CN112568901A - Gait symmetry and consistency evaluation method based on multiple sensors - Google Patents

Gait symmetry and consistency evaluation method based on multiple sensors Download PDF

Info

Publication number
CN112568901A
CN112568901A CN202011427022.8A CN202011427022A CN112568901A CN 112568901 A CN112568901 A CN 112568901A CN 202011427022 A CN202011427022 A CN 202011427022A CN 112568901 A CN112568901 A CN 112568901A
Authority
CN
China
Prior art keywords
gait
symmetry
data
consistency
model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202011427022.8A
Other languages
Chinese (zh)
Inventor
吕士云
郇战
耿宏杨
高歌
李华昊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Changzhou University
Original Assignee
Changzhou University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Changzhou University filed Critical Changzhou University
Priority to CN202011427022.8A priority Critical patent/CN112568901A/en
Publication of CN112568901A publication Critical patent/CN112568901A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
    • A61B5/112—Gait analysis
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
    • A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
    • A61B5/6813—Specially adapted to be attached to a specific body part
    • A61B5/6828—Leg
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Veterinary Medicine (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Physics & Mathematics (AREA)
  • Public Health (AREA)
  • Pathology (AREA)
  • Physiology (AREA)
  • Signal Processing (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Psychiatry (AREA)
  • Dentistry (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Power Engineering (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

The invention relates to a gait symmetry and consistency evaluation method based on multiple sensors, which comprises the following steps: s1, preprocessing and dividing data; s2, establishing a regression angle deviation model for the periodic data in the step S1; s3, establishing an evaluation strategy for the model in the step S2, and measuring gait symmetry and motion consistency; modeling the lower leg and thigh data of 8 subjects in 5-group weight-bearing mode, the evaluation results showed that 37.5% and 62.5% of subjects showed good symmetry in the lower leg and thigh region, respectively, and 87.5% of subjects showed symmetry in the thigh region that remained unchanged or better relative to the lower leg; furthermore, the motion consistency of the thigh is up to 90%, significantly higher than 77.5% of the calf. The invention provides a new idea for gait analysis of wearable equipment and has important guiding significance for understanding of load-bearing gait.

Description

Gait symmetry and consistency evaluation method based on multiple sensors
Technical Field
The invention belongs to the technical field of behavior recognition performance evaluation, and relates to a load-bearing gait analysis method for wearable equipment, in particular to a gait symmetry and consistency evaluation method based on multiple sensors.
Background
Clinical studies have found that symmetry and consistency studies in patients with neurological or motor disorders, including lower limb amputation, anterior cruciate ligament reconstruction, hip or knee replacement, chronic idiopathic neck pain or stroke, etc., reveal physical health and pathological features. The kinematics research finds that the evaluation of the gait symmetry and consistency is helpful for detecting the physical consumption condition, the action accuracy degree and the like of the athlete, such as discussing the asymmetry of the athlete in the run-up process to adjust the run-up speed, analyzing the gait variability and symmetry of a race walker to distinguish a high-grade athlete from an elite-grade young race walker, comparing the gait symmetry of a long-distance runner at different running distances to research the action variability and the like. Some researchers have also studied activities of daily activities, such as discussing how walking speed changes step size, how arm swing amplitude affects lower limb asymmetry, what is different in barefoot and shoe gait symmetry, etc. Therefore, the gait symmetry and consistency have important significance for clinical medicine, kinematics, biomechanics and the like.
Gait in daily behavioral activities can reflect real life more realistically than medical research and kinematics research. Weight-bearing walking, particularly hand-held heavy walking, is a common daily activity widely existing in groups such as students, office workers, shoppers, athletes, and medical patients. The gait is easy to be abnormal due to long-term load bearing, and even the waist and back muscle ache or the lower limb joint injury is caused. Therefore, the influence of the load-bearing behaviors on the gait of the lower limbs needs to be researched, and the gait difference caused by different load-bearing behaviors is measured through symmetry and consistency.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: in order to overcome the defects in the prior art, the invention provides a gait symmetry and consistency evaluation method which is based on a regression angle deviation model (RAOM) of a multi-sensor, establishes a symmetry grading rule through deviation angles under different loading modes, and measures the movement consistency by using position span under multiple experiments.
The technical scheme adopted by the invention for solving the technical problems is as follows: a gait symmetry and consistency assessment method based on multiple sensors comprises the following steps: preprocessing and segmenting data, establishing a regression angle deviation model and establishing an evaluation strategy.
The data preprocessing and segmenting step comprises:
1) and data acquisition: acquiring multi-sensor data of crus and thighs, and extracting acceleration and angular velocity data of three experiments of eight subjects under five loading modes;
2) and calculating the signal amplitude: the method performs amplitude scalar calculation on inertial sensor data, and the signal peak values of acceleration and angular velocity are calculated as follows:
Figure BDA0002825354640000021
Figure BDA0002825354640000022
wherein A isx、AyAnd AzAcceleration data of x, y, z axes, Gx、GyAnd GzAngular velocity data for x, y, and z axes, respectively;
3) fitting and resampling: the method adopts three-spline linear interpolation to perform function fitting to prevent data loss in the acquisition process, and the gait data of the left and right lower limbs are synchronized by resampling;
4) and low-pass filtering: the load walking process is a stable process, and signals are mainly distributed in a low-frequency part, so that the method adopts a zero-phase forward and reverse third-order Butterworth low-pass filter, and the cut-off frequency is 5 Hz;
5) and segmentation: firstly, gait event detection is carried out, and each gait cycle is divided according to the landing point of the heel. The gait event step mainly comprises three steps of peak value detection, abnormal value elimination and missing point supplement;
the step of establishing a regression angle offset model comprises:
1) and establishing a model: suppose that
Figure BDA0002825354640000031
And
Figure BDA0002825354640000032
respectively represent the left and right gait data of the subject S in the Uth group gait mode, and
Figure BDA0002825354640000033
and
Figure BDA0002825354640000034
the original data is subjected to mean value removal to obtain a new vector, which is defined as:
Figure BDA0002825354640000035
Figure BDA0002825354640000036
wherein i represents the i-th sample point of the subject S in the weight-bearing state U, i is 1,2, …,101 × N, S is 1,2, …,5, U is 1,2, …,5, and N is the number of cycles of the subject S in the weight-bearing mode U.
And (3) constructing a linear regression model by taking the left gait data as an independent variable and the right gait data as a dependent variable:
Figure BDA0002825354640000037
wherein, beta0S,UAnd beta1S,UIntercept and regression coefficient, ε, of subject S in the Uth set of gait patterns, respectivelyS,UFor random error term, with a single gait of a single subjectBy way of example, equation (5) is simplified as:
Yi＝β0+β1Li+e (6)
wherein, YiAs right side gait data, LiLeft side gait data, e residual term, regression coefficient beta1And intercept beta0Respectively representing offset and amplitude difference;
2) and (3) testing the model: when checking the interpreted variable YiAnd interpreting variable XiIf there is linear relation, give the original hypothesis H0:β 10 and alternate hypothesis H1:β1Not equal to 0; if it is assumed that H is0If the sum of squared deviations (SST) is established, the linear regression model is shown to lose significance, and F statistic is constructed through a decomposition method of the sum of squared deviations (SST):
Figure BDA0002825354640000038
wherein p is the number of variable parameters, n is the number of samples, and n-p-1 is the degree of freedom. The larger the F statistic is, the larger X isiFor YiThe greater the impact of (a), the more likely the original hypothesis is to be rejected; the closer the F statistic is to 0, the more X is declarediFor YiThe smaller the influence of (a), the more likely the original assumption is to hold;
at the same time, using the goodness of fit R2And R2-degree of fit of adjusted validation model:
Figure BDA0002825354640000041
Figure BDA0002825354640000042
where SSR is a regression sum of squares, SST is a dispersion sum of squares, SSE is a residual sum of squares, and n is the number of samples. R2The fitting degree of the regression equation to the observed value is reflected, and the closer the value is to 1, the SSR accounts for the SSTThe higher the ratio, the better the fit of the model to the sample points;
3) and residual analysis: the residual being observed values
Figure BDA0002825354640000043
(denoted Y) and the regression estimate Y, i.e., e-Y (I-H) Y, where H is the residual matrix, also known as the hat matrix, and H-X (X)TX)-1XT. The method comprises the following steps of utilizing information provided by residual errors to investigate the reasonability of model hypothesis and the reliability of data, namely residual error analysis, and comprising the steps of residual error normality test, independence test and variance homogeneity test;
4) and data diagnosis: generally, for the establishment of the model, the most ideal state is that each sample has a certain influence on the statistical inference, but the influence is not expected to be too large, otherwise, the analysis of the model is influenced, so that the model lacks stability, therefore, outliers, strong influence points and high lever points in the data are eliminated, and the model effect after the outliers are eliminated is observed:
student residual error (r)i) Is a commonly used index for detecting outliers and is the ratio of the residual to the estimated standard deviation. From the nature of the residual, Var (e) can be derivedi)＝σ2(1-hii) Wherein h isiiIs the diagonal element of the hat matrix H, then riIs defined as:
Figure BDA0002825354640000044
hat statistics (h)i) The existence of high leverage points can be judged, the high leverage points refer to outliers of an independent variable factor space and are formed by combining abnormal independent variable values, and when the independent variables of some data points are far away from other points, the points are called the high leverage points. h isiIs defined as follows:
Figure BDA0002825354640000051
wherein n is the number of samples, XiFor the value of each of the independent variables,
Figure BDA0002825354640000052
is the mean value. Generally, a hat value greater than 3 times the mean hat value indicates a high leverage point.
Coker distance (D)i) The degree of influence of a single sample on the whole regression model is described, and the larger the library k distance is, the larger the influence of the sample on the model is. Thus, the method uses the Cocky distance to find the strong point of influence, DiIs defined as follows:
Figure BDA0002825354640000053
wherein r isiIs a student residual, p is the number of variable parameters, s2Is the Mean Square Error (MSE). In general, for n sample sizes and p variable parameters, a value of the Cocky statistic greater than 4/(n-p-1) indicates a strong point of influence;
the step of establishing an evaluation policy comprises:
1) calculating an offset angle: known regression coefficient beta1The included angle between any two linear regression equations can be obtained
Figure BDA0002825354640000054
Is a vector with a direction, and is used for measuring
Figure BDA0002825354640000055
To
Figure BDA0002825354640000056
The offset between the two or more of the first and second,
Figure BDA0002825354640000057
is an action GiThe regression curve of (a) is obtained,
Figure BDA0002825354640000058
is an action GjThe regression curve of (2). Offset distance
Figure BDA0002825354640000059
The direction of the offset is marked as "+",
Figure BDA00028253546400000510
the offset direction is noted as "-". In this method, a direction of "+" represents a left bias and a direction of "-" represents a right bias. Offset distance D between different loading states of subjectijIs defined as:
Figure BDA00028253546400000511
wherein the content of the first and second substances,
Figure BDA00028253546400000512
and
Figure BDA00028253546400000513
for subject Sith group action GiAnd jth action GjI, j belongs to {1,2,3,4,5}, and i ≠ j;
2) establishing a symmetry grade division rule: the symmetry grade of the subject is measured by the following rules, namely whether the deviation directions of the loads on different sides are consistent, whether the deviation distance of the loads on the same side is in direct proportion to the load degree, and whether the deviation direction is consistent with the load direction. According to the sequence of the given rules, 4 symmetry levels are divided, specifically:
step 1: if it is not
Figure BDA0002825354640000061
And is
Figure BDA0002825354640000062
The symmetry grade is marked as C, otherwise, D;
step 2: on the premise that the symmetry level satisfies C, if
Figure BDA0002825354640000063
And is
Figure BDA0002825354640000064
The symmetry level is designated as B;
step 3: on the premise that the symmetry level satisfies B, if
Figure BDA0002825354640000065
And is
Figure BDA0002825354640000066
The symmetry rating is designated as a;
3) calculating the position span: the position span refers to the range of position change of the subject in the action, and if the positions of Sub1 in three experiments when walking without load are 2, 4, and 4, respectively, the position span is 2. The span is 0, then the action consistency is considered to be excellent; if the span is 1 or 2, the action consistency is considered to be higher; the span is 3 and 4, the action is considered to be less consistent. The movement consistency of the subjects was evaluated based on the position span of the five weight bearing states in three experiments.
The invention has the beneficial effects that: the method for evaluating the lower limb movement consistency and the gait symmetry provides a new technical scheme for the health monitoring and the rehabilitation training of clinical patients, and fully utilizes the information of each sampling point by establishing a model with good fitting capacity on left and right gait data; in addition, through the concentration degree of the data, the abnormal state of the testee in the gait phase can be clearly observed, so that a clinician can quantify the lower limb movement with more confidence and measure the gait ability of the patient.
Drawings
The invention is further illustrated with reference to the following figures and examples.
Fig. 1 is a block diagram of the present invention.
FIG. 2 is a schematic diagram of the visualization of the data preprocessing and segmentation according to the present invention.
FIG. 3 is a graph showing the results of the residual analysis of the present invention.
FIG. 4 is a comparison chart before and after the abnormal value is eliminated.
FIG. 5 is a schematic view of the position span of the subject under different weight bearing conditions in the present invention.
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic views illustrating only the basic structure of the present invention in a schematic manner, and thus show only the constitution related to the present invention.
A method for evaluating gait symmetry and consistency based on multiple sensors as shown in fig. 1, comprising the following steps:
s1, preprocessing gait data: in the process of weight-bearing walking, four inertial sensors are adopted to acquire data of 3-axis acceleration and 3-axis angular velocity of the middle parts of the lower leg and the thigh of a subject, and the sampling frequency is 250 Hz.
In this example, 8 healthy young people were selected as subjects (average age: 23. + -.2 years, height: 171. + -.15 CM; weight: 70. + -.16 KG), and walking data of 5 groups of healthy young people under load were obtained, and the walking data were recorded as the weights of no load, 2.5KG for right-hand load, 5KG for right-hand load, 2.5KG for left-hand load, and 5KG for left-hand load, and recorded as G1, G2, G3, G4, and G5, and the experiment was repeated 3 times for each group. A notebook computer bag is used as a carrier for loading heavy objects, and a laboratory walkway with the length of 12 meters is used as an experimental environment. After a series of data preprocessing and segmentation steps, the visualization result shown in fig. 2 is obtained.
S2, establishing a regression angle deviation model (RAOM): table 1 shows the results of the model tests, for the experimental data of this study, the F-values correspond to P<0.05, reject the original hypothesis H0The explanatory variables are considered to have a significant effect on the explained variables. Meanwhile, the goodness-of-fit R-square and the adjusted R-square are both 89.7%, which indicates that the fitting degree of the linear regression to the model is high.
Table 1 results of model testing
Figure BDA0002825354640000071
And SS: the sum of squares; df: degree of freedom; MS: mean square error
Fig. 3 shows the results of residual analysis of the left and right side gait data. Fig. 3(a) is a normal probability graph of residuals, in which the horizontal axis represents student residuals sorted from small to large, and the vertical axis represents normal scores. The normal score is from the expected value of the sample order statistic conforming to the standard normal distribution, if the number of samples is n, the normal score of the i-th residual after sorting is the i/n quantile of the standard normal distribution. If the residual errors obey standard normal distribution, the sorted residual errors are approximately the same as a normal score, under the assumption of normality, all points in the scatter diagram approximate a straight line, and the student residual errors are considered to accord with the normality; FIG. 3(b) shows student residuals according to the observation sequence, under the assumption of error independence, all residuals in the graph can be observed to randomly fluctuate around 0, and the residuals are considered to be independent from each other; in fig. 3(c), the horizontal axis represents the fitting value and the vertical axis represents the residual error. Under standard assumptions, the student residuals are not correlated with the fit values, and the points in the graph should also be randomly distributed. It is clear that the points in the graph are mostly distributed between ± 2 times the standard deviation and there is no significant systematic trend, so the residuals are considered to satisfy homogeneity of variance.
Generally, for the establishment of the model, the optimal condition is that each sample has a certain influence on the statistical inference, but the influence is not expected to be too large, otherwise, the analysis of the model is influenced, and the model lacks stability. The sample may be disturbed due to a difference in load or an influence of noise. And eliminating abnormal values in the model by adopting student residual errors, lever point detection and Cock distance analysis. And comparing the regression models in different loading modes before and after the interference points are eliminated, and finding that the model effect is optimized after the interference points are eliminated. The effect before and after the interference point is eliminated is shown by taking one subject as an example. As shown in FIG. 4, the line of the direct proportional function is the symmetry line in the ideal state, G1 is the regression line in the no-load state, G2 and G3 are the regression lines for the weights 2.5KG and 5KG held by the right hand, respectively, and G4 and G5 are the regression lines for the weights 2.5KG and 5KG held by the left hand, respectively. It is observed that the deviation rules of the subject under different load conditions are difficult to be seen from the five lines shown in fig. 4(a), after the interference points are eliminated, the symmetrical deviation distance of the gait is observed to be larger along with the increase of the degree of holding the weight on the load side, and the deviation directions of different side loads have obvious distinction degree relative to the non-load condition.
S3, establishing an evaluation strategy: the method measures the movement consistency of the subject among 3 experiments through position span. The position span refers to the position change range of the subject under the action, as shown in table 2, Sub1 to Sub8 are subject numbers, G1 to G5 are action numbers, T1 to T3 are trial repetition numbers, and the numbers in the table represent the positions of the action (sequentially marked as 1,2,3,4 and 5 from left to right). In Table 2, the span of Sub1 when walking without load is 2, and the span when walking with a 2.5KG weight on the right hand is 1. The span is 0, then the action consistency is considered to be excellent; if the span is 1 or 2, the action consistency is considered to be higher; the span is 3 and 4, the action is considered to be less consistent. Fig. 5 shows a position span diagram of different movements of 8 subjects, with the horizontal axis representing the subject numbers and the vertical axis representing the span, and colors representing different weight bearing modes. The figure clearly shows the level of motion consistency for the subjects, as the subject with Sub5 has significantly worse motion consistency than the subject with Sub6, and Sub6 has excellent motion consistency at 5KG for right and 5KG for left hand. Analysis of the position span plot showed that the consistency of motion of the subject's lower leg was 77.5%. Meanwhile, the thigh data are analyzed for motion consistency, and the motion consistency of the thigh is up to 90%.
TABLE 2 span of positions
Figure BDA0002825354640000091
Table 3 shows the evaluation results of the symmetry levels of the calf and thigh of the subject shown in table 4, based on the deviation distance statistic under different load conditions and the rule of dividing the statistic and the symmetry levels. The levels "a" and "B" are considered close to the ideal case with good symmetry, and the results "C" and "D" are considered out of the ideal case with poor symmetry. The results in table 4 show that 37.5% of the subjects had better symmetry in the symmetry class of the lower leg and 62.5% of the subjects had better symmetry in the symmetry class of the upper leg; when walking with a right or left hand holding a weight, the symmetry of the thigh is increased by 25% and 12.5% respectively over the symmetry of the calf. The overall evaluation of symmetry showed that 87.5% of the subjects had a stable or better degree of thigh symmetry relative to the lower leg and an improved degree of thigh symmetry relative to the lower leg of 12.5%.
TABLE 3 statistics
Figure BDA0002825354640000101
Figure BDA0002825354640000102
TABLE 4 evaluation results of symmetry level
Figure BDA0002825354640000103
The above analysis shows that there is a symmetrical shift in gait in the natural walking state, with 87.5% of subjects having their calves in a left-leaning gait and 75% of subjects having their thighs in a right-leaning gait. The human gait is influenced by sensory input factors such as the interaction between joints, the distribution of muscles, and vision, so that the following rules are that the deviation directions of loads on the same side are consistent, the deviation distance of loads on the same side is in direct proportion to the load degree, and the deviation direction is consistent with the load direction, and only exist in a few healthy subjects, and the symmetry of the subjects is considered to be consistent with the ideal condition and high. Under different loading modes, the symmetry and consistency of the thigh are better than those of the shank part, and the right-side loading has better regularity than no loading. These conclusions are closely related to the swing amplitude of the legs and the degree of adaptation to the weight, and are in line with the real situation.
In light of the foregoing description of the preferred embodiment of the present invention, many modifications and variations will be apparent to those skilled in the art without departing from the spirit and scope of the invention. The technical scope of the present invention is not limited to the content of the specification, and must be determined according to the scope of the claims.

Claims (4)

1. A gait symmetry and consistency assessment method based on multiple sensors is characterized by comprising the following steps: comprises the following steps:
s1, preprocessing and dividing data;
s2, establishing a regression angle deviation model for the periodic data of the step S1;
and S3, establishing an evaluation strategy for the model formed in the step S2, and measuring gait symmetry and motion consistency.
2. The method of multi-sensor based gait symmetry and consistency assessment according to claim 1, characterized by: the step S1 specifically includes:
s1.1, data acquisition: acquiring multi-sensor data of crus and thighs, and extracting acceleration and angular velocity data of three experiments of eight subjects under five loading modes;
s1.2, pretreatment: the method comprises four steps of signal amplitude calculation, interpolation fitting, resampling and low-pass filtering;
s1.3, segmentation: and detecting gait events of the time sequence after the low-pass filtering, and dividing gait cycles according to the detected heel landing points, wherein the gait event detection comprises peak value detection, abnormal value elimination and missing point supplement.
3. The method of multi-sensor based gait symmetry and consistency assessment according to claim 1, characterized by: the step S2 specifically includes:
s2.1, establishing a model: constructing a linear regression model by taking the left gait data as an independent variable and the right gait data as a dependent variable;
s2.2, model checking: testing of the interpreted variable Y Using F statisticsiAnd interpreting variable XiWhether or not there is a linear relationship, and at the same time, using the goodness of fit R2And R2-degree of fit of the adjusted validation model;
s2.3, residual analysis: the method comprises the following steps of utilizing information provided by residual errors to investigate the reasonability of model hypothesis and the reliability of data, namely residual error analysis, and comprising the steps of residual error normality test, independence test and variance homogeneity test;
s2.4, data diagnosis: and eliminating outliers, strong influence points and high leverage points in the data.
4. The method of multi-sensor based gait symmetry and consistency assessment according to claim 1, characterized by: the step S3 specifically includes:
s3.1, calculating an offset angle: the deviation angle is defined as an included angle between any two linear regression equations, in the method, the direction is "+" which represents leftward deviation, and the direction is "-" which represents rightward deviation;
s3.2, establishing a symmetry grade division rule: the symmetry grade of the testee is measured by the following rules, namely whether the deviation directions of loads on different sides are consistent, whether the deviation distance of loads on the same side is in direct proportion to the load degree, whether the deviation direction is consistent with the load direction, and 4 symmetry grades A, B, C, D are divided according to the sequence of the given rules;
s3.3, calculating position span: the position span refers to the position variation range of the subject under the action, and if the span is 0, the action consistency is considered to be excellent; if the span is 1 or 2, the action consistency is considered to be higher; the movement consistency is considered to be poor when the span is 3 and 4, and the movement consistency of the subject is evaluated according to the position span of five loading states in three experiments.
CN202011427022.8A 2020-12-09 2020-12-09 Gait symmetry and consistency evaluation method based on multiple sensors Pending CN112568901A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011427022.8A CN112568901A (en) 2020-12-09 2020-12-09 Gait symmetry and consistency evaluation method based on multiple sensors

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011427022.8A CN112568901A (en) 2020-12-09 2020-12-09 Gait symmetry and consistency evaluation method based on multiple sensors

Publications (1)

Publication Number Publication Date
CN112568901A true CN112568901A (en) 2021-03-30

Family

ID=75127965

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011427022.8A Pending CN112568901A (en) 2020-12-09 2020-12-09 Gait symmetry and consistency evaluation method based on multiple sensors

Country Status (1)

Country Link
CN (1) CN112568901A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113674796A (en) * 2021-08-17 2021-11-19 安渡生物医药(杭州)有限公司 Method for establishing drug-resistant antibody calculation threshold group and system for realizing method
CN114287921A (en) * 2021-12-23 2022-04-08 常州信息职业技术学院 Gait bilateral similarity analysis method, device and system
CN114376566A (en) * 2022-02-16 2022-04-22 常州大学 Symmetry evaluation method for lower limb segments during hand load
CN115372294A (en) * 2022-09-15 2022-11-22 中国市政工程东北设计研究总院有限公司 Graphite tube stability discrimination method

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106937872A (en) * 2017-04-20 2017-07-11 杭州电子科技大学 A kind of gait bilateral symmetric property evaluation method based on regression curve
CN110123317A (en) * 2019-04-28 2019-08-16 华东交通大学 Merge the knee joint contact force calculation method of artificial fish-swarm and random forests algorithm
CN111568436A (en) * 2020-05-29 2020-08-25 常州大学 Gait symmetry evaluation method based on regression rotation angle

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106937872A (en) * 2017-04-20 2017-07-11 杭州电子科技大学 A kind of gait bilateral symmetric property evaluation method based on regression curve
CN110123317A (en) * 2019-04-28 2019-08-16 华东交通大学 Merge the knee joint contact force calculation method of artificial fish-swarm and random forests algorithm
CN111568436A (en) * 2020-05-29 2020-08-25 常州大学 Gait symmetry evaluation method based on regression rotation angle

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
ZHAN HUAN等: "An Evaluation Strategy for the Symmetry and Consistency of Lower Limb Segments During Upper Limb Loading", 《IEEE SENSORS JOURNAL》 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113674796A (en) * 2021-08-17 2021-11-19 安渡生物医药(杭州)有限公司 Method for establishing drug-resistant antibody calculation threshold group and system for realizing method
CN113674796B (en) * 2021-08-17 2024-02-20 安渡生物医药(杭州)有限公司 Method for establishing drug-resistant antibody calculation threshold group and system for realizing method
CN114287921A (en) * 2021-12-23 2022-04-08 常州信息职业技术学院 Gait bilateral similarity analysis method, device and system
CN114376566A (en) * 2022-02-16 2022-04-22 常州大学 Symmetry evaluation method for lower limb segments during hand load
CN115372294A (en) * 2022-09-15 2022-11-22 中国市政工程东北设计研究总院有限公司 Graphite tube stability discrimination method

Similar Documents

Publication Publication Date Title
CN112568901A (en) Gait symmetry and consistency evaluation method based on multiple sensors
Hurmuzlu et al. On the measurement of dynamic stability of human locomotion
CN106667493A (en) Human body balance assessment system and assessment method
CN104997523B (en) A kind of upper limb rehabilitation robot rehabilitation training motor function evaluation method
Nigg et al. Development of a symmetry index using discrete variables
Shourijeh et al. A forward-muscular inverse-skeletal dynamics framework for human musculoskeletal simulations
Jiang et al. The elderly fall risk assessment and prediction based on gait analysis
CN111568436B (en) Gait symmetry evaluation method based on regression rotation angle
Tomescu et al. The effects of filter cutoff frequency on musculoskeletal simulations of high-impact movements
Rekant et al. Inertial measurement unit-based motion capture to replace camera-based systems for assessing gait in healthy young adults: Proceed with caution
Kloeckner et al. Prediction of ground reaction forces and moments during walking in children with cerebral palsy
Ulman et al. Using gait variability to predict inter-individual differences in learning rate of a novel obstacle course
CN111540465A (en) Method for predicting movement injury risk of college student male football players by neural network model
Kato et al. Estimating a joint angle by means of muscle bulge movement along longitudinal direction of the forearm
Mathur et al. Gait classification of stroke survivors-An analytical study
CN110755084B (en) Motion function assessment method and device based on active-passive and staged actions
Tong et al. The influence of treadmill on postural control
Bensalma et al. Graphical-based multivariate analysis for knee joint clinical and kinematic data correlation assessment
CN115399791B (en) Method and system for evaluating functions of lower limbs of stroke based on myoelectric motion multi-data fusion
Akhavanfar et al. Evaluation of spinal force normalization techniques
Jia et al. Principal component analysis to guide the reduce the risk of fall
Eskandari et al. Comparative evaluation of different spinal stability metrics
Sari et al. Predicting effect of physical factors on tibial motion using artificial neural networks
Habibi et al. Health Level Classification of Motor Stroke Patients Based on Flex Sensor Using Fuzzy Logic Method
Çetiner et al. Tibial rotation assessment using artificial neural networks

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20210330

WD01 Invention patent application deemed withdrawn after publication