A mouse swimming-based exercise state analysis method and system

Abstract

The application discloses a behavior analysis method and system based on a swimming state of a mouse. The behavior analysis method based on the swimming state of the mouse comprises the following steps: constructing a motion scene of mouse swimming, which comprises a motion endurance evaluation motion scene and a motion coordination motion scene; acquiring posture video data of the mouse in the swimming state in the motion scene; identifying positions of each skeletal joint of the mouse based on the posture video data, and proposing coordinate values of the corresponding skeletal joint at different moments; establishing a data collection of the coordinate values of each skeletal joint of the mouse based on the coordinate values; identifying a time frequency and a number of behavior occurrences of the corresponding skeletal joint based on the data collection, comparing differences between various behaviors, and generating motion change rule data of the swimming state of the mouse. The application can effectively improve the problems that a human subjective factor is relatively large and an observation level is single when analyzing the swimming motion state of the mouse at present.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of animal motion state analysis, in particular to a motion state analysis method and system based on mouse swimming. BACKGROUND

[0002] Many nervous system diseases (such as multiple sclerosis, Huntington's disease and spinal cord injury) and musculoskeletal diseases (such as arthritis, ligament injury, etc.) can cause obvious motor dysfunction in patients and mouse models, and change the general movement pattern. At present, the evaluation of mouse model movement endurance and movement coordination is used for the research of the intervention effect of some extracted components, small molecule substances, hormones, etc. on mice, such as angelica polysaccharide, Jinqiu capsule and decapeptide CMS001, etc.

[0003] At present, the behavior analysis method for evaluating the movement endurance and movement coordination of mouse models at home and abroad for drug screening and pharmacodynamic detection research mainly includes forced swimming test, weight-bearing swimming test, rotating rod test, water immersion test, etc., among which the most classic one is the weight-bearing swimming test. Swimming is a typical form of movement behavior. For example, the mouse has the characteristics of strong hind limb paddling, consistent and coordinated movement, which provides the main power for forward movement, while the front paws are usually stationary below the chest, and the horizontal sinusoidal oscillation of the tail and occasional front limb paddling may be important for navigation and stability during swimming. By observing the behavior indicators of mice such as exhaustion time, swimming speed, swimming path, etc., the movement endurance and movement coordination of mice are evaluated, and movement defects are detected.

[0004] At present, the evaluation methods for evaluating the movement endurance and movement coordination of mouse models are: scale scoring method, video acquisition method. Among them, the results obtained by the scale scoring method are often nonlinear and subjective, with low sensitivity and easy to be disturbed by compensatory movement. In addition, the method of recording the process of completing a specific behavior task of the mouse model by video acquisition does not process the mouse model hair color, and has a processing method of removing the original hair of the model mouse and marking the key points with color. These two methods both have the problems of large subjectivity in judging the position of key points, easy to cause the death of mice, and high cost. At the same time, the operation of removing the hair is not only complex, but also will interfere with the normal physiological state of the mouse, and further affect the behavior performance of the mouse and interfere with the experimental results. Although the video acquisition method records a large amount of behavior information of the mouse model, due to the inaccuracy of direct observation by the naked eye and the limitation of small number and single level of behavior indicators, it is difficult to interpret the vast amount of behavior information contained in the video to a greater extent. At present, the observation of mouse behavior based on human eyes has the defects of time-consuming and laborious, large subjectivity, non-uniform standard, small number of indicators, single level, low accuracy, poor experimental repeatability, etc., and it is difficult to compare the observation results of different laboratories, which increases the difficulty of behavior analysis. SUMMARY

[0005] In view of the above prior art defects, the purpose of the present application is to disclose a mouse swimming-based motion state analysis method and system to improve the current problems of greater human subjective factors and single observation level in analyzing the swimming motion state of mice.

[0006] To achieve the above object and other related objects, the present application discloses a mouse swimming state-based behavior analysis method, which comprises:

[0007] Constructing a mouse swimming motion scene, which includes a motion endurance evaluation motion scene and a motion coordination motion scene;

[0008] Obtaining posture video data of a mouse in a swimming state in the motion scene;

[0009] Based on the posture video data, identifying the positions of each skeletal joint of the mouse, and proposing the coordinate values of the corresponding skeletal joint at different time points;

[0010] Based on the coordinate values, establishing a data collection of the coordinate values of each joint of the mouse;

[0011] Based on the data collection, identifying the time frequency and number of occurrence of behaviors of the corresponding skeletal joint, comparing the differences between various behaviors, and generating motion change rule data of the mouse swimming state.

[0012] In one scheme of the present application, in constructing the motion endurance evaluation motion scene, it comprises:

[0013] Constructing a swimming area of the mouse, and setting a video acquisition device for acquiring the state of the mouse on one side of the swimming area;

[0014] Presetting the water temperature range and the water depth range of the swimming area, wherein the water temperature range comprises 25±1℃, and the water depth comprises 35±1cm;

[0015] Setting a counterweight, and the mass of the counterweight is 5% of the body weight of the mouse.

[0016] In one scheme of the present application, in constructing the motion coordination motion scene, it comprises:

[0017] Constructing a rectangular test box, and the test box is made of transparent material;

[0018] The video acquisition device is located at the bottom of the test box, and is used for acquiring posture video data of the mouse in the test box during motion;

[0019] A first adjusting mirror;

[0020] A second adjustment mirror is located to the side of the first adjustment mirror, and the second adjustment mirror and the first adjustment mirror are set at an angle; and

[0021] The third adjustment mirror is located on the other side of the first adjustment mirror, and the third adjustment mirror and the first adjustment mirror are set at an angle.

[0022] The angle between the second or third adjusting mirror and the first adjusting mirror can be adjusted.

[0023] In one aspect of the present invention, in the step of identifying the positions of various skeletal joints of the mouse based on the posture video data and proposing the coordinate values ​​of the corresponding skeletal joints at different times:

[0024] The locations of the skeletal joints include: the tip of the nose, left ear, right ear, left eye, right eye, center of the head, neck, center of the torso, root of the tail, middle of the tail, tip of the tail, iliac crest, buttock, hind ankle, tip of the hind toe, shoulder, wrist, elbow, and tip of the front toe; and the coordinate system corresponding to the coordinate values ​​of the skeletal joints is set by the resolution of the video acquisition device.

[0025] In one aspect of the present invention, based on the mouse trunk skeleton, the skeletal joints are classified and mapped with corresponding trunk classification labels.

[0026] In one aspect of the present invention, in the step of identifying the frequency and number of behaviors occurring at corresponding skeletal joint points based on the data set, comparing the differences among various behaviors, and generating data on the movement change patterns of mouse swimming states, the data on the movement change patterns of mouse swimming states includes:

[0027] Mouse exercise endurance data, including exercise duration, rest duration, trunk angle, yaw angle, extension angle, tail flexion angle, swimming distance, and swinging distance; and

[0028] Mouse motor coordination data, including time index data, vertical joint movement and height index data, stride length and extension distance index data, paw-to-body center angle index data, and body segment angle index data.

[0029] The present invention also provides a behavioral analysis system based on the swimming state of mice, which includes an experimental module, a data transmission module, a video analysis module, a storage module, and an output module;

[0030] The video analysis module is used to analyze the tracking video data of the mouse, and includes:

[0031] Acquire posture video data of mice swimming in a motion scenario;

[0032] Based on the posture video data, the positions of each skeletal joint of the mouse are identified, and the coordinate values of the corresponding skeletal joint at different time points are proposed;

[0033] Based on the coordinate values, a data set of the coordinate values of each joint of the mouse is established;

[0034] Based on the data set, the time and frequency of the behavior of the corresponding skeletal joint are identified, the differences between various behaviors are compared, and the motion change rule data of the swimming state of the mouse is generated.

[0035] The application discloses a motion state analysis method and system based on mouse swimming, which is based on two mouse swimming motion modes of weight-bearing swimming and spontaneous swimming, and identifies and tracks multiple body joints, captures fine behavior indicators of the mouse from multiple dimensions, quantifies the head, trunk and limb motion of the mouse, and comprehensively analyzes different motion modes of the mouse through fine behavior indicators such as motion speed, motion path, forelimb swing speed and tail position, so as to evaluate the motion endurance and motion coordination of the mouse. Meanwhile, the design of the three mirrors reduces the use of the camera, reduces the site requirement of the experiment, and reduces the manpower and material requirement of the experiment. BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0037] Fig. 1 A schematic diagram of a mouse swimming scene in an embodiment of the behavior analysis method based on the mouse swimming state of the present application;

[0038] Fig. 2 A flowchart in an embodiment of the behavior analysis method based on the mouse swimming state of the present application;

[0039] Fig. 3 A module diagram in an embodiment of the behavior analysis system based on the mouse swimming state of the present application.

[0040] Element number explanation

[0041] 101, first adjusting mirror; 102, second adjusting mirror; 103, third adjusting mirror;

[0042] 100, experiment module; 200, data acquisition module; 300, data transmission module; 400, video analysis module; 500, storage module; 600, output module. DETAILED DESCRIPTION

[0043] The present application is described herein with reference to specific embodiments thereof, which are illustrative of the principles of the present application. Other advantages and embodiments of the present application will become apparent to those skilled in the art from the following detailed description, taken in conjunction with the accompanying drawings. The detailed description and specific examples are given for the purpose of illustration only and are not intended to limit the scope of the present application.

[0044] Referring to Figs. 1-3 It is to be understood that the structures, proportions, sizes, etc. shown in the drawings are to be interpreted as illustrative only and the scope of the present application is not to be limited to the specific illustrations set forth herein. Any modification, change, substitution, or variation of the structures, proportions, sizes, etc. shown in the drawings, as well as any other modifications, changes, substitutions, or variations of the methods and / or techniques described herein, are intended to be included within the scope of the present application.

[0045] Referring to Figs. 1-2 As shown in the drawings, the present application discloses a behavior analysis method based on mouse swimming state, which can be used to improve the problem that the subjective factor is large and the observation level is single when analyzing the swimming state of the mouse. Specifically, by comprehensively and accurately analyzing the swimming behavior of the mouse in different forms of swimming movement, it is necessary to comprehensively evaluate the exercise tolerance and exercise coordination of the mouse, and it is also a prerequisite to link the neuroanatomical changes with the functional results. Among them, for the exercise state of the mouse, it includes normal exercise state and weight-bearing exercise state. By analyzing and processing the two exercise states of the mouse, the accuracy of the behavior data of the mouse can be effectively improved.

[0046] Referring to Fig. 2 As shown in the drawings, in an embodiment, the behavior analysis method based on mouse swimming state includes the following steps.

[0047] Firstly, step S10 is executed to construct the exercise scene of the mouse swimming, which includes the exercise tolerance evaluation exercise scene and the exercise coordination exercise scene. It can be understood that the exercise tolerance evaluation exercise scene is used to acquire the exercise behavior data of the mouse in the weight-bearing state, and the exercise coordination exercise scene is used to acquire the exercise behavior data of the mouse in the normal exercise state.

[0048] Specifically, in constructing the exercise endurance evaluation exercise scene, a swimming area for mice can be allowed to be constructed, and a video acquisition device for acquiring the exercise state of the mice is arranged on one side of the swimming area of the mice. The specific signal and category of the video acquisition device can be allowed to be defined according to actual needs.

[0049] The swimming area of the mice can be a cylindrical barrel made of glass, for example, the diameter of the cylindrical barrel is 25 cm, and the height is 40 cm. At the same time, the video acquisition device for acquiring the exercise state of the mice can be located on one side of the barrel, so as to ensure that the mice are located in the capture picture of the video acquisition device during the swimming process.

[0050] It should be noted that when analyzing the exercise endurance of the mice, the weight of the weight is 5% of the body weight of the mice. At the same time, the water temperature in the cylindrical barrel is 25±1℃, and the depth of the water in the cylindrical barrel is 35±1cm.

[0051] Specifically, in constructing the exercise coordination exercise scene, a rectangular test box can be allowed to be constructed, and the test box is made of transparent material. Then, the video acquisition device is arranged at the bottom of the test box, which is used to acquire the posture video data of the mice in the test box during exercise.

[0052] It should be noted that a plurality of adjusting mirrors can be arranged on one side of the test box. The plurality of adjusting mirrors can include a first adjusting mirror 101, a second adjusting mirror 102, and a third adjusting mirror 103, and the second adjusting mirror 102 and the third adjusting mirror 103 are respectively located on both sides of the first adjusting mirror 101. At the same time, the second adjusting mirror 102 and the third adjusting mirror 103 are respectively arranged at an angle with the first adjusting mirror 101, and the second adjusting mirror 102 and the third adjusting mirror 103 are respectively hinged with the first adjusting mirror 101.

[0053] In the mouse motor coordination scene, the video acquisition device is perpendicular to the first adjustment mirror 101, and the first adjustment mirror 101 is parallel to the water surface of the test box. Therefore, when the mouse swims in the test box, the video acquisition device can be allowed to acquire the posture video data of the mouse in the abdominal direction. At the same time, the second adjustment mirror 102 and the third adjustment mirror 103 are located on both sides of the video acquisition device, and the second adjustment mirror 102 and the third adjustment mirror 103 are symmetrically arranged. It should be noted that the mouse image located in the second adjustment mirror 102 or the third adjustment mirror 103 is located within the acquisition range of the video acquisition device, so that the video acquisition device can be allowed to acquire the swimming posture of the mouse on the side through the second adjustment mirror 102 or the third adjustment mirror 103. Therefore, by arranging the second adjustment mirror 102 and the third adjustment mirror 103, the video acquisition device can simultaneously acquire the posture data of the bottom and two sides of the mouse during swimming.

[0054] Then, step S20 is performed to acquire the posture video data of the mouse in the swimming state in the movement scene. After the mouse swimming movement scene is constructed, the mouse is placed in the corresponding movement scene. When the mouse swims, the video acquisition device acquires the movement state of the mouse, and further obtains the posture video data of the mouse in the swimming state.

[0055] Further, step S30 is performed to identify the positions of the respective skeletal joints of the mouse based on the posture video data, and to propose the coordinate values of the corresponding skeletal joints at different time points.

[0056] Specifically, the positions of the skeletal joints of the mouse include the nose tip, the left ear, the right ear, the left eye, the right eye, the head center, the neck, the trunk center, the tail root, the middle of the tail, the tail end, the iliac crest, the hip, the hind ankle, the hind toe tip, the shoulder, the wrist, the elbow, and the front toe tip. After the video acquisition device acquires the posture video data of the mouse swimming, the posture video data of the mouse is analyzed and processed to identify the positions of the respective skeletal joints of the mouse. The coordinate system corresponding to the coordinate values of the skeletal joints is set by the resolution of the video acquisition device.

[0057] It should be noted that the cloud analysis system can be allowed to analyze and process the posture video data of the mouse to identify the positions of the skeletal joints of the mouse. The positions of the skeletal joints of the mouse can be allowed to be based on the image frames of the video recorded by the video acquisition device to establish a coordinate system. According to the established coordinate system, the specific coordinates of the respective skeletal joints of the mouse in the corresponding coordinate system are further determined.

[0058] The coordinate system is established based on the photo size of each frame of the posture video data of the mouse. For example, when the resolution of each frame of the posture video data of the mouse is 1920*1080, a two-dimensional coordinate system can be established with the 1920 value as the Y axis and the 1080 value as the X axis. Therefore, the corresponding bone joints of the mouse correspond to specific coordinate data at the corresponding time.

[0059] To improve the analysis effect of specific data of the mouse in the swimming state, the plurality of bone joints of the mouse can be classified and mapped with the corresponding trunk classification label based on the trunk skeleton of the mouse. For example, the head-neck segment, shoulder-elbow segment, tail base-tail end segment and the like can be defined as trunk labels, and the bone joints of the mouse meeting the corresponding trunk labels can be labeled. By classifying the plurality of bone joints of the mouse, the observation of the swimming posture of the mouse can be effectively improved.

[0060] Further, step S40 is performed to establish a data set of the coordinate values of the respective joints of the mouse based on the coordinate values.

[0061] Specifically, after obtaining the coordinate values of the respective bone joints of the mouse, a data set of the respective bone joints of the mouse can be established.

[0062] Finally, step S50 is performed to identify the time and frequency of the behavior of the corresponding bone joint, compare the differences between various behaviors, and generate the motion change rule data of the swimming state of the mouse based on the data set. The motion change rule data of the swimming state of the mouse includes mouse motion endurance data and mouse motion coordination data. The mouse motion endurance data includes motion time, rest time, trunk angle, deflection angle, stretching angle, tail bending angle, swimming motion distance, and swing motion distance. The mouse motion coordination data includes time index data, vertical joint motion and height index data, step length and stretching distance index data, paw and body center angle index data, and body segment angle index data.

[0063] Specifically, for mouse exercise endurance data, the exercise duration includes the frame number corresponding to the mouse tail root speed greater than or equal to 1 cm / s, and the total exercise duration is calculated. The static duration includes the frame number corresponding to the mouse tail root speed less than 1 cm / s, and the total static duration is calculated. The trunk angle includes taking the mouse nose coordinate point A, taking the mouse tail root coordinate point B, taking the mouse tail root as the x-axis vertical line, and taking the mouse nose as the y-axis vertical line, and the two vertical lines intersect at point C, tan a = BC / AC, and the angle a is the average trunk angle. The deflection angle includes the midpoint of the two hind limbs as point A, and the angle between the line connecting the nose and point A and the Y-axis is the deflection angle. The stretch angle includes the angle formed by the mouse tail root-trunk center-nose three points. The tail bending angle includes the angle formed by the tail root-tail middle-tail tip three points. The swimming movement distance includes the sum of the X value difference of the mouse tail root coordinates of adjacent two frames, that is, (X2-X1)+(X3-X2)+…+(Xn+1-Xn). And the swing movement distance includes the sum of the Y value difference of the mouse tail root coordinates of adjacent two frames, that is, (Y2-Y1)+(Y3-Y2)+…+(Yn+1-Yn).

[0064] Further, for mouse exercise coordination data, it includes time index data, vertical joint movement and height index data, step and stretch distance index data, paw and body center angle index data, and body segment angle index data.

[0065] In an embodiment, the time type index can include average step time, left front limb duration, left rear limb duration, right front limb duration, right rear limb duration, average step length, left front rear right synchronization ratio, and right front left rear synchronization ratio. Among them, the left front limb duration includes the time average value of the Xth step of the left front limb-the time average value of the X-1th step of the left front limb; the left rear limb duration includes the time average value of the Xth step of the left rear limb-the time average value of the X-1th step of the left rear limb; the right front limb duration includes the time average value of the Xth step of the right front limb-the time average value of the X-1th step of the right front limb; the right rear limb duration includes the time average value of the Xth step of the right rear limb-the time average value of the X-1th step of the right rear limb; the average step length includes the length average value of the Xth step-the length average value of the X-1th step; the left front rear right synchronization ratio includes: 1-time(left front limb right rear limb synchronization / time(left front limb right rear limb synchronization+left front limb right rear limb asynchronization)); and the right front left rear synchronization ratio: 1-time(right front limb left rear limb synchronization / time(right front limb left rear limb synchronization+right front limb left rear limb asynchronization)).

[0066] Further, the vertical joint movement and height index data includes:

[0067] 1. The average movement height of the left rear ankle is the average((the height of the left rear ankle recognition point)) / step number;

[0068] 2. Left rear ankle vertical movement height, which is max (height of left rear ankle identification point) - min (height of left rear ankle identification point) / number of steps;

[0069] 3. Left rear toe tip average movement height, which is average ((height of left rear toe tip identification point)) / number of steps;

[0070] 4. Left rear toe tip vertical movement height, which is max (height of left rear toe tip identification point) - min (height of left rear toe tip identification point) / number of steps;

[0071] 5. Left front toe tip average movement height, which is average ((height of left front toe tip identification point)) / number of steps;

[0072] 6. Left front toe tip vertical movement height, which is max (height of left front toe tip identification point) - min (height of left front toe tip identification point) / number of steps;

[0073] 7. Left head average movement height, which is average ((height of head center identification point)) / number of steps;

[0074] 8. Left hip average movement height, which is average ((height of left hip identification point)) / number of steps;

[0075] 9. Left hip vertical movement height, which is max (height of left hip tip identification point) - min (height of left hip tip identification point) / number of steps;

[0076] 10. Left forelimb average movement height, which is average ((height of left forelimb identification point)) / number of steps;

[0077] 11. Left forelimb vertical movement height, which is max (height of left forelimb tip identification point) - min (height of left forelimb tip identification point) / number of steps;

[0078] 12. Right rear ankle average movement height, which is average ((height of right rear ankle identification point)) / number of steps;

[0079] 13. Right rear ankle vertical movement height, which is max (height of right rear ankle identification point) - min (height of right rear ankle identification point) / number of steps;

[0080] 14. Right rear toe tip average movement height, which is average ((height of right rear toe tip identification point)) / number of steps;

[0081] 15. Right rear toe tip vertical movement height, which is max (height of right rear toe tip identification point) - min (left rear toe tip identification point) / number of steps;

[0082] 16. Average height of right front toe tip movement, which is average (height of right front toe tip identification point) / number of steps;

[0083] 17. Vertical height of right front toe tip movement, which is max (height of right front toe tip identification point) - min (height of right front toe tip identification point) / number of steps;

[0084] 18. Average height of right front leg movement, which is average (height of right front leg identification point) / number of steps; and

[0085] 19. Vertical height of right front leg movement, which is max (height of right front leg tip identification point) - min (height of right front leg tip identification point) / number of steps.

[0086] Further, the step length and stride distance indicator data includes:

[0087] 1. Average length of left hind leg, which is average (distance (last 1 step of left hind leg - 1st step of left hind leg));

[0088] 2. Median length of left hind leg movement, which is median (distance (last 1 step of left hind leg - 1st step of left hind leg));

[0089] 3. Total horizontal movement of left hind leg, which is max (distance (last 1 step of left hind leg - 1st step of left hind leg));

[0090] 4. Maximum extension of left hind leg, which is max (distance (last extension of left hind leg to beginning of extension of left hind leg));

[0091] 5. Maximum contraction of left hind leg, which is max (distance (last contraction of left hind leg to beginning of contraction of left hind leg));

[0092] 6. Average length of left front leg, which is average (distance (last 1 step of left front leg - 1st step of left front leg));

[0093] 7. Median length of left front leg movement, which is median (distance (last 1 step of left front leg - 1st step of left front leg));

[0094] 8. Total horizontal movement of left front leg, which is max (distance (last 1 step of left front leg - 1st step of left front leg));

[0095] 9. Maximum extension of left front leg, which is max (distance (last extension of left front leg to beginning of extension of left front leg));

[0096] 10. Maximum contraction of left front leg, which is max (distance (last contraction of left front leg to beginning of contraction of left front leg));

[0097] 11. Average length of right hind leg, which is average (distance (last 1 step of left hind leg - 1st step of left hind leg));

[0098] 12. Right hind limb movement length median, which is median (distance (last 1 step of left hind limb - 1st step of left hind limb));

[0099] 13. Right hind limb total horizontal movement, which is max (distance (last 1 step of left hind limb - 1st step of left hind limb));

[0100] 14. Right hind limb maximum extension, which is max (distance (last extension of left hind limb to the beginning of extension of left hind limb));

[0101] 15. Right hind limb maximum contraction, which is max (distance (last contraction of right hind limb to the beginning of contraction of right hind limb));

[0102] 16. Right front limb average length, which is average (distance (last 1 step of right front limb - 1st step of right front limb));

[0103] 17. Right front limb movement length median, which is median (distance (last 1 step of right front limb - 1st step of right front limb));

[0104] 18. Right front limb total horizontal movement, which is max (distance (last 1 step of right front limb - 1st step of right front limb));

[0105] 19. Right front limb maximum extension, which is max (distance (last extension of right front limb to the beginning of extension of right front limb));

[0106] 20. Right front limb maximum contraction, which is max (distance (last contraction of right front limb to the beginning of contraction of right front limb));

[0107] 21. Right hip total horizontal movement, which is max (distance (last contraction of right hip to the beginning of contraction of right hip));

[0108] 22. Right hip average length, which is average (distance (last contraction of right hip to the beginning of contraction of right hip)).

[0109] Further, the paw and body center angle indicators include:

[0110] 1. Left hind limb average angle, which is average (atan2 (left hind limb coordinate Y value - body center coordinate Y value, left hind limb coordinate X value - body center coordinate X value)) / step number;

[0111] 2. Left hind limb maximum angle, which is max (atan2 (left hind limb coordinate Y value - body center coordinate Y value, left hind limb coordinate X value - body center coordinate X value)) / step number;

[0112] 3. left hind leg minimum angle, which is min(atan2(left hind leg coordinate Y value - body center coordinate Y value, left hind leg coordinate X value - body center coordinate X value)) / step number;

[0113] 4. left front leg average angle, which is average(atan2(left front leg coordinate Y value - body center coordinate Y value, left front leg coordinate X value - body center coordinate X value)) / step number;

[0114] 5. left front leg maximum angle, which is max(atan2(left front leg coordinate Y value - body center coordinate Y value, left front leg coordinate X value - body center coordinate X value)) / step number;

[0115] 6. left front leg minimum angle, which is min(atan2(left front leg coordinate Y value - body center coordinate Y value, left front leg coordinate X value - body center coordinate X value)) / step number;

[0116] 7. right hind leg average angle, which is average(atan2(right hind leg coordinate Y value - body center coordinate Y value, right hind leg coordinate X value - body center coordinate X value)) / step number;

[0117] 8. right hind leg maximum angle, which is max(atan2(right hind leg coordinate Y value - body center coordinate Y value, right hind leg coordinate X value - body center coordinate X value)) / step number;

[0118] 9. right hind leg minimum angle, which is min(atan2(right hind leg coordinate Y value - body center coordinate Y value, right hind leg coordinate X value - body center coordinate X value)) / step number;

[0119] 10. right front leg average angle, which is average(atan2(right front leg coordinate Y value - body center coordinate Y value, right front leg coordinate X value - body center coordinate X value)) / step number;

[0120] 11. right front leg maximum angle, which is max(atan2(right front leg coordinate Y value - body center coordinate Y value, right front leg coordinate X value - body center coordinate X value)) / step number;

[0121] 12. right front leg minimum angle, which is min(atan2(right front leg coordinate Y value - body center coordinate Y value, right front leg coordinate X value - body center coordinate X value)) / step number.

[0122] Further, the body segment angle index data includes:

[0123] 1. Left hind hip-ankle-toe average angle, which is the average of (atan2((left hind toe Y coordinate - left hind ankle Y coordinate), (left hind toe X coordinate - left hind ankle X coordinate)) - atan2((hip Y coordinate - hind ankle Y coordinate), (hip X coordinate - hind ankle X coordinate))) / step count;

[0124] 2. Left front elbow-wrist-toe tip average angle, which is the average of (atan2((left front toe tip Y coordinate - left front wrist Y coordinate), (left front toe tip X coordinate - left front wrist X coordinate)) - atan2((left front elbow Y coordinate - left front wrist Y coordinate), (left front elbow X coordinate - left front wrist X coordinate))) / step count;

[0125] 3. Left front shoulder-elbow-wrist average angle, which is the average of (atan2((left front wrist Y coordinate - left front coordinate elbow Y coordinate), (left front wrist X coordinate - left front coordinate elbow X coordinate)) - atan2((left front shoulder Y coordinate - left front elbow coordinate Y coordinate), (left front shoulder X coordinate - left front elbow coordinate X coordinate))) / step count;

[0126] 4. Left hind hip-ankle-toe angle maximum, which is max(atan2((left hind toe Y coordinate - left hind ankle Y coordinate), (left hind toe X coordinate - left hind ankle X coordinate)) - atan2((left hind hip Y coordinate - left hind ankle Y coordinate), (left hind hip X coordinate - left hind ankle X coordinate))) / step count;

[0127] 5. Left front shoulder-elbow-wrist angle maximum, which is max(atan2((left front wrist Y coordinate - left front elbow coordinate Y coordinate), (left front wrist X coordinate - left front elbow coordinate X coordinate)) - atan2((left front shoulder coordinate Y coordinate - left front elbow coordinate Y coordinate), (left front shoulder coordinate X coordinate - left front elbow coordinate X coordinate)) / step count;

[0128] 6. Left hind hip-ankle-toe angle minimum, which is min(atan2((left hind toe Y coordinate - left hind ankle Y coordinate), (left hind toe X coordinate - left hind ankle X coordinate)) - atan2((left hind hip coordinate Y coordinate - left hind ankle Y coordinate), (left hind hip coordinate X coordinate - left hind ankle X coordinate))) / step count;

[0129] 7. Left front elbow-wrist-toe tip angle minimum, which is min(atan2((left front toe tip Y coordinate - left front wrist Y coordinate), (left front toe tip X coordinate - left front wrist X coordinate)) - atan2((left front elbow Y coordinate - left front wrist Y coordinate), (left front elbow X coordinate - left front wrist X coordinate))) / step count;

[0130] 8. left front shoulder-elbow-wrist angle minimum, which is min(atan2((left front wrist Y value - left front elbow Y value), (left front wrist X value - left front elbow X value)) - atan2((left front shoulder Y value - left front elbow Y value), (left front shoulder X value - left front elbow X value))) / step count;

[0131] 9. right rear hip-ankle-toe average angle, which is average(atan2((right rear toe Y value - right rear ankle Y value), (right rear toe X value - right rear ankle X value)) - atan2((hip Y value - rear ankle Y value), (hip X value - rear ankle X value)) / step;

[0132] 10. right front elbow-wrist-toe tip average angle, which is average(atan2((right front toe tip Y value - right front wrist Y value), (right front toe tip X value - right front wrist X value)) - atan2((right front elbow Y value - right front wrist Y value), (right front elbow X value - right front wrist X value)) / step count;

[0133] 11. right front shoulder-elbow-wrist average angle, which is average(atan2((right front wrist Y value - right front elbow Y value), (right front wrist X value - right front elbow X value)) - atan2((right front shoulder Y value - right front elbow Y value), (right front shoulder X value - right front elbow X value)) / step count;

[0134] 12. right rear hip-ankle-toe angle maximum, which is max(atan2((right rear toe Y value - right rear ankle Y value), (right rear toe X value - right rear ankle X value)) - atan2((right rear hip Y value - right rear ankle Y value), (right rear hip X value - right rear ankle X value)) / step count;

[0135] 13. right front shoulder-elbow-wrist angle maximum, which is max(atan2((right front wrist Y value - right front elbow Y value), (right front wrist X value - right front elbow X value)) - atan2((right front shoulder Y value - right front elbow Y value), (right front shoulder X value - right front elbow X value)) / step count;

[0136] 14. right rear hip-ankle-toe angle minimum, which is min(atan2((right rear toe Y value - right rear ankle Y value), (right rear toe X value - right rear ankle X value)) - atan2((right rear hip Y value - right rear ankle Y value), (right rear hip X value - right rear ankle X value)) / step count;

[0137] 15. A minimum right front elbow wrist toe tip angle, which is min(atan2((right front toe tip coordinate Y value - right front wrist coordinate Y value), (right front toe tip coordinate X value - right front wrist coordinate X value)) - atan2((right front elbow Y value - right front wrist Y value), (right front elbow X value - right front wrist X value)) / step number;

[0138] 16. A minimum right front shoulder elbow wrist angle, which is min(atan2((right front wrist coordinate Y value - right front elbow coordinate Y value), (right front wrist coordinate X value - right front elbow coordinate X value)) - atan2((right front shoulder coordinate Y value - right front elbow coordinate Y value), (right front shoulder coordinate X value - right front elbow coordinate X value)) / step number.

[0139] In an embodiment, the present application further provides a behavior analysis system based on mouse swimming state, which comprises an experiment module 100, a data acquisition module 200, a data transmission module 300, a video analysis module 400, a storage module 500, and an output module 600.

[0140] Specifically, the experiment module 100 is used to realize the movement of the mouse, and the swimming video data of the mouse is acquired by the data acquisition module 200. The data transmission module 300 is used to transmit the posture video data of the swimming of the mouse to the video analysis module 400. The video analysis module 400 is used to analyze the swimming video data of the mouse, which comprises:

[0141] acquiring the posture video data of the mouse in the swimming state in the movement scene;

[0142] based on the posture video data, identifying the positions of each skeletal joint of the mouse, and proposing the coordinate values of the corresponding skeletal joint at different time points;

[0143] based on the coordinate values, establishing the data collection of the coordinate values of each joint of the mouse;

[0144] based on the data collection, identifying the time frequency and the number of occurrence of the behavior of the corresponding skeletal joint, comparing the differences between various behaviors, and generating the movement change rule data of the swimming state of the mouse.

[0145] It should be noted that the movement change rule data of the swimming state of the mouse generated by the video analysis module 400 is stored in the storage module 500, and is output by the output module 600.

[0146] In summary, the application discloses a motion state analysis method and system based on mouse swimming, which is based on two mouse swimming modes (weight-bearing swimming and spontaneous swimming), and a plurality of body joint points are identified and tracked, fine behavior indexes of the mouse are captured from multiple dimensions, motion of a head, a trunk and four limbs of the mouse is quantified, different motion modes of the mouse are comprehensively analyzed through fine behavior indexes such as motion speed, motion path, forelimb swing speed and tail position, and motion endurance and motion coordination of the mouse are evaluated. Meanwhile, the design of the three mirrors reduces the use of the camera, reduces the site requirement of the experiment, and reduces the manpower and material requirement of the experiment.

[0147] Therefore, the problem that a subjective factor is large and observation levels are single in analyzing the swimming motion state of the mouse at present can be effectively improved.

[0148] Therefore, the application effectively overcomes some practical problems in the prior art, and has high utilization value and use significance.

[0149] The above examples only exemplarily illustrate the principles and effects of the application, and are not used to limit the application. Any person skilled in the art can modify or change the above examples without departing from the spirit and category of the application. Therefore, all equivalent modifications or changes completed by those skilled in the art without departing from the spirit and technical thought disclosed by the application should be covered by the claims of the application.

Claims

1. A method for analyzing the exercise state based on mouse swimming, characterized by, The application comprises: constructing a swimming exercise scene for mice, including an exercise endurance evaluation exercise scene and an exercise coordination exercise scene; acquiring posture video data of mice in a swimming state in the exercise scene; based on the posture video data, identifying the positions of each skeletal joint of the mice, and proposing coordinate values of the corresponding skeletal joints at different time points; based on the coordinate values, establishing a data collection of coordinate values of each joint of the mice; based on the data collection, identifying the duration and frequency of occurrence of behaviors of the corresponding skeletal joints, comparing the differences between various behaviors, and generating exercise change rule data of the swimming state of the mice; in the step of identifying the duration and frequency of occurrence of behaviors of the corresponding skeletal joints based on the data collection, comparing the differences between various behaviors, and generating exercise change rule data of the swimming state of the mice, the exercise change rule data of the swimming state of the mice comprises: mouse exercise endurance data, including exercise duration, static duration, trunk angle, deflection angle, stretching angle, tail bending angle, swimming exercise distance, and swing exercise distance; and mouse exercise coordination data, including time index data, vertical joint exercise and height index data, step length and stretching distance index data, paw and body center angle index data, and body segment angle index data.

2. The mouse swimming-based exercise state analysis method according to claim 1, characterized by, In constructing the exercise endurance evaluation exercise scene, it comprises: constructing a swimming area for mice, and a video acquisition device for acquiring the state of the mice is arranged on one side of the swimming area; presetting the water temperature range and the water depth range of the swimming area, wherein the water temperature range comprises 25±1℃, and the water depth comprises 35±1cm; setting a counterweight, and the mass of the counterweight is 5% of the body weight of the mouse.

3. The mouse swimming-based exercise state analysis method according to claim 1, characterized by, In constructing the exercise coordination exercise scene, it comprises: constructing a rectangular test box, and the test box is made of transparent material; a video acquisition device is arranged at the bottom of the test box, which is used to acquire posture video data of the mouse in the test box during exercise; a first adjusting mirror; a second adjusting mirror is arranged on the side of the first adjusting mirror, and the second adjusting mirror and the first adjusting mirror are arranged at an included angle; and a third adjusting mirror is arranged on the other side of the first adjusting mirror, and the third adjusting mirror and the first adjusting mirror are arranged at an included angle; wherein the included angle between the second adjusting mirror or the third adjusting mirror and the first adjusting mirror can be adjusted.

4. The mouse swimming-based exercise state analysis method according to claim 1, characterized by, In the step of identifying the positions of each skeletal joint of the mice based on the posture video data, and proposing coordinate values of the corresponding skeletal joints at different time points: the positions of the skeletal joints include: nose tip, left ear, right ear, left eye, right eye, head center, neck, trunk center, tail root, middle tail, tail tip, iliac crest, hip, hind ankle, hind toe tip, shoulder, wrist, elbow, and front toe tip; and the coordinate system corresponding to the coordinate values of the skeletal joints is set by the resolution of the video acquisition device.

5. The mouse swimming-based exercise state analysis method according to claim 1, wherein Based on the mouse trunk skeleton, the skeletal joints are classified and mapped to corresponding trunk classification labels.

6. A system for implementing the mouse swimming-based exercise state analysis method according to any one of claims 1 to 5, characterized by, The system comprises an experiment module, a data acquisition module, a data transmission module, a video analysis module, a storage module, and an output module. The video analysis module is configured to analyze the swimming video data of the mouse, and comprises: acquiring posture video data of the mouse in a swimming state in a motion scene; based on the posture video data, identifying the positions of each skeletal joint of the mouse, and proposing coordinate values of the corresponding skeletal joint at different time points; based on the coordinate values, establishing a data collection of the coordinate values of each joint of the mouse; based on the data collection, identifying the duration and frequency of the behavior of the corresponding skeletal joint, comparing the differences between different behaviors, and generating motion change rule data of the swimming state of the mouse.

Images