the gait cycle is the time period from the start of one side foot touching the ground to the end of the same side foot touching the ground again;

[0035] The pre-trained LSTM model is trained based on historical gait data and corresponding feature vectors; the LSTM network is suitable for capturing the time dependency of gait data. First, the gait data is normalized to convert data of different magnitudes into a unified dimension to ensure the training effect of the model; then, the gait data is used as the input of the LSTM to learn the dynamic changes of each step and its trend over time; finally, the LSTM model outputs the time series feature vector of the gait data after processing through the time dimension.

[0036] The present invention utilizes a flexible array pressure walkway and a posture sensor to comprehensively collect walking data and motion data, including multiple key time points such as the time when a walking instruction is issued, the moment when the foot is lifted, and the moment when the foot touches the ground, as well as multiple parameters such as the activity angles, accelerations, and angular velocities of the hip joint, knee joint, and ankle joint, thereby improving the comprehensiveness and reliability of data collection. The multi-dimensional data collection method ensures that the walking state of the subject is fully reflected from multiple angles, avoiding errors that may be caused by a single data source. At the same time, the flexible array pressure walkway can provide position information of the plantar pressure center, which is helpful for accurately evaluating gait characteristics. The posture sensor can provide multi-dimensional motion data, which increases the dimension of evaluation and makes the evaluation more accurate.

[0037] Preferably, the time domain and frequency domain analysis of the motion data includes:

[0038] S1: extracting motion data collected in real time;

[0039] S2: Obtain corresponding motion signal data by performing time domain analysis and frequency domain analysis on the motion data; wherein the motion signal data includes: peak-to-peak value, root mean square, skewness, kurtosis, approximate entropy, average amplitude change rate, main frequency component, signal power, frequency index and spectrum entropy;

[0040] S3: By inputting the motion signal data into a pre-trained feature extraction model, a feature vector corresponding to the motion signal data is obtained.

[0041] It should be noted that the peak-to-peak value reflects the amplitude of joint movement. The greater the intensity of movement, the higher the peak-to-peak value. The root mean square indicates the energy of joint movement. The greater the energy, the higher the root mean square value. Skewness and kurtosis are used to evaluate the smoothness of movement. Skewness reflects symmetry, and kurtosis reflects whether there are sudden changes in movement. Approximate entropy reflects the regularity and coordination of movement. A lower value indicates a more stable movement, while a higher value indicates a complex or unstable movement. Average amplitude change rate shows the frequency of movement changes. The greater the change, the more unstable the movement may be. The main frequency component is used to evaluate the rhythm and periodicity of gait. The higher the frequency, the faster the rhythm. Signal power reflects the concentration of movement energy. The higher the power, the more intense the movement. Frequency index and spectral entropy measure the coordination and complexity of movement. The higher the spectral entropy, the more complex the movement, indicating an increase in irregularity.

[0042] The pre-trained feature extraction model is obtained by training an artificial intelligence model based on historical gait data, historical motion signal data, feature vectors corresponding to historical gait data, and feature vectors corresponding to historical motion signal data; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network.

[0043] The present invention extracts multiple features such as peak-to-peak value, root mean square, skewness, kurtosis, approximate entropy, average amplitude change rate, main frequency component, signal power, frequency index and spectrum entropy by analyzing gait data and motion data in time domain and frequency domain. These features cover the amplitude, energy, stability, regularity, frequency, rhythm, energy concentration, complexity and coordination of motion, making the evaluation more comprehensive.

[0044] Preferably, the evaluating and analyzing the gait data includes:

[0045] By evaluating and analyzing the gait evaluation parameters, evaluation parameters are obtained; wherein the evaluation parameters include: periodicity, stability and symmetry;

[0046] The periodicity of the gait is determined by calculating the ratio of the swing time TE to the support time TD; the closer the ratio of each step is, the better the periodicity of the gait is;

[0047] The stability of the gait is determined based on the starting time TC, the stride W and the step length L; wherein, the shorter the starting time, the smaller the stride and the more consistent the step length, the better the stability of the gait;

[0048] The symmetry of the two feet is determined based on the difference DX of the two feet. The closer the value of the difference DX of the two feet is to 1, the better the symmetry of the two feet. The difference contribution ratio of the support time TD, the swing time TE, the step length L, and the stride W is 1:1:1:1.

[0049] By inputting the evaluation parameters into the pre-trained feature extraction model, the feature vector corresponding to the evaluation parameters is obtained.

[0050] The present invention obtains evaluation parameters, including periodicity, stability and symmetry, by evaluating and analyzing gait data, thereby further enriching the evaluation dimensions and making the evaluation results more comprehensive and accurate.

[0051] Preferably, the rehabilitation score is obtained by weighted summation based on the gait data, motion signal data and evaluation data, including:

[0052] The feature vector corresponding to the gait parameter is marked as Xi, the feature vector corresponding to the motion signal data is marked as Yj, and the feature vector corresponding to the evaluation parameter is marked as Zg; wherein Xi represents the feature vector corresponding to the i-th gait parameter; Yj represents the feature vector corresponding to the j-th motion signal data; Zg represents the feature vector corresponding to the g-th evaluation parameter;

[0053] The rehabilitation score is calculated by the formula: K = ∑(Ai×Xi)+∑(Bj×Yj)+∑(Cg×Zg); wherein, i = 1, 2, …, n, n is the total number of eigenvectors corresponding to gait parameters; j = 1, 2, …, m, m is the total number of eigenvectors corresponding to motion signal data; g = 1, 2, …, r, r is the total number of eigenvectors corresponding to estimation parameters; Ai, Bj, Cg are weight coefficients, and ∑Ai+∑Bj+∑Cg=1.

[0054] It should be noted that the weight coefficient is set by historical data or actual experience; Ai is the weight coefficient of the feature vector corresponding to the i-th gait parameter; Bj is the weight coefficient of the feature vector corresponding to the j-th motion signal data; Cg is the weight coefficient of the feature vector corresponding to the g-th estimation parameter.

[0055] The present invention obtains the rehabilitation score by the weighted summation method, which integrates the gait parameters, motion signal data and evaluation data, making the evaluation result more comprehensive and accurate, and at the same time being able to fully reflect the rehabilitation status of the subject, ensuring the comprehensiveness and accuracy of the evaluation result.

[0056] A second aspect of the present invention provides a quantitative lower limb rehabilitation assessment method, comprising:

[0057] Step 1: Collect the walking data and motion data of the subject in real time during walking;

[0058] Step 2: Calculate the gait data of the subject during walking based on the walking data;

[0059] Step 3: Obtain motion signal data by performing time domain and frequency domain analysis on the motion data;

[0060] Step 4: Obtain evaluation data by evaluating and analyzing the gait data;

[0061] Step 5: Obtain a rehabilitation score through a weighted summation method based on the gait data, motion signal data, and evaluation data, and generate a lower limb rehabilitation evaluation report based on the rehabilitation score.

[0062] Compared with the prior art, the present invention has the following beneficial effects:

[0063] The prior art mainly evaluates the symmetry, periodicity and stability of walking through gait information, but does not consider the evaluation of dynamic characteristics of walking, such as intensity, energy, smoothness, regularity, frequency, rhythm, energy concentration, complexity and coordination of movement, resulting in an incomplete evaluation of lower limb rehabilitation, and thus an inaccurate evaluation result of lower limb rehabilitation. The present invention collects walking data and motion data of the subject in real time during walking; calculates the gait data of the subject during walking based on the walking data; obtains motion signal data by performing time domain and frequency domain analysis on the motion data; obtains evaluation data by evaluating and analyzing the gait data; and obtains a rehabilitation score by a weighted summation method based on the gait data, motion signal data and evaluation data, and generates a lower limb rehabilitation evaluation report based on the rehabilitation score, thereby solving the technical problem that the prior art does not have high evaluation accuracy of lower limb rehabilitation. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0065] Figure 1 A schematic diagram of a system module according to an embodiment of the present invention;

[0066] Figure 2 A schematic diagram of a walking assessment device according to an embodiment of the present invention;

[0067] Figure 3 A schematic diagram of walking time division according to an embodiment of the present invention;

[0068] Figure 4 Schematic diagram of the method steps of an embodiment of the present invention.

[0069] In the attached drawings, 100 is a flexible array pressure walkway; 101 is a standing starting area; 102 is a walking test area; 103 is an array acquisition control unit; 201, 202, 203, 204, 205, 206 are posture sensors; 300 is a lower limb rehabilitation comprehensive test and analysis software. DETAILED DESCRIPTION

[0070] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] See also Figure 1 , the first aspect of the present invention provides a quantitative lower limb rehabilitation assessment system, including: a data acquisition module, a data analysis module and a lower limb rehabilitation assessment module;

[0072] Data acquisition module: used to collect the walking data and motion data of the subject in real time during walking;

[0073] Data analysis module: Calculate the gait data of the subject during walking based on the walking data; obtain motion signal data by performing time domain and frequency domain analysis on the motion data; and,

[0074] By evaluating and analyzing the gait data, evaluation data is obtained;

[0075] Lower limb rehabilitation assessment module: The rehabilitation score is obtained by weighted summation based on gait data, motion signal data and assessment data, and a lower limb rehabilitation assessment report is generated based on the rehabilitation score.

[0076] See also Figure 2 , real-time collection of walking data and motion data of the subject during walking, including:

[0077] The walking data of the subject during walking is obtained through a flexible array pressure walkway; wherein the walking data includes: the time when the walking instruction is issued, the moment of lifting the foot, the moment of touching the ground, the moment of leaving the ground, the moment of landing, the position at the moment of landing, the position at the moment of leaving the ground, the plantar pressure center at the moment of landing, and the plantar pressure center at the moment of leaving the ground;

[0078] The motion data of the subjects during walking is collected in real time through the posture sensor; the motion data includes: hip joint activity Euler angle, knee joint activity Euler angle, ankle joint activity Euler angle, hip joint acceleration, knee joint acceleration, ankle joint acceleration, hip joint angular velocity, knee joint angular velocity, ankle joint angular velocity.

[0079] The hip joint activity Euler angle is the angle between the thigh and the vertical axis in the sagittal plane, and the angle is defined as "positive" when flexing forward and "negative" when extending backward; the knee joint activity Euler angle is the angle between the calf and the extension line of the thigh, and the angle value is defined to be always "positive"; the ankle joint activity Euler angle is the angle between the fifth toe of the foot and the normal line perpendicular to the calf, and the angle is defined as "positive" when dorsiflexing upward and "negative" when plantar flexing downward.

[0080] See also Figure 3 , based on the walking data, the gait data of the subject during walking is calculated, including:

[0081] Gait parameters and gait assessment parameters;

[0082] Gait parameters include:

[0083] Reaction time TA = foot lifting time T2 - walking command issuance time T1;

[0084] Action time TB = time of leaving the ground T4 - time of lifting the foot T2;

[0085] Gait cycle TF = ground contact time T7 - ​​ground contact time T3;

[0086] Pace S = stride W / (landing time T6 - ground departure time T4);

[0087] Step frequency F = 1 / gait period TF;

[0088] Gait assessment parameters include:

[0089] Starting time TC = ground-leave time T4 - walking command issuance time T1;

[0090] Support time TD = landing time T6 - ground contact time T5;

[0091] Swing time TE = landing time T5 - lift-off time T4;

[0092] Step length L = the position at the time of landing T6 - the position at the time of lifting off T4;

[0093] Stride W = plantar pressure center at landing time T6 - plantar pressure center at lift-off time T4;

[0094] Bipedal difference DX=left foot parameter / right foot parameter, wherein both left foot parameter and right foot parameter include: support time TD, swing time TE, step length L, and stride W.

[0095] By inputting the gait parameters into the pre-trained LSTM model, the feature vector corresponding to the gait parameters is obtained.

[0096] By performing time domain and frequency domain analysis on motion data, including:

[0097] S1: extracting motion data collected in real time;

[0098] S2: Obtain corresponding motion signal data by performing time domain analysis and frequency domain analysis on the motion data; wherein the motion signal data includes: peak-to-peak value, root mean square, skewness, kurtosis, approximate entropy, average amplitude change rate, main frequency component, signal power, frequency index and spectrum entropy;

[0099] S3: By inputting the motion signal data into a pre-trained feature extraction model, a feature vector corresponding to the motion signal data is obtained.

[0100] By evaluating and analyzing gait data, including:

[0101] By evaluating and analyzing the gait evaluation parameters, evaluation parameters are obtained; wherein the evaluation parameters include: periodicity, stability and symmetry;

[0102] The periodicity of the gait is determined by calculating the ratio of the swing time TE to the support time TD; the closer the ratio of each step is, the better the periodicity of the gait is;

[0103] The stability of the gait is determined based on the starting time TC, the stride W and the step length L; wherein, the shorter the starting time, the smaller the stride and the more consistent the step length, the better the stability of the gait;

[0104] The symmetry of the two feet is determined based on the difference DX of the two feet. The closer the value of the difference DX of the two feet is to 1, the better the symmetry of the two feet. The difference contribution ratio of the support time TD, the swing time TE, the step length L, and the stride W is 1:1:1:1.

[0105] By inputting the evaluation parameters into the pre-trained feature extraction model, the feature vector corresponding to the evaluation parameters is obtained.

[0106] The rehabilitation score is obtained by weighted summation based on gait data, motion signal data and assessment data, including:

[0107] The feature vector corresponding to the gait parameter is marked as Xi, the feature vector corresponding to the motion signal data is marked as Yj, and the feature vector corresponding to the evaluation parameter is marked as Zg; wherein Xi represents the feature vector corresponding to the i-th gait parameter; Yj represents the feature vector corresponding to the j-th motion signal data; Zg represents the feature vector corresponding to the g-th evaluation parameter;

[0108] The rehabilitation score is calculated by the formula: K = ∑(Ai×Xi)+∑(Bj×Yj)+∑(Cg×Zg); wherein, i = 1, 2, …, n, n is the total number of eigenvectors corresponding to gait parameters; j = 1, 2, …, m, m is the total number of eigenvectors corresponding to motion signal data; g = 1, 2, …, r, r is the total number of eigenvectors corresponding to estimation parameters; Ai, Bj, Cg are weight coefficients, and ∑Ai+∑Bj+∑Cg=1.

[0109] For example, suppose there is a 50-year-old male patient who is undergoing rehabilitation training three months after surgery and needs to conduct a quantitative assessment of his lower limb rehabilitation. The details are as follows:

[0110] 1. Data acquisition module;

[0111] Walking data collection:

[0112] Equipment configuration: Use a flexible array pressure walkway, including a standing starting area and a walking test area, equipped with an array-type acquisition control unit.

[0113] Data collection: Record the process of the patient walking from the standing starting area to the walking test area, and obtain the time of issuing the walking command, the moment of lifting the foot, the moment of touching the ground, the moment of leaving the ground, the moment of landing, the position at the moment of landing, the position at the moment of leaving the ground, the plantar pressure center at the moment of landing, and the plantar pressure center at the moment of leaving the ground.

[0114] Sports data collection:

[0115] Equipment configuration: Use posture sensors, tied to the patient's left and right thighs at the lower 1 / 3, left and right tibia at the upper 1 / 3, and the second metatarsal.

[0116] Data collection: record the Euler angles of hip joint activity, knee joint activity, ankle joint activity, hip joint acceleration, knee joint acceleration, ankle joint acceleration, hip joint angular velocity, knee joint angular velocity, and ankle joint angular velocity during the patient's walking.

[0117] 2. Data analysis module;

[0118] Gait data calculation:

[0119] Gait parameters:

[0120] Reaction time TA = foot lifting time T2 - walking command issuance time T1;

[0121] Action time TB = time of leaving the ground T4 - time of lifting the foot T2;

[0122] Gait cycle TF = ground contact time T7 - ​​ground contact time T3;

[0123] Pace S = stride W / (landing time T6 - ground departure time T4);

[0124] Step frequency F = 1 / gait period TF;

[0125] Gait assessment parameters:

[0126] Starting time TC = ground-leave time T4 - walking command issuance time T1;

[0127] Support time TD = landing time T6 - ground contact time T5;

[0128] Swing time TE = landing time T5 - lift-off time T4;

[0129] Step length L = the position at the time of landing T6 - the position at the time of lifting off T4;

[0130] Stride W = plantar pressure center at landing time T6 - plantar pressure center at lift-off time T4;

[0131] Bipedal difference DX = left foot parameters / right foot parameters (support time TD, swing time TE, step length L, stride W);

[0132] Motion signal data extraction:

[0133] By performing time domain analysis and frequency domain analysis on the motion data, corresponding motion signal data is obtained;

[0134] Time domain characteristics:

[0135] Peak-to-peak value: reflects the amplitude of joint movement. The greater the exercise intensity, the higher the peak-to-peak value.

[0136] Root mean square: represents the energy of joint movement. The greater the energy, the higher the root mean square value.

[0137] Skewness and kurtosis: used to assess the smoothness of movement; skewness reflects symmetry, and kurtosis reflects whether there are sudden changes in movement;

[0138] Approximate entropy: reflects the regularity and coordination of movement. Lower values ​​indicate more stable movement, while higher values ​​indicate complex or unstable movement.

[0139] Average amplitude change rate: shows the frequency of motion changes. The greater the change, the more unstable the motion may be.

[0140] Frequency domain characteristics:

[0141] Main frequency component: used to evaluate the rhythm and periodicity of gait. The higher the frequency, the faster the rhythm.

[0142] Signal power: reflects the concentration of motion energy. The higher the power, the stronger the motion.

[0143] Frequency index and spectral entropy: measure the coordination and complexity of movement; the higher the spectral entropy, the more complex the movement, indicating increased irregularity;

[0144] Evaluation parameters:

[0145] By evaluating and analyzing the gait evaluation parameters, the evaluation parameters are obtained; the specific parameters are as follows:

[0146] The periodicity of the gait is determined by calculating the ratio of the swing time TE to the support time TD; the closer the ratio of each step is, the better the periodicity of the gait is;

[0147] The stability of the gait is determined based on the starting time TC, the stride W and the step length L; wherein, the shorter the starting time, the smaller the stride and the more consistent the step length, the better the stability of the gait;

[0148] The symmetry of the two feet is determined based on the difference DX of the two feet. The closer the value of the difference DX of the two feet is to 1, the better the symmetry of the two feet. The difference contribution ratio of the support time TD, the swing time TE, the step length L, and the stride W is 1:1:1:1.

[0149] For example: Assume the left foot parameters are:

[0150] Support time TD = 0.4 seconds;

[0151] Swing time TE = 0.2 seconds;

[0152] Step length L = 0.2 m;

[0153] Stride W = 0.1 m;

[0154] Assume the right foot parameters are:

[0155] Support time TD = 0.4 seconds;

[0156] Swing time TE = 0.2 seconds;

[0157] Step length L = 0.2 m;

[0158] Stride W = 0.1 m;

[0159] Therefore, the bipedal difference DX is:

[0160] DX_TD=left foot TD / right foot TD=1;

[0161] DX_TE = left foot TE / right foot TE = 1;

[0162] DX_L = left foot L / right foot L = 1;

[0163] DX_W = left footstep W / right footstep W = 1;

[0164] Combined calculation:

[0165] Since the variance contribution ratio of each parameter is 1:1:1:1, this means that we can combine these variance DX values ​​by simple averaging.

[0166] DX total = (1 + 1 + 1 + 1) / 4 = 1;

[0167] Feature vector generation:

[0168] The gait parameters are input into the pre-trained LSTM model to obtain the feature vector corresponding to the gait parameters.

[0169] The motion signal data is input into a pre-trained feature extraction model (such as CNN or DBN) to obtain a feature vector corresponding to the motion signal data.

[0170] The evaluation parameters are input into the pre-trained feature extraction model to obtain the feature vector corresponding to the evaluation parameters.

[0171] 3. Lower limb rehabilitation assessment module;

[0172] Recovery score calculation:

[0173] Feature vector notation:

[0174] The feature vector corresponding to the gait parameters is labeled Xi;

[0175] The feature vector corresponding to the motion signal data is marked as Yj;

[0176] The eigenvector corresponding to the evaluation parameter is labeled Zg;

[0177] Calculation formula: K = ∑(Ai×Xi)+∑(Bj×Yj)+∑(Cg×Zg);

[0178] Assume the eigenvector is:

[0179] Xi = [0.5, 1.3, 1.2, 0.83, 0.83] (gait parameter feature vector);

[0180] Yj = [0.15, 0.08, 0.05, 2.2, 0.02, 0.005, 1.5, 0.03, 0.7, 0.05] (motion signal data feature vector);

[0181] Zg = [0.5, 0.4, 0.2, 0.2, 0.1, 1] (evaluation parameter feature vector);

[0182] Assume the weight coefficient is:

[0183] Ai=[0.05,0.07,0.06,0.03,0.04];

[0184] Bj=[0.02,0.07,0.03,0.07,0.04,0.05,0.05,0.03,0.06,0.04];

[0185] Cg=[0.05,0.03,0.04,0.06,0.08,0.03];

[0186] Substituting the above eigenvector and weight coefficient into the formula K = ∑(Ai×Xi)+∑(Bj×Yj)+∑(Cg×Zg), we can get K≈0.63.

[0187] Generate an assessment report:

[0188] The rehabilitation score K is calculated and a comprehensive lower limb rehabilitation assessment report is generated, including the overall rehabilitation score and the individual scores and details of each part. The report covers the results of gait analysis, joint range of motion analysis, periodicity, stability and symmetry assessment, and provides recommended rehabilitation measures and improvement plans. These data provide an objective reference for clinicians and rehabilitation therapists, which is conducive to the formulation and adjustment of rehabilitation plans, thereby improving rehabilitation effects.

[0189] See also Figure 4 The second aspect of the present invention provides a quantitative lower limb rehabilitation assessment method, comprising:

[0190] Step 1: Collect the walking data and motion data of the subject in real time during walking;

[0191] Step 2: Calculate the gait data of the subject during walking based on the walking data;

[0192] Step 3: Obtain motion signal data by performing time domain and frequency domain analysis on the motion data;

[0193] Step 4: Obtain evaluation data by evaluating and analyzing the gait data;

[0194] Step 5: Obtain a rehabilitation score through a weighted summation method based on the gait data, motion signal data, and evaluation data, and generate a lower limb rehabilitation evaluation report based on the rehabilitation score.

[0195] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0196] Working principle of the present invention:

[0197] The present invention collects walking data and motion data of a subject in real time during walking; calculates gait data of the subject during walking based on the walking data; obtains motion signal data by performing time domain and frequency domain analysis on the motion data; obtains evaluation data by performing evaluation analysis on the gait data; and obtains a rehabilitation score by a weighted summation method based on the gait data, motion signal data and evaluation data, and generates a lower limb rehabilitation evaluation report based on the rehabilitation score, thereby solving the technical problem that the prior art has low evaluation accuracy for lower limb rehabilitation.

[0198] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A quantitative lower limb rehabilitation assessment system, characterized in that: include: Data collection module, data analysis module and lower limb rehabilitation assessment module; Data acquisition module: used to collect the walking data and motion data of the subject in real time during walking; Data analysis module: Calculate the gait data of the subject during walking based on the walking data; obtain motion signal data by performing time domain and frequency domain analysis on the motion data; as well as, By evaluating and analyzing the gait data, evaluation data is obtained; Lower limb rehabilitation assessment module: The rehabilitation score is obtained by weighted summation based on gait data, motion signal data and assessment data, and a lower limb rehabilitation assessment report is generated based on the rehabilitation score.

2. A quantitative lower limb rehabilitation assessment system according to claim 1, characterized in that: The real-time collection of walking data and motion data of the subject during walking includes: The walking data of the subject during walking is obtained through a flexible array pressure walkway; wherein the walking data includes: the time when the walking instruction is issued, the moment of lifting the foot, the moment of touching the ground, the moment of leaving the ground, the moment of landing, the position at the moment of landing, the position at the moment of leaving the ground, the plantar pressure center at the moment of landing, and the plantar pressure center at the moment of leaving the ground; The motion data of the subjects during walking is collected in real time through the posture sensor; the motion data includes: hip joint activity Euler angle, knee joint activity Euler angle, ankle joint activity Euler angle, hip joint acceleration, knee joint acceleration, ankle joint acceleration, hip joint angular velocity, knee joint angular velocity, ankle joint angular velocity.

3. A quantitative lower limb rehabilitation assessment system according to claim 2, characterized in that: The hip joint activity Euler angle is the angle between the thigh and the vertical axis in the sagittal plane, and it is defined that the angle is "positive" when flexing forward and the angle is "negative" when extending backward; the knee joint activity Euler angle is the angle between the calf and the extension line of the thigh, and the angle value is always "positive"; the ankle joint activity Euler angle is the angle between the fifth toe of the foot and the normal line perpendicular to the calf, and it is defined that the angle is "positive" when dorsiflexing upward and the angle is "negative" when plantar flexing downward.

4. A quantitative lower limb rehabilitation assessment system according to claim 1, characterized in that: The step of calculating the gait data of the subject during walking based on the walking data includes: Gait parameters and gait assessment parameters; Gait parameters include: Reaction time TA = foot lifting time T2 - walking command issuance time T1; Action time TB = time of leaving the ground T4 - time of lifting the foot T2; Gait cycle TF = ground contact time T7 - ​​ground contact time T3; Pace S = stride W / (landing time T6 - ground departure time T4); Step frequency F = 1 / gait period TF; Gait assessment parameters include: Starting time TC = ground-leave time T4 - walking command issuance time T1; Support time TD = landing time T6 - ground contact time T5; Swing time TE = landing time T5 - lift-off time T4; Step length L = the position at the time of landing T6 - the position at the time of lifting off T4; Stride W = plantar pressure center at landing time T6 - plantar pressure center at lift-off time T4; Bipedal difference DX = left foot parameter / right foot parameter, where both left foot parameter and right foot parameter include: support time TD, swing time TE, step length L, stride W; By inputting the gait parameters into the pre-trained LSTM model, the feature vector corresponding to the gait parameters is obtained.

5. A quantitative lower limb rehabilitation assessment system according to claim 1, characterized in that: The time domain and frequency domain analysis of the motion data includes: S1: extracting motion data collected in real time; S2: Obtain corresponding motion signal data by performing time domain analysis and frequency domain analysis on the motion data; wherein the motion signal data includes: peak-to-peak value, root mean square, skewness, kurtosis, approximate entropy, average amplitude change rate, main frequency component, signal power, frequency index and spectrum entropy; S3: By inputting the motion signal data into a pre-trained feature extraction model, a feature vector corresponding to the motion signal data is obtained.

6. A quantitative lower limb rehabilitation assessment system according to claim 1, characterized in that: The evaluation and analysis of gait data includes: By evaluating and analyzing the gait evaluation parameters, evaluation parameters are obtained; The periodicity of gait was determined by calculating the ratio of swing time TE and stance time TD; Determine the stability of the gait based on the starting time TC, stride W and step length L; The symmetry of the two feet is determined based on the difference DX of the two feet; among which, the difference contribution ratio of the support time TD, the swing time TE, the step length L, and the stride W is 1:1:1:1; By inputting the evaluation parameters into the pre-trained feature extraction model, the feature vector corresponding to the evaluation parameters is obtained.

7. A quantitative lower limb rehabilitation assessment system according to claim 1, characterized in that: The rehabilitation score is obtained by weighted summation based on the gait data, the motion signal data and the evaluation data, including: The feature vector corresponding to the gait parameter is marked as Xi, the feature vector corresponding to the motion signal data is marked as Yj, and the feature vector corresponding to the evaluation parameter is marked as Zg; wherein Xi represents the feature vector corresponding to the i-th gait parameter; Yj represents the feature vector corresponding to the j-th motion signal data; Zg represents the feature vector corresponding to the g-th evaluation parameter; The rehabilitation score is calculated by the formula: K = ∑(Ai×Xi)+∑(Bj×Yj)+∑(Cg×Zg); wherein, i = 1, 2, …, n, n is the total number of eigenvectors corresponding to gait parameters; j = 1, 2, …, m, m is the total number of eigenvectors corresponding to motion signal data; g = 1, 2, …, r, r is the total number of eigenvectors corresponding to estimation parameters; Ai, Bj, Cg are weight coefficients, and ∑Ai+∑Bj+∑Cg=1.

8. A quantitative lower limb rehabilitation assessment method, applied to a quantitative lower limb rehabilitation assessment system according to any one of claims 1 to 7, characterized in that: include: Step 1: Collect the walking data and motion data of the subject in real time during walking; Step 2: Calculate the gait data of the subject during walking based on the walking data; Step 3: Obtain motion signal data by performing time domain and frequency domain analysis on the motion data; Step 4: Obtain evaluation data by evaluating and analyzing the gait data; Step 5: Obtain a rehabilitation score through a weighted summation method based on the gait data, motion signal data, and evaluation data, and generate a lower limb rehabilitation evaluation report based on the rehabilitation score.

Images