A device and method integrating neck motion trajectory collection and motion symmetry analysis

Abstract

The application relates to a device and a method integrating neck movement trajectory collection and movement symmetry analysis, and relates to the fields of three-dimensional measurement and neck movement evaluation. In the application, data collection is carried out by using a binocular camera, so that the instantaneous speed and the instantaneous acceleration at each collection moment can be conveniently measured, the speed and acceleration curves in the movement process are obtained, the force degree of each neck position of a to-be-evaluated person can be evaluated according to the curves, and the force degree of the neck position of the to-be-evaluated person at the moment is obtained. The application proposes a new measurement method for the symmetry of the movement trajectory, and combines the existing curve similarity and kinematics characteristics, so that the symmetry of the neck movement trajectory and the movement speed is more comprehensively and meticulously described, and the existing movement symmetry description method is enriched. The application can effectively extract the kinematics indexes of the neck, and evaluate the symmetry and stability of the neck movement in each direction.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of three-dimensional measurement and neck movement evaluation, and particularly relates to a device and method integrating neck movement trajectory collection and cervical vertebra movement symmetry analysis. BACKGROUND

[0002] Symptoms of neck diseases are often reflected in changes in neck movement characteristics. There are many methods for measuring neck movement, such as cervical vertebra activity measuring instruments, photographic analysis devices, etc. These devices can measure movement limit angles in various directions, or use X-ray, ultrasound, etc. to accurately measure movement limit angles and extract movement trajectories, thereby describing cervical vertebra activity to some extent, and achieving the purpose of auxiliary diagnosis and disease screening according to movement limit angles and cervical vertebra movement trajectories.

[0003] However, most of the existing devices only analyze the start and end states, and although the movement limit angle can be obtained, the entire movement process is ignored, or after the movement trajectory is extracted, no detailed trajectory curve analysis is performed, so that the movement speed, movement symmetry, etc. at each moment in the movement cannot be clearly reflected, and therefore the understanding of cervical vertebra activity under the existing technology is not comprehensive and intuitive. SUMMARY

[0004] In view of the existing problems, the present application provides a device integrating neck movement trajectory collection and movement symmetry analysis, which can effectively extract neck kinematic indexes and evaluate the symmetry and stability of neck movement in various directions.

[0005] A device integrating neck movement trajectory collection and movement symmetry analysis, characterized in that the collection and analysis device comprises a marker ball fixing cap 8 for marking the position of a head feature; a light supplementing lamp 13 for improving data quality; a darkroom 9 for improving data quality; a binocular camera 10 for collecting data; a processing computer 11 for storing and processing data; and a data transmission line 12 for connecting the binocular camera and the processing computer.

[0006] A device integrating neck movement trajectory collection and movement symmetry analysis, characterized in that the binocular camera is connected to the data transmission line 12 through an RJ-45 crystal head, and the other end of the data transmission line 12 is connected to any port of the processing computer 11 through an RJ-45 crystal head. The data transmission line 12 transmits data through a 10Base-T twisted pair line.

[0007] The device integrates neck movement trajectory collection and movement symmetry analysis, characterized by collecting and analyzing three kinds of neck movements, namely left and right lateral flexion, forward flexion and backward extension, and left and right rotation. The initial state of the three kinds of neck movements is sitting in a neutral position. The left and right lateral flexion movement is flexing the neck to the left limit position in the coronal plane, returning to the neutral position, and then flexing to the right limit position and returning to the neutral position. The forward flexion and backward extension movement is flexing forward to the limit position in the sagittal plane, returning to the neutral position, and then extending backward to the limit position and returning to the neutral position. The left and right rotation movement is rotating the neck to the left limit position in the transverse plane, returning to the neutral position, and then rotating to the right limit position and returning to the neutral position.

[0008] The device integrates neck movement trajectory collection and movement symmetry analysis, characterized by marking six feature positions during the movement. The marking ball fixed cap is composed of a strap 6, an inverted "T" fixed frame 5, and three retroreflective marker balls 4 fixed to the three ends of the inverted "T" fixed frame 5, which mark the left, right, and upper three feature positions of the head, respectively. In addition, the spinous processes of the second, fifth, and seventh cervical vertebrae are marked with retroreflective marker points to mark the three feature positions.

[0009] The device integrates neck movement trajectory collection and movement symmetry analysis, characterized by a binocular camera 10 made of two GIGE industrial cameras, with a collection pixel of 1280*1024 and a collection frame rate of 33 frames / second.

[0010] The device integrates neck movement trajectory collection and movement symmetry analysis, characterized by a marker point batch recognition and marking software designed for calculating the three-dimensional coordinates of the marker points in the dynamic data stream. The software is installed on a processing computer 11 and includes a marker point recognition and labeling module 14 for recognizing the two-dimensional coordinates of the marker points, an anomaly detection module 15 for detecting the marker point recognition error frames, a display interaction module 16 for displaying the recognition results and user interaction, and a three-dimensional measurement module 17 for calculating the three-dimensional coordinates of the marker points. The relationship between the modules is as follows: the marker point recognition and labeling module 14 recognizes the two-dimensional coordinates of the marker points, the anomaly detection module 15 detects anomalies according to the two-dimensional coordinate data and frame collection sequence, and if an anomaly is detected, the abnormal frame is delivered to the display interaction module 16 for user marker point recognition and labeling operation; if there is no anomaly or the user operation is completed, the two-dimensional coordinate data is delivered to the three-dimensional measurement module 17; the three-dimensional measurement module 17 calculates the three-dimensional coordinates of the marker points according to the triangulation principle.

[0011] The device integrates neck movement trajectory collection and movement symmetry analysis, characterized by trajectory extraction and analysis software designed for three kinds of neck movement trajectory extraction and analysis. The software is installed on a processing computer 11 and includes an analysis module 18 for extracting and analyzing left and right lateral flexion, forward flexion and backward extension, and rotation movement trajectories, the analysis module including three sub-modules corresponding to the three movements respectively; a display interaction module 19 for user selection of operation directory and display of movement analysis results. The user selects a target file directory and neck movement to be analyzed through the display interaction module 19, which is handed over to the corresponding sub-module of the analysis module 18 for trajectory extraction and analysis, and the results are displayed to the display interaction module 19.

[0012] The method integrates neck movement trajectory collection and movement symmetry analysis, characterized by using the following steps:

[0013] S1, the subject wears a marker ball fixing cap, and reflective marker points are pasted on the spinous processes of the 7th cervical vertebra, the 5th cervical vertebra and the 2nd cervical vertebra, respectively, in a dark room with a fill light turned on, left and right lateral flexion, forward flexion and backward extension, and left and right rotation movements are performed, and a binocular stereovision device is used for recording.

[0014] S2, using the marker point batch recognition and labeling software to recognize and label all the pictures collected by the binocular camera, and then calculating the three-dimensional coordinates of the marker points;

[0015] S3, arranging the coordinates of each marker point frame by frame to obtain the three-dimensional movement trajectory of the marker position;

[0016] S4, using the trajectory extraction and analysis software to analyze the space movement trajectory and extract the index.

[0017] The index extracted in the above step S4 includes: for left and right lateral flexion and left and right rotation movement, the left and right movement limit angle and the ratio, the swing amplitude during left and right movement, the left and right movement average distance, the left and right movement average acceleration ratio, the neck movement imbalance coefficient are calculated respectively, and the left and right movement velocity and acceleration curves are drawn; for forward flexion and backward extension movement, the forward flexion and backward extension limit angle, the swing amplitude during forward flexion and backward extension, the forward flexion and backward extension average speed ratio are calculated respectively, and the forward flexion and backward extension trajectory curve, the forward flexion and backward extension velocity and acceleration curve are drawn.

[0018] The movement symmetry measurement method based on circle fitting and space distance extracted by the above step S4 is used, taking left and right lateral flexion as an example, including the following sub-steps:

[0019] S41, fitting the left and right lateral flexion trajectories into circular arcs, then the radius corresponding to the neutral position is R M , the central angles θ l , θ r corresponding to the left and right lateral flexion trajectories are calculated respectively, then the left and right lateral flexion angle ratio θratio The calculation method is as follows:

[0020]

[0021] S42. Move the left flexion trajectory to R M Let θ be the axis of reflection. r <θ l , for θ r Rounding [θ] r ], Draw rays outward from the center of the arc at 1-degree intervals (from R M (starting from the ray), total [θ] r Each ray intersects two trajectory curves at two points. Calculate the spatial distances d1, d2, ..., d between the two points. [θr] Calculate the average distance D across all distances as a measure of the symmetry of the trajectory curve;

[0022] S43. Ratio of left and right lateral flexion average accelerations: If the lengths of the left and right lateral flexion trajectories are S... l S r The total number of frames for left and right lateral flexion is I. l I r If the camera frame rate K = 33, then the left-side bending average acceleration... Right-sided mean velocity Then the ratio of the mean lateral flexion accelerations to A ratio The calculation method is as follows:

[0023]

[0024] A ratio As a measure of the symmetry of the magnitude of the force exerted by the subject during left and right movements;

[0025] S44, Spatial location P of the marker point in the nth frame n (x n ,y n ,z n ), spatial location P of the (n+1)th frame n+1 (x n+1 ,y n+1 ,z n+1 If the instantaneous velocity V at this moment is given, then... n for:

[0026]

[0027] This allows us to plot the left and right lateral flexion velocity curves. By differentiating the velocity curves, we can plot the left and right lateral flexion acceleration curves, which can be used to visually display the motion state.

[0028] S45, calculate the left and right lateral flexion space trajectory similarity measure index Sim, which evaluates the symmetry of the two trajectory curves as a whole, the calculation method of Sim is S451-S452. Integrate the motion symmetry description array [theta ratio , D, A ratio , Sim], which describes the motion symmetry in the left and right lateral flexion process from multiple directions;

[0029] S46, after the calculation of the motion symmetry index, obtain the relationship model between the neck motion imbalance coefficient and each influencing factor in the symmetry description array through the SVR neural network. In the experiment, 71 measurement data (50 normal neck motion data and 21 limited neck motion data) are used as the training set, and through cross-validation, the regression model with Gaussian kernel function has the best prediction result, and the average measurement error mae, mean square error mse and determination coefficient r 2 are 0.114, 0.018 and 0.911 respectively. After determining the model, any symmetry description array can be input to calculate the neck motion imbalance coefficient (close to 0 for good neck motion symmetry, close to 1 for serious neck motion asymmetry), for example:

[0030] input = [0.03549, 12.25, 0.088044, 66.76] -> output = [-0.02085159]

[0031] The data is normal person data, and the output is -0.02.

[0032] The calculation method of the left and right lateral flexion space trajectory similarity measure index Sim is as follows:

[0033] S451, first, the shortest path planning is performed on the two space curves, the left lateral flexion space point sequence L i (i = 1, 2, ……n), the right lateral flexion space point sequence R j (j = 1, 2, ……m), the distance corresponding to the points at different positions of the two sequences is calculated to form a distance matrix, and the element (L i , R j ) in the matrix corresponds to the Euclidean distance of the space point L i and the space point R j ;

[0034] S452, define the path W, W is a path from (L0, R0) to (L n , R m ) in the distance matrix. Traverse all the paths existing in the distance matrix to find the path with the smallest sum of path points, and the average of all path points on this path is the index Sim.

[0035] The present application has the following advantages:

[0036] (1) The data collection of the present application uses a binocular stereo vision measurement camera, which can obtain the three-dimensional spatial coordinates of any spatial point in the field of view. With the three-dimensional spatial coordinates and the collection rate, the instantaneous velocity and the instantaneous acceleration at each collection time can be easily calculated, and the velocity and acceleration curves during the movement process can be drawn. Furthermore, the force degree of each neck position of the to-be-tested person can be evaluated from the curves, and the force degree of the neck position of the to-be-tested person at the moment can be obtained.

[0037] (2) The present application proposes a new method for measuring the symmetry of the movement trajectory. The symmetry degree of the neck movement trajectory and the movement velocity is more comprehensive and detailed by combining the existing curve similarity and kinematic characteristics, which enriches the existing movement symmetry description method. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 , three kinds of neck movements (left and right lateral flexion, forward and backward extension, left and right rotation) are shown;

[0039] Figure 2 , dynamic marker point identification program flow chart;

[0040] Figure 3 , marker ball fixing cap;

[0041] Figure 4 , neck movement collection system;

[0042] Figure 5 , marker point batch identification and marking software composition;

[0043] Figure 6 , trajectory extraction and analysis software composition.

[0044] BRIEF DESCRIPTION OF DRAWINGS:

[0045] 1, left and right lateral flexion trajectory; 2, forward and backward extension trajectory; 3, left and right rotation trajectory; 4, reflective marker ball; 5, inverted "T" fixing frame; 6, binding belt; 7, to-be-tested person; 8, marker ball fixing cap; 9, darkroom; 10, binocular camera; 11, processing computer; 12, data transmission line; 13, light supplement lamp; 14, marker point identification label module; 15, abnormality detection module; 16, marker point batch identification and marking software display interaction module; 17, three-dimensional measurement module; 18, trajectory analysis module; 19, trajectory extraction and analysis software display interaction module. DETAILED DESCRIPTION

[0046] In order to make the technical problems, technical solutions and advantages of the present application more clear, the following will be described in detail in combination with the drawings and specific embodiments.

[0047] As Figure 4As shown in the figure, a device integrating neck movement trajectory collection and movement symmetry analysis, characterized in that the collection and analysis device comprises a marker ball fixing cap 8 for marking head feature positions; a light supplement lamp 13 for improving data quality; a darkroom 9 for improving data quality; a binocular camera 10 for collecting data, a specific embodiment being two GIGE industrial cameras with the same parameters, collecting pixels of 1280*1024 and collecting frame rate of 33 frames / second; a processing computer 11 for storing and processing data; a data transmission line 12 connecting the binocular camera and the processing computer using an RJ-45 crystal head, a specific embodiment being a 10Base-T twisted pair line.

[0048] A device integrating neck movement trajectory collection and movement symmetry analysis, characterized in that three kinds of neck movements can be collected and analyzed, respectively being left and right lateral flexion, forward flexion and backward extension, and left and right rotation, as shown in the figure. Figure 1 As shown in the figure, the initial state of the three kinds of neck movements is a sitting neutral position, the left and right lateral flexion movement is that the neck is flexed to the left limit position in the coronal plane and then returned to the neutral position, and then flexed to the right limit position and returned to the neutral position; the forward flexion and backward extension movement is that the neck is first flexed to the limit position in the sagittal plane and then returned to the neutral position, and then extended to the limit position and returned to the neutral position; the left and right rotation movement is that the neck is first rotated to the left limit position in the transverse plane and then returned to the neutral position, and then rotated to the right limit position and returned to the neutral position.

[0049] As shown in the figure, a device integrating neck movement trajectory collection and movement symmetry analysis, characterized in that the device is designed for calculating the three-dimensional coordinates of the marker points in the dynamic data stream. Figure 5 As shown in the figure, a device integrating neck movement trajectory collection and movement symmetry analysis, characterized in that the device is designed for calculating the three-dimensional coordinates of the marker points in the dynamic data stream.

[0050] As shown in the figure, a device integrating neck movement trajectory collection and movement symmetry analysis, characterized in that the device is designed for calculating the three-dimensional coordinates of the marker points in the dynamic data stream. Figure 6As shown, a device integrating neck movement trajectory collection and movement symmetry analysis, characterized in that the trajectory extraction and analysis software for three kinds of neck movement trajectory extraction and analysis is designed. The software is installed on a processing computer 11, including an analysis module 18 for extracting and analyzing left and right lateral flexion, forward flexion and backward extension, and left and right rotation movement trajectories, the analysis module including three sub-modules corresponding to the three movements respectively; a display interaction module 19 for user to select operation directory and display movement analysis results. The user selects the target file directory and the neck movement to be analyzed through the display interaction module 19, and the corresponding sub-module in the analysis module 18 extracts and analyzes the trajectory, and displays the results to the display interaction module 19.

[0051] A method integrating neck movement trajectory collection and movement symmetry analysis, the use steps are as follows:

[0052] S1, the subject wears a marker ball fixing cap, and the spinous processes of the 7th cervical vertebra, the 5th cervical vertebra and the 2nd cervical vertebra are pasted with reflective marker points, in a dark room and the light supplement lamp is turned on, left and right lateral flexion, forward flexion and backward extension, and left and right rotation movement are performed, and a binocular stereovision device is used for recording.

[0053] S2, using the marker point batch recognition and labeling software to identify and label all the pictures collected by the binocular camera, and then calculating the three-dimensional coordinates of the marker points;

[0054] S3, arranging the coordinates of each marker point frame by frame to obtain the three-dimensional movement trajectory of the marker position;

[0055] S4, using the trajectory extraction and analysis software to analyze and extract the index of the space movement trajectory.

[0056] Step S2 adopts the designed marker point batch recognition and labeling software algorithm, as shown in the figure, including the following sub-steps: Figure 2

[0057] S21: sequentially read all the picture streams and perform ellipse detection and recognition on the contour of each frame of picture as a marker point target, and sort all the targets according to prior knowledge after detection;

[0058] S22: throw the first frame of picture, and display the coordinates of the identified marker points with a red cross in the user window, if the identification and sorting are correct, the user confirms, if not, the user uses the left mouse button and the keyboard "↑" "↓" keys to select and sort the marker points;

[0059] S23: after confirming that the marker points on the first frame of picture are correct, read backward, and the subsequent pictures are taken as the basis according to the coordinates of the marker points on the previous frame, if the coordinates of the marker points P h and the coordinates of the previous frame P h-1 satisfy ​If the recognition is incorrect, an exception is thrown;

[0060] S24: After the exception is thrown, the exception frame is displayed in the user window, and the user can adjust the marker point coordinates and the serial number according to the method in S22 until the last frame;

[0061] S25: After all the marker point coordinates are recognized, according to the coordinate recognition result of a marker point P1: left picture (u l , v l ), right picture (u r , v r ), the spatial coordinates of which are (X, Y, Z), according to the binocular vision principle, there are and for the right camera, there are where f l , f r , R and T are the left and right camera intrinsic matrix, rotation matrix and translation matrix obtained through camera calibration, respectively. The spatial coordinates (X, Y, Z) can be obtained by combining the two equations.

[0062] The above step S4 adopts the extracted motion symmetry measurement method based on circle fitting and spatial distance, taking left and right lateral flexion as an example, including the following sub-steps:

[0063] S41, fit the left and right lateral flexion trajectories as circular arcs, then the radius corresponding to the neutral position is R M , respectively calculate the central angle angles θ l and θ r corresponding to the left and right lateral flexion trajectories, then the calculation method of the left and right lateral flexion angle ratio θ ratio is as follows:

[0064]

[0065] S42, fold the left lateral flexion trajectory with R M as the axis, set θ r < θ l , take the integer part of θ r [θ r ], and draw a ray outward every 1 degree from the center of the circular arc (starting from the ray where R M is located), a total of [θ r ] rays, each ray intersects the two trajectory curves at two points, and the spatial distances d1, d2, …, d [θr] between the two points are calculated, and the average value D of all distances is taken as the symmetry measurement value of the trajectory curves;

[0066] S43, left and right lateral flexion average acceleration ratio: if the lengths of the left and right lateral flexion trajectories are S l and S r , and the total number of frames of left and right lateral flexion is I l and Ir If the camera frame rate K = 33, then the left-side bending average acceleration... Right-sided mean velocity Then the ratio of the mean lateral flexion accelerations to A ratio The calculation method is as follows:

[0067]

[0068] A ratio As a measure of the symmetry of the magnitude of the force exerted by the subject during left and right movements;

[0069] S44, Spatial location P of the marker point in the nth frame n (x n ,y n ,z n ), spatial location P of the (n+1)th frame n+1 (x n+1 ,y n+1 ,z n+1 If the instantaneous velocity V at this moment is given, then... n for:

[0070]

[0071] This allows us to plot the left and right lateral flexion velocity curves. By differentiating the velocity curves, we can plot the left and right lateral flexion acceleration curves, which can be used to visually display the motion state.

[0072] S45. Calculate the similarity index Sim for the left and right lateral flexion spatial trajectories. This index evaluates the symmetry of the two trajectory curves as a whole. The calculation method of Sim is shown in S451-S452. Integrate the motion symmetry description array [θ] ratio D, A ratio Sim describes the motion symmetry during left and right lateral flexion from multiple directions;

[0073] S46. After calculating the motion symmetry index, the relationship model between the neck motion imbalance coefficient and the influencing factors in the symmetry description array is obtained through an SVR neural network. In the experiment, 71 measurement data points (50 cases of normal neck motion and 21 cases of restricted neck motion) were used as the training set. Cross-validation showed that the regression model using the Gaussian kernel function yielded the best prediction results. The regression model had the following mean measurement error (mae), root mean square error (mse), and coefficient of determination (r). 2 The values ​​are 0.114, 0.018, and 0.911, respectively. After determining the model, input any symmetry description array to calculate its neck movement imbalance coefficient (close to 0 indicates good neck movement symmetry, close to 1 indicates severe neck movement asymmetry). For example:

[0074] input=[0.03549,12.25,0.088044,66.76]→output=[-0.02085159]

[0075] This data is from a normal person, and the output is -0.002.

[0076] The calculation method for the Sim index, a measure of similarity between left and right lateral flexion trajectories, is as follows:

[0077] S451. First, perform shortest path planning on the two space curves. Let L be the sequence of space points on the left side. i (i = 1, 2, ..., n), right-side buckling space point sequence R j (j=1,2,……m), calculate the distances between different positions in the two sequences, forming a distance matrix, where the elements (L) i R j ) corresponds to spatial point L i and spatial point R j The Euclidean distance;

[0078] S452. Define path W, where W is a path in the distance matrix from (L0, R0) to (L... n R m The path is found by traversing all paths in the distance matrix and finding the path with the minimum sum of path points. The average of all path points on this path is the index Sim.

Claims

1. A method of integrating neck motion trajectory acquisition and motion symmetry analysis, characterized in that The use steps are as follows: S1, the subject wears a marker ball fixing cap, and reflective marker points are pasted on the spinous processes of the 7th cervical vertebra, the 5th cervical vertebra and the 2nd cervical vertebra, respectively, in a dark room with a light supplement lamp turned on, left and right lateral flexion, forward and backward extension, left and right rotation movements are performed, and a binocular stereoscopic vision device is used for recording; S2, the marker point batch recognition and marking software is used to identify and label all the pictures collected by the binocular camera, and then the three-dimensional coordinates of the marker points are calculated; S3, the coordinates of each marker point are arranged frame by frame to obtain the three-dimensional motion trajectory of the marker position; S4, the trajectory extraction and analysis software is used to analyze the space motion trajectory and extract indexes; The indexes extracted in step S4 include: for left and right lateral flexion and left and right rotation movement, the left and right motion limit angle and the ratio, the left and right motion process swing amplitude, the left and right motion trajectory average distance, the left and right motion average acceleration ratio, the neck movement imbalance coefficient are calculated, respectively, and the left and right motion velocity and acceleration curves are drawn; for forward and backward extension movement, the forward and backward extension limit angle, the forward and backward extension process swing amplitude, the forward and backward extension average speed ratio are calculated, respectively, and the forward and backward extension trajectory curve, the forward and backward extension velocity and acceleration curve are drawn; The motion symmetry measurement method based on circle fitting and space distance is used in the above step S4, when calculating the left and right lateral flexion, including the following sub-steps: S41. Fit the left and right flexion trajectories to circular arcs, then the radius corresponding to the neutral position is R. M Calculate the central angle θ corresponding to the left and right lateral flexion trajectories respectively. l θ r Then the ratio of the left and right lateral flexion angles to θ ratio The calculation method is as follows: S42, the left side of the track to R M θ r θ l θ r θ r θ M θ r d1, d2……d [θr] D S43, left and right lateral bending average acceleration ratio: if the left and right lateral bending trajectory lengths are S l , S r , the left and right lateral bending total frame numbers are I l , I r , the camera frame rate K = 33, then the left lateral bending average acceleration right lateral bending average acceleration , the left and right lateral bending average acceleration ratio A ratio is calculated as follows: A ratio as a measure of the symmetry of the force exerted by the subject during the left-right movement; S44, mark the point the n frame space position P n (x n ,y n ,z n ), the n+1 frame space position P n+1 (x n+1 ,y n+1 ,z n+1 ), then the instantaneous speed V n at this moment is: The left and right lateral flexion velocity curve is drawn, the derivative of the velocity curve is calculated, the left and right lateral flexion acceleration curve is drawn, and the motion state is intuitively displayed; S45, calculate the left and right lateral flexion space trajectory similarity measure Sim, which evaluates the symmetry of the two trajectory curves as a whole, see S451-S452 for the calculation method of Sim; integrate the motion symmetry description array [θ ratio , D, A ratio , Sim], which describes the motion symmetry in the left and right lateral flexion process from multiple directions; S46, after the calculation of the motion symmetry index is completed, the relationship model between the neck movement imbalance coefficient and the influence factors in the symmetry description array is obtained through the SVR neural network; The calculation method of the left and right lateral flexion space trajectory similarity measurement index Sim is: S451、First, the shortest path planning is performed on two spatial curves, and let the left side flexion spatial point sequence L i , i = 1, 2, … n; the right side flexion spatial point sequence R j , j = 1, 2, … m; the distance corresponding to the points at different positions of the two sequences is calculated, to form a distance matrix, and the element (L i , R j ) in the matrix corresponds to the Euclidean distance between the spatial point L i and the spatial point R j ; S452, define a path W, W is a path from (L0, R0) to (L n , m ) in the distance matrix; traverse all paths in the distance matrix to find a path with the smallest sum of path points, the average of all path points on this path is the indicator Sim.

Images