CN115631532A - Training action classification and completion degree scoring method in rehabilitation training system - Google Patents
Training action classification and completion degree scoring method in rehabilitation training system Download PDFInfo
- Publication number
- CN115631532A CN115631532A CN202211110704.5A CN202211110704A CN115631532A CN 115631532 A CN115631532 A CN 115631532A CN 202211110704 A CN202211110704 A CN 202211110704A CN 115631532 A CN115631532 A CN 115631532A
- Authority
- CN
- China
- Prior art keywords
- action
- training
- state
- joint
- rehabilitation
- 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
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/23—Recognition of whole body movements, e.g. for sport training
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0021—Tracking a path or terminating locations
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0062—Monitoring athletic performances, e.g. for determining the work of a user on an exercise apparatus, the completed jogging or cycling distance
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B71/00—Games or sports accessories not covered in groups A63B1/00 - A63B69/00
- A63B71/06—Indicating or scoring devices for games or players, or for other sports activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
Landscapes
- Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physical Education & Sports Medicine (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Psychiatry (AREA)
- Human Computer Interaction (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Social Psychology (AREA)
- Rehabilitation Tools (AREA)
Abstract
The invention discloses a training action classification and completion degree scoring method in a rehabilitation training system, which is applied to a rehabilitation training doctor-patient interaction system and comprises the following steps: step one, a patient guides a video to complete a rehabilitation training action according to a guide instruction and a standard training action preset in a system, and action data information is collected; classifying the training actions acquired in the step one according to a preset training action classification criterion, and adopting different completeness scoring methods for different classifications; and step three, extracting proper action characteristics according to the training actions collected in the step one, and calculating the completion degree score of the current training action by combining the completion degree scoring method corresponding to the step two. The invention can obtain a training score which is more in line with the rehabilitation training condition, and the score can help the rehabilitation doctor to more intuitively know the rehabilitation training condition of the patient; meanwhile, the patient can know whether the training action is standard or not, so that the patient can correct the training action in time.
Description
Technical Field
The invention relates to the field of intelligent medical treatment, in particular to a training action classification and completion degree scoring method in a rehabilitation training system.
Background
With the continuous aggravation of the aging phenomenon of the population, the prevalence rate of cardiovascular and cerebrovascular diseases in social groups is gradually increased, and the number of people who are accidentally disabled due to frequent traffic accidents is increased. In the treatment process of the paralyzed and disabled patients, the rehabilitation training of the patients is particularly important, and a large amount of medical resources need to be invested. In the face of the above situation, there is an urgent need for a system capable of automatically guiding a patient to complete rehabilitation training.
The traditional rehabilitation training plan comprises a set of systematic actions of arm joint extension, elbow joint rotation, limb stretching, horse step squatting, bow step leg pressing and the like, wherein the set of actions is formulated by a rehabilitation doctor and guides a patient to quantitatively complete the actions every day according to the plan; the standard degree of the rehabilitation training action needs to be judged by a professional rehabilitation doctor, and as the training actions are various and the judgment of the training action completion degree has strong professionality and subjectivity, a large number of professional medical workers are needed to carry out training guidance and training effect evaluation on the rehabilitation training patient; in the face of the shortage of medical care personnel, a method capable of automatically guiding rehabilitation training of patients and scoring rehabilitation training actions of the patients is urgently needed.
Because the rehabilitation training actions are various and the judgment standards of different rehabilitation training actions are different, and the scores of the rehabilitation training have strong subjectivity and professionality, a rehabilitation training expert with a large amount of experience can accurately give the training scores. The existing rehabilitation training action completeness grade calculating method does not consider the complex situation in practical application to carry out classification processing on rehabilitation training actions, and the same or similar method is adopted to calculate training grades for different types of actions, so that reasonable and accurate training action completeness grade cannot be obtained. The Chinese patent with publication number CN108346457A discloses a rehabilitation training evaluation method, device and system, wherein the scheme is to obtain the conformity between the current limb movement and the standard limb movement by comparing the limb movement of a patient with the preset standard limb movement, and generate the rehabilitation training score of the current limb according to the conformity. Although the scheme can process the rehabilitation training scores of patients, the scheme is only suitable for processing the score calculation problem of the rehabilitation training action of a single joint or limb, is not suitable for processing the score calculation problem of the rehabilitation training action of a plurality of joints or limb combined actions, and is not suitable for processing the score calculation problem of the rehabilitation training action of the continuous state training action. The Chinese patent with the publication number of CN113707268A discloses a rehabilitation training evaluation method and system, and the scheme obtains the standard degree of the training action of a patient by performing motion analysis on a video and comparing the training action of the patient with a preset motion track, so as to obtain the rehabilitation training score of the patient. Although the scheme can solve the problem of the evaluation calculation of the rehabilitation training scores of the single-joint training action and the multi-joint combined training action, the problem of the evaluation calculation of the rehabilitation training scores of the continuous state training action still cannot be solved. Therefore, a method for carrying out system analysis on the rehabilitation training action and reasonably and accurately evaluating the completion degree of the rehabilitation training action is still lacked at present.
Therefore, the computer technology is used for classifying the existing rehabilitation training actions, and the rehabilitation action completeness is scored by adopting a more targeted method for different action types.
Disclosure of Invention
The invention aims to solve the defects of the prior art and provides a training action classification and completion degree scoring method in a rehabilitation training system.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a training action classification and completion degree scoring method in a rehabilitation training system comprises the following steps:
the method comprises the following steps: acquiring action data information of a patient when the patient conducts rehabilitation training action according to a guide instruction and a standard training action guide video preset in the system;
step two: classifying the action data information collected in the step one according to a preset training action classification criterion, and selecting different completeness scoring methods according to different classifications;
step three: and (4) extracting action characteristics according to the action data information collected in the step one, and calculating the completion degree score of the current training action by combining the completion degree scoring method corresponding to the step two.
Preferably, the specific method for acquiring motion data in the first step is as follows: collecting rehabilitation training action videos of patients, and obtaining the coordinate information of the bone points of the patients in the training process by using a bone point recognition algorithm as action data.
Preferably, the preset training action classification criteria specifically include: the training device comprises a single-joint training action, a multi-joint combined training action and a continuous state training action, wherein the single-joint training action refers to a process action of moving from a starting position to an ending position by a single joint or limbs; the multi-joint combination training motion refers to a process motion of moving from a starting position to an ending position by combining a plurality of joints or limbs; the continuous state training motion refers to a state motion in which a plurality of joints or limbs of the body move to a certain state and are maintained for a certain period of time.
Preferably, the extracted action features include: angles between joints, normalized distances between bone points, completion time of training actions, duration of action state.
Preferably, the single joint training action is calculated and scored according to the action amplitude and time of the single joint in the process, and the specific scoring calculation method comprises the following steps: detecting the minimum value and the maximum value of the action amplitude of a single joint in the whole movement process, and obtaining the action amplitude of the patient training according to the difference value of the minimum value and the maximum value and combining the action completion time for grading, wherein the specific grading formula is as follows:
wherein F single Scoring of the single joint training movements; phi is the amplitude value of single joint or limb movement in the rehabilitation training process, phi s The standard value of the action amplitude of a single joint or limb in the rehabilitation training action process; t is the time for the patient to complete the single joint or limb training process, T s Standard time for completing the action of a single joint or limb training process; a. The 1 The weight value of the influence of the action amplitude on the rehabilitation training score is obtained; a. The 2 The weight value of the influence of the action completion time on the rehabilitation training score is used.
Preferably, the multi-joint cooperative coordination is adopted for the multi-joint combined training action, and the action amplitude of each joint is comprehensively considered when the comprehensive scoring calculation is performed according to the action amplitude of each joint, and the specific calculation mode is as follows: detecting the minimum value and the maximum value of the action amplitude of each joint in the whole movement process, obtaining the action amplitude of each joint in the patient training action according to the difference value of the minimum value and the maximum value, distributing different weights to the action amplitudes of each joint, and scoring by combining the action finishing time, wherein the specific scoring formula is as follows:
wherein F multi Scoring of the multi-joint combination training movements;
score the ith joint or limb movement during rehabilitation training movements individually, and 1<i<N, n is the number of joints or limb movements involved in the current multi-joint compound training movement; phi i The amplitude value of the ith joint or limb movement in the rehabilitation training movement process; phi si The standard amplitude value of the ith joint or limb movement in the rehabilitation training movement process; t is a unit of i The time for the patient to complete the ith joint or limb movement; t is si Standard time to complete the ith joint or limb movement;
the weight value of the influence of the ith joint or limb action amplitude on the rehabilitation training score is obtained;
the weight value of the impact of the ith joint or limb action completion time on the rehabilitation training score is obtained; b i And scoring the impact weight value of the ith joint or limb action on the overall score.
Preferably, the action for the persistent state should be qualified when reaching the standard action state and keeping for a certain time, and the specific calculation mode is as follows: detecting motion amplitude values of one or more joints in the whole motion process, and selecting a stable peak value as the motion state amplitude of the one or more joints in the patient training motion; and the value is scored according to the duration of the action state, and the specific scoring formula is as follows:
wherein F state A score for a persistence state training action; k is an action state matching coefficient of the persistent state training action; t is the action state duration of the duration state training action; t is a unit of s Training a standard action state duration of the action for the duration state; sim (Q, Q) s ) Training the similarity between the action state of the action and the standard action state for the duration state, and starting to accumulate the duration time of the action state when the similarity of the action state is greater than a similarity threshold value; TH is an action state similarity threshold of the continuous state training action; m is the number of action state features involved in the current persistence state training action judged by expert experience; a. The j Training the impact weight value of the jth action state feature in the action on the overall action state similarity for the persistence state, and 1<＝j<= m; q is a state feature vector composed of all j action state features involved in the persistent state training action, wherein delta j The j action state characteristic in the training action; q s A standard state feature vector consisting of all j standard action state features involved in the persistence state training action, wherein δ s j Is the jth standard motion state feature in the training motion.
The invention has the beneficial effects that:
the invention classifies different rehabilitation training actions in advance, scores different types of actions in a more targeted scoring mode, and meanwhile, scores can be calculated by selecting different action characteristics when calculating the scores, so that the scores obtained by calculation are more accurate and are more consistent with the judgment of medical care personnel in the subjective sense.
When the method is applied to calculating the rehabilitation training action score, a more accurate and more visual and reasonable score result can be obtained; the medical staff can more accurately analyze the rehabilitation training condition of the patient through the score; meanwhile, the patient can more intuitively know the completion condition of each rehabilitation training action of the patient instead of a general score, so that the patient can conveniently improve the rehabilitation training action of the patient.
Drawings
The features and advantages of the present invention will be more clearly understood by reference to the accompanying drawings, which are illustrative and not to be construed as limiting the invention in any way, and in which:
FIG. 1 is an overall system block diagram of a specific embodiment of the present invention;
FIG. 2 is a flow chart illustrating the use of a patient terminal in an embodiment of the present invention;
FIG. 3 is a schematic diagram of information of virtual human skeletal points according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating a process of right arm lift operation in accordance with an embodiment of the present invention;
FIG. 5 is a diagram illustrating the operation of the right leg during the step and leg pressing operation in accordance with the present invention;
FIG. 6 is a diagram illustrating the squatting motion of the horse in accordance with an embodiment of the present invention;
the reference numbers in the figures are: 1 denotes a nose skeleton point, 2 denotes a right shoulder skeleton point, 3 denotes a left shoulder skeleton point, 4 denotes a right elbow skeleton point, 5 denotes a left elbow skeleton point, 6 denotes a right wrist skeleton point, 7 denotes a left wrist skeleton point, 8 denotes a right hip skeleton point, 9 denotes a left hip skeleton point, 10 denotes a right knee skeleton point, 11 denotes a left knee skeleton point, 12 denotes a right ankle skeleton point, and 13 denotes a left ankle skeleton point.
Detailed Description
The following description and the drawings sufficiently illustrate specific embodiments of the invention to enable those skilled in the art to practice them.
It should be understood that the described embodiments are only some embodiments of the invention, and not all embodiments. Based on the embodiments of the present invention, those skilled in the art can obtain other embodiments without creative efforts, which belong to the protection scope of the present invention.
The following description refers to the accompanying drawings in which the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of systems and methods consistent with certain aspects of the invention, as detailed in the appended claims.
In the description of the present invention, it is to be understood that the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art. In addition, in the description of the present invention, "a plurality" means two or more unless otherwise specified. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship.
Referring to fig. 1, an interactive system for rehabilitation action training according to an embodiment of the present invention includes a training plan setting module, a training result query module, a training plan acquisition module, a training action data acquisition module, a training action classification and scoring module, a training result query and upload module, a medical staff terminal, a patient staff terminal, a system administrator terminal, and a server system,
the medical staff terminal is used by rehabilitation doctors and other staff, the rehabilitation doctors can select rehabilitation training actions and make rehabilitation training plans for patients at the terminal, meanwhile, the rehabilitation doctors can check the rehabilitation training result data of the patients at the terminal, the terminal comprises a training plan setting module and a training result query module,
the training plan setting module is used for making a proper rehabilitation training plan for the patient. The specific implementation mode is as follows: the rehabilitation doctor selects proper rehabilitation training actions according to the injury, illness, disability and the like of the patient, determines the daily completion times and the training time of the training actions according to the physical condition of the patient, and finally sends the whole training plan to the patient through the server system,
the training result query module is used for facilitating a rehabilitation doctor to check the training result of the rehabilitation action of the patient. The module can be used for a rehabilitation doctor to check the scores of different rehabilitation actions of a patient, provide a training suggestion or modify a training plan for the patient according to the training result of the patient,
the patient person terminal is used for rehabilitation training of patients or family members thereof, and the specific use flow is shown in fig. 2. The patient and the family members thereof can check the rehabilitation training plan of the patient at the terminal and complete the rehabilitation training action in the plan according to the guidance. The built-in grading system of terminal can grade according to the completion condition of rehabilitation training action to show rehabilitation training action score at patient's terminal for looking over, rehabilitation training action score can be preserved and upload to medical personnel's terminal simultaneously, in order to supply the training condition that rehabilitation doctor looked over the disease. The terminal comprises a training plan acquisition module, a training action classification and scoring module and a training result query module and a reporting module,
the training plan acquisition module is used for acquiring and checking a rehabilitation training plan formulated by a rehabilitation doctor. The rehabilitation training plan comprises a plurality of rehabilitation training actions, the patient can check the training actions in the rehabilitation training plan, the completion times of the actions, the completion standard of the actions and the training guidance of a rehabilitation doctor through the training plan acquisition module,
the training action data acquisition module is used for acquiring rehabilitation training action data information of the patient. In the embodiment, the action video of the patient is acquired through the camera device, and then the coordinate information of each bone point of the patient in the rehabilitation action process is acquired through the human bone point recognition algorithms such as BlazePose and the like. And obtaining rehabilitation training action data information of the patient by processing the coordinate information of the bone points.
It should be noted that, in the present embodiment, a blazepos human bone point recognition algorithm is used, which can recognize 33 bone points in total, but in practical applications, some useless coordinate information of bone points may be discarded according to circumstances, and the bone point information used in the present invention is shown in fig. 3. However, in practical applications, the human bone point recognition algorithm may also be openpos, and the human bone point recognition algorithm is not limited in this embodiment.
The training action classification scoring module is used for classifying and scoring the rehabilitation training action of the patient. Because rehabilitation training actions are various and complex, and a single scoring method cannot accurately and effectively evaluate the completion degree and accuracy of patient training actions, different rehabilitation training actions need to be pre-classified, and then different action categories need to be scored in different ways to obtain more accurate training completion degree scores.
The rehabilitation training actions can be roughly classified into three major categories, namely single-joint training actions, multi-joint combination training actions and continuous state training actions, and the accuracy and the completeness of the training actions can be effectively evaluated by adopting a targeted calculation method for different categories of training actions.
The single joint training action comprises actions such as flexion, extension, rotation and the like of trunk joints, the actions have no fixed initial state or fixed ending state, and the calculation and the scoring are only needed to be carried out by combining the action amplitude and the action finishing time of the single joint in the rehabilitation training action process.
The score calculation formula is as follows:
wherein F single Scoring for single joint training movements; phi is the amplitude value of single joint or limb movement in the rehabilitation training process, phi s The standard value of the action amplitude of a single joint or limb in the rehabilitation training action process; t is the time for the patient to complete the single joint or limb training process, T s Standard time for completing the action of a single joint or limb training process; a. The 1 The weight value of the influence of the action amplitude on the rehabilitation training score is obtained; a. The 2 The weight value of the influence of the action completion time on the rehabilitation training score; the judgment standard of the joint or limb action amplitude adopts different judgment characteristics according to different joint or limb actions.
The score calculation process is described in detail below by taking the lifting motion of the right arm joint as an example, the lifting motion process of the right arm joint is as shown in fig. 4, the right arm is gradually lifted in the whole process from (a) an initial state, through (b) a middle state motion to (c) a final state, and the right arm joint is trained.
The patient completes the lifting action of the right arm joint with the right side of the body facing the camera equipment according to the guidance of the patient personnel terminal, and the bone point coordinate data of the patient in the action process is acquired through the training action data acquisition module.
For the lifting action of the right arm joint, the coordinate information of three skeleton points needs to be collected, which are respectively as follows:
the abscissa of the right hip bone point and the ordinate of the right hip bone point are hereinafter abbreviated as RHx and RHy;
the abscissa and ordinate of the right shoulder skeleton point are abbreviated as RSx and RSy;
the abscissa of the right elbow skeleton point and the ordinate of the right elbow skeleton point are abbreviated as REx and REy below;
the calculation formula of the right arm joint action angle W is as follows:
calculating the angle of the right arm joint in the whole lifting action process of the right arm joint according to the formula, and processing to obtain the maximum value W of the angle of the right arm joint in the action process max And a minimum value W min The time T for finishing the action and the amplitude value W of the angle change of the right arm joint in the preset standard action process s And a standard action completion time T s The score is obtained by comparison and calculation, and the influence of the action amplitude on the rehabilitation training score can be obtained through a preset weight value A 1 To change; the influence of the action completion time on the rehabilitation training score can be realized through a preset weight value A 2 To change.
The specific score calculation formula is as follows:
the score calculation process of the right elbow joint rotation motion described above is only one embodiment of motion score calculation for a single joint training process. The action of the single-joint training process only comprises the rotation action of the elbow joint of the right hand, the influence factors of the scoring only comprise the angle of the joint, and the normalized distance of the bone joint can be used for scoring for some actions.
The multi-joint combined training action comprises actions such as bow step leg pressing, double-arm stretching and the like, the actions have no fixed initial state or fixed end state, and the multi-joint cooperation is needed in the rehabilitation training action process, so the action amplitude of each joint is comprehensively considered when the actions are calculated and evaluated.
The score calculation formula is as follows:
wherein F multi Scoring of the multi-joint combination training movements;
score the ith joint or limb movement during rehabilitation training movements individually, and 1<i<= n, n is the number of joints or limb movements involved in the current multi-joint combined training movement; phi i The amplitude value of the ith joint or limb movement in the rehabilitation training movement process; phi si The standard amplitude value of the ith joint or limb movement in the rehabilitation training movement process; t is i The time for the patient to complete the motion of the ith joint or limb; t is si Standard time to complete the ith joint or limb movement;
the weight value of the influence of the ith joint or limb action amplitude on the rehabilitation training score is obtained;
the weight value of the impact of the ith joint or limb action completion time on the rehabilitation training score is obtained; b is i The influence weight value of the ith joint or limb action score on the overall score is given; the judgment standard of the joint or limb movement amplitude adopts different judgment characteristics according to different joint or limb movements.
The scoring calculation process is elaborated in detail by taking the right leg bow step leg pressing movement as an example, the right leg bow step leg pressing movement completion scoring is judged according to expert experience requirements by using two joints or limb movement characteristics, namely the bending angle of the right knee joint and the normalized distance between the left ankle and the right ankle, so that the value of n in the formula is 2. The right leg arch step leg pressing action process is as shown in fig. 5, from (a) an initial state, through (b) an intermediate state, to (c) a final state, in the whole process, the right knee joint is pressed by both hands to continuously squat, meanwhile, the left foot and the right foot are separated by a certain distance, and the right knee joint and the hip joint are trained.
The patient completes the right leg bending and leg pressing action towards the camera equipment with the right side of the body according to the guidance of the patient person terminal, and the bone point coordinate data of the patient in the action process is obtained through the training action data obtaining module.
For the right leg bow step leg pressing action, the coordinate information of four skeleton points needs to be collected, which is as follows:
the abscissa and ordinate of the right hip bone point, hereinafter abbreviated as RHx and RHy;
the abscissa of the right knee skeleton point and the ordinate of the right knee skeleton point are abbreviated as RKx and RKy;
the abscissa of the right ankle bone point and the ordinate of the right ankle bone point, hereinafter referred to as RAx and RAy;
the abscissa of the left ankle bone point and the ordinate of the left ankle bone point are abbreviated as LAx and LAy;
the calculation formula of the action angle W of the knee joint of the right leg is as follows:
the calculation formula of the normalized distance D of the left ankle joint and the right ankle joint is as follows:
calculating the angle of the knee joint of the right leg in the whole bow step leg pressing action process according to the formula, and processing to obtain the maximum value W of the angle of the knee joint of the right leg in the action process max And a minimum value W min And time T for completing right knee joint action 1 Amplitude value W of the angle change of the knee joint of the right leg during the same predetermined standard movement s And the standard right knee joint action completion time T s1 Carrying out comparison calculation to obtain a score; the influence of the right knee joint angle on the rehabilitation training score can be realized through a preset weight value
The influence of the motion completion time of the right knee joint on the rehabilitation training score can be changed through a preset weight value
To be changed. Meanwhile, for the bow step leg pressing action, the normalized distance D of the left and right ankle joints when the training action is finished and the normalized distance D of the left and right ankle joints in the standard action s The rehabilitation training effect is also influenced; the influence of the normalized distance of the left ankle joint and the right ankle joint on the rehabilitation training score can be realized through a preset weight value
The influence of the motion completion time of the left foot and the right foot on the rehabilitation training score can be changed through the preset weight value
The influence of the score of the action completion degree of the right knee joint on the overall score can be changed through a preset weight value B 1 The influence of the distance score of the left foot and the right foot on the overall score can be changed through a preset weight value B 2 The specific score calculation formula is as follows:
the above-described score calculation process of the bow step leg pressing motion is only one embodiment of the motion score calculation of the multi-joint combination training process. The actions of the multi-joint combined training process only comprise the bow step leg pressing action, and the influence factors of the combined scoring also only comprise the angle of the joint and the normalized distance of the joint.
The continuous state training action comprises the actions of squatting by horse, lifting hands, standing and the like, and the patient with the action only needs to reach a certain action state and keep a certain time to be qualified.
The score calculation formula is as follows:
wherein F state A score for a persistence state training action; k is an action state matching coefficient of the persistent state training action; t is the action state duration of the duration state training action; t is s Training a standard action state duration of the action for the duration state; sim (Q, Q) s ) Training the similarity between the action state of the action and the standard action state for the duration state, and starting to accumulate the duration time of the action state when the similarity of the action state is greater than a similarity threshold value; TH is an action state similarity threshold of the continuous state training action; m is the number of action state features involved in the current persistence state training action judged by expert experience; a. The j Training the impact weight value of the jth action state feature in the action on the overall action state similarity for the persistence state, and 1<＝j<= m; q is a state feature vector consisting of all j action state features involved in the persistent state training action, wherein delta j The j action state characteristic in the training action; q s A standard state feature vector consisting of all j standard action state features involved in the persistence state training action, wherein δ s j As the jth standard action shape in the training actionState features; different continuous state training actions need to select different action state characteristics according to expert experience.
The scoring calculation process is elaborated in detail by taking the squatting movement of the horse as an example, the scoring of the completion degree of the squatting movement of the horse is judged by using three joint or limb movement characteristics according to the expert experience, namely the bending angle of the right knee joint, the bending angle of the left knee joint and the vertical angle of the upper limb, so that the value of m in the formula is 3. As shown in fig. 6, the process of the horse step squat movement is from (a) an initial state, through (b) an intermediate state to (c) a final state, the knees are continuously bent in the whole process, meanwhile, the upper half body is kept relatively vertical to the ground, and the knees are trained.
The patient uses the right side of the body to face the camera equipment according to the guidance of the patient person terminal, completes the squatting action of the horse, and obtains the bone point coordinate data of the patient in the action process through the training action data obtaining module. Aiming at the squatting action of the horse, the coordinate information of three skeleton points needs to be collected, which are respectively as follows:
the abscissa and ordinate of the right hip bone point, hereinafter abbreviated as RHx and RHy;
the abscissa of the right knee skeleton point and the ordinate of the right knee skeleton point are abbreviated as RKx and RKy;
the abscissa of the right ankle bone point and the ordinate of the right ankle bone point, hereinafter referred to as RAx and RAy;
the abscissa of the left hip bone point and the ordinate of the left hip bone point are abbreviated as LHx and LHy in the following;
the abscissa and ordinate of the left knee skeleton point are abbreviated as LKx and LKy;
the abscissa of the left ankle skeleton point and the ordinate of the left ankle skeleton point are abbreviated as LAx and LAy;
the abscissa of the right shoulder skeleton point and the ordinate of the right shoulder skeleton point are abbreviated as RSx and RSy below;
the squatting action of the horse step requires that the knee joint is bent to a certain angle, the upper limb is kept upright, the state is met, the squatting action lasts for a period of time, and the squatting action is judged to be full score without judging the middle process of the action.
Right leg knee joint action angle W 1 The calculation formula is as follows:
left leg knee joint action angle W 2 The calculation formula is as follows:
vertical angle W of upper limb 3 The calculation formula is as follows:
the specific score calculation formula is as follows:
wherein F 3 Scoring of squat movements of the horse; k is an action state matching coefficient of the squatting action of the horse step; t is the action state duration time of the squatting action of the horse step; t is s The standard action state duration time of the squatting action of the horse step; sim (Q, Q) s ) The similarity between the action state of the horse step squat action and the standard action state is obtained, and when the similarity of the action state is greater than a similarity threshold value, the action state duration time is accumulated; TH is an action state similarity threshold of the squatting action of the horse step; q is a state feature vector composed of the motion state features of the right knee joint, the left knee joint and the upper limb in the squatting motion of the horse step, wherein W 1 、W 2 、W 3 The characteristics of the action states of the right knee joint, the left knee joint and the upper limb in the squatting action of the horse step are respectively; q s A standard state feature vector composed of the motion state features of the right knee joint, the left knee joint and the upper limb in the squatting motion of the horse step, wherein W s1 、W s2 、W s3 Are respectively a horseStandard motion state characteristics of the right knee joint, the left knee joint and the upper limb in the step-and-squat motion; a. The 1 、A 2 、A 3 The weight values of the influence of the action states of the right knee joint, the left knee joint and the upper limb on the similarity of the overall action state in the squatting action of the horse step are respectively.
The above-mentioned score calculating process of the horse step squat action is only one embodiment of the score calculating of the state training action, and the state training action only comprises the horse step squat action.
The training result inquiry and report module is used for inquiring the score of rehabilitation training of a patient, and the patient can choose to finish rehabilitation training action again if the score is too low, so that the action is more standard and a better rehabilitation training effect is obtained; if the patient is satisfied with the score, the score of the rehabilitation training action and the rehabilitation training action data can be stored through the server system and uploaded to the medical staff terminal, so that a rehabilitation doctor can provide a rehabilitation training suggestion or modify a rehabilitation training plan.
The server system is used for completing the transmission and storage of the interactive data among the medical staff terminals, the patient staff terminals and the system management staff terminals.
The system administrator terminal is used by a rehabilitation training interaction system administrator, the administrator can complete the setting of a rehabilitation training plan by a rehabilitation doctor and the modification and deletion of rehabilitation training action scores uploaded by patient personnel through the terminal, and meanwhile, the administrator can realize the user registration and logout of the medical personnel terminal and the patient personnel terminal through the terminal.
The technical features of the above examples can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, however, as long as there is no contradiction between the combinations of the technical features, the scope of the present description should be considered as being described in the present specification.
Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations fall within the scope defined by the appended claims.
Claims (7)
1. A training action classification and completion degree scoring method in a rehabilitation training system is characterized by comprising the following steps:
the method comprises the following steps: acquiring action data information of a patient when the patient conducts rehabilitation training action according to a guide instruction and a standard training action guide video preset in the system;
step two: classifying the action data information collected in the step one according to a preset training action classification criterion, and selecting different completeness scoring methods according to different classifications;
step three: and (4) extracting action features according to the action data information collected in the first step, and calculating the completion degree score of the current training action by combining the completion degree scoring method corresponding to the second step.
2. The method for classifying training actions and scoring the completeness in a rehabilitation training system according to claim 1, wherein the step one for collecting the action data comprises the following specific steps: the rehabilitation training action video of the patient is collected, and the bone point identification algorithm is used for obtaining the bone point coordinate information of the patient in the training process and using the bone point coordinate information as action data.
3. The method for classifying training actions and scoring the completeness in a rehabilitation training system according to claim 1, wherein the preset training action classification criteria specifically include: the training device comprises a single-joint training action, a multi-joint combined training action and a continuous state training action, wherein the single-joint training action refers to a process action of moving from a starting position to an ending position by a single joint or limbs; the multi-joint combination training motion refers to a process motion of moving from a starting position to an ending position by combining a plurality of joints or limbs; the continuous state training motion refers to a state motion in which a plurality of joints or limbs of the body move to a certain state and are maintained for a certain period of time.
4. The method for training motion classification and completeness scoring in a rehabilitation training system according to claim 1, wherein the extracted motion features comprise: angles between joints, normalized distances between skeletal points, completion time of training actions, duration of action state.
5. The method for classifying and scoring the completion of a training session in a rehabilitation training system according to claim 3, wherein the specific scoring formula for the training session of a single joint is as follows:
wherein F single Scoring of the single joint training movements; phi is the amplitude value of single joint or limb movement in the rehabilitation training process, phi s The standard value of the action amplitude of a single joint or limb in the rehabilitation training action process; t is the time for the patient to finish the action of the single joint or limb training process, and Ts is the standard time for finishing the action of the single joint or limb training process; a. The 1 The weight value of the influence of the action amplitude on the rehabilitation training score; a. The 2 The weight value of the influence of the action completion time on the rehabilitation training score is used.
6. The method for classifying and scoring the completion of the training movements in the rehabilitation training system according to claim 3, wherein the specific scoring formula of the multi-joint combination training movement is as follows:
wherein F multi Scoring of the multi-joint combination training movements;
score for the ith joint or limb movement during rehabilitation training movements individually, and 1<i<N, n is that the current multi-joint combined training action involves a joint orThe number of limb movements; phi i The amplitude value of the ith joint or limb movement in the rehabilitation training movement process; phi si The standard amplitude value of the ith joint or limb movement in the rehabilitation training movement process; t is i The time for the patient to complete the motion of the ith joint or limb; t is si Standard time to complete the ith joint or limb movement;
the weight value of the influence of the ith joint or limb action amplitude on the rehabilitation training score is obtained;
the influence weight value of the ith joint or limb action completion time on the rehabilitation training score is obtained; b is i And scoring the impact weight value of the ith joint or limb action on the overall score.
7. The method for classifying and scoring the completion of a training session in a rehabilitation training system as claimed in claim 3, wherein the specific scoring formula for the duration training session is:
wherein F state A score for a persistence state training action; k is an action state matching coefficient of the persistent state training action; t is the action state duration of the duration state training action; t is s Training a standard action state duration of the action for the duration state; sim (Q, Q) s ) Training the similarity between the action state of the action and the standard action state for the duration state, and starting to accumulate the duration time of the action state when the similarity of the action state is greater than a similarity threshold value; TH is an action state similarity threshold of the continuous state training action; m is the number of action state features involved in the current persistence state training action judged by expert experience; a. The j Training actions for persistent statesThe influence weight values of the j action state characteristics on the overall action state similarity are 1<＝j<= m; q is a state feature vector consisting of all j action state features involved in the persistent state training action, wherein delta j The j action state characteristic in the training action; q s A standard state feature vector consisting of all j standard action state features involved in the persistence state training action, wherein delta sj Is the jth standard motion state feature in the training motion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211110704.5A CN115631532A (en) | 2022-09-13 | 2022-09-13 | Training action classification and completion degree scoring method in rehabilitation training system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211110704.5A CN115631532A (en) | 2022-09-13 | 2022-09-13 | Training action classification and completion degree scoring method in rehabilitation training system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115631532A true CN115631532A (en) | 2023-01-20 |
Family
ID=84902536
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211110704.5A Pending CN115631532A (en) | 2022-09-13 | 2022-09-13 | Training action classification and completion degree scoring method in rehabilitation training system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115631532A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117038006A (en) * | 2023-07-21 | 2023-11-10 | 筋斗云易行科技(西安)有限责任公司 | Method for performing rehabilitation training AI auxiliary diagnosis decision after upper and lower limb orthopedics operation |
| CN117078976A (en) * | 2023-10-16 | 2023-11-17 | 华南师范大学 | Action scoring method, action scoring device, computer equipment and storage medium |
| CN117357103A (en) * | 2023-12-07 | 2024-01-09 | 山东财经大学 | CV-based limb movement training guiding method and system |
-
2022
- 2022-09-13 CN CN202211110704.5A patent/CN115631532A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117038006A (en) * | 2023-07-21 | 2023-11-10 | 筋斗云易行科技(西安)有限责任公司 | Method for performing rehabilitation training AI auxiliary diagnosis decision after upper and lower limb orthopedics operation |
| CN117078976A (en) * | 2023-10-16 | 2023-11-17 | 华南师范大学 | Action scoring method, action scoring device, computer equipment and storage medium |
| CN117357103A (en) * | 2023-12-07 | 2024-01-09 | 山东财经大学 | CV-based limb movement training guiding method and system |
| CN117357103B (en) * | 2023-12-07 | 2024-03-19 | 山东财经大学 | CV-based limb movement training guiding method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115631532A (en) | Training action classification and completion degree scoring method in rehabilitation training system | |
| Wang et al. | Quantitative assessment of upper-limb motor function for post-stroke rehabilitation based on motor synergy analysis and multi-modality fusion | |
| CN114067358B (en) | Human body posture recognition method and system based on key point detection technology | |
| CN108983973B (en) | Control method of humanoid smart myoelectric artificial hand based on gesture recognition | |
| CN110047591B (en) | Method for evaluating posture of doctor in surgical operation process | |
| CN112057040B (en) | Upper limb movement function rehabilitation evaluation method | |
| CN103440498A (en) | Surface electromyogram signal identification method based on LDA algorithm | |
| CN113707268A (en) | Rehabilitation training evaluation method and system | |
| CN110991268B (en) | Depth image-based Parkinson hand motion quantization analysis method and system | |
| CN110755085B (en) | Motion function evaluation method and equipment based on joint mobility and motion coordination | |
| CN113974612A (en) | Automatic assessment method and system for upper limb movement function of stroke patient | |
| CN114373530A (en) | Limb rehabilitation training system and method | |
| CN111883229A (en) | Intelligent movement guidance method and system based on visual AI | |
| CN115497626A (en) | Body health assessment method based on joint point identification | |
| CN114099234A (en) | Intelligent rehabilitation robot data processing method and system for assisting rehabilitation training | |
| Senanayake et al. | A knowledge-based intelligent framework for anterior cruciate ligament rehabilitation monitoring | |
| CN109126045A (en) | intelligent motion analysis and training system | |
| CN114202772B (en) | Reference information generation system and method based on artificial intelligence and intelligent medical treatment | |
| CN114533042A (en) | Multi-source feature-based automatic evaluation method for physical dysfunction of lumbago patients | |
| CN108447562B (en) | User exercise capacity assessment method and system | |
| JP7012631B2 (en) | Object identification device, identification system and identification method | |
| CN110895964A (en) | Electroencephalogram-based rehabilitation training assessment method and system | |
| CN116510249A (en) | Hand virtual rehabilitation training system and training method based on electromyographic signals | |
| CN114240934B (en) | Image data analysis method and system based on acromegaly | |
| CN115067934A (en) | Hand motion function analysis system based on machine intelligence |
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 |