CN117612669A - Rehabilitation training safety assessment method based on wearable equipment - Google Patents
Rehabilitation training safety assessment method based on wearable equipment Download PDFInfo
- Publication number
- CN117612669A CN117612669A CN202311343794.7A CN202311343794A CN117612669A CN 117612669 A CN117612669 A CN 117612669A CN 202311343794 A CN202311343794 A CN 202311343794A CN 117612669 A CN117612669 A CN 117612669A
- Authority
- CN
- China
- Prior art keywords
- training
- user
- motion
- rehabilitation
- function evaluation
- 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
- 238000012549 training Methods 0.000 title claims abstract description 234
- 238000000034 method Methods 0.000 title claims abstract description 72
- 230000033001 locomotion Effects 0.000 claims abstract description 222
- 238000011156 evaluation Methods 0.000 claims abstract description 119
- 230000000694 effects Effects 0.000 claims abstract description 78
- 230000033764 rhythmic process Effects 0.000 claims abstract description 58
- 230000008569 process Effects 0.000 claims abstract description 39
- 238000001514 detection method Methods 0.000 claims abstract description 8
- 230000006870 function Effects 0.000 claims description 106
- 230000006378 damage Effects 0.000 claims description 30
- 230000009471 action Effects 0.000 claims description 13
- 230000002787 reinforcement Effects 0.000 claims description 13
- 238000005457 optimization Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 8
- 230000000739 chaotic effect Effects 0.000 claims description 4
- 230000007659 motor function Effects 0.000 claims description 4
- 238000010219 correlation analysis Methods 0.000 claims description 3
- 230000001133 acceleration Effects 0.000 description 31
- 231100000915 pathological change Toxicity 0.000 description 12
- 230000036285 pathological change Effects 0.000 description 12
- 238000013528 artificial neural network Methods 0.000 description 9
- 230000008859 change Effects 0.000 description 9
- 210000003127 knee Anatomy 0.000 description 9
- 210000000689 upper leg Anatomy 0.000 description 9
- 210000003423 ankle Anatomy 0.000 description 8
- 208000027418 Wounds and injury Diseases 0.000 description 7
- 230000001186 cumulative effect Effects 0.000 description 6
- 230000003183 myoelectrical effect Effects 0.000 description 6
- 238000012360 testing method Methods 0.000 description 6
- 210000002414 leg Anatomy 0.000 description 5
- 238000003759 clinical diagnosis Methods 0.000 description 4
- 239000000284 extract Substances 0.000 description 4
- 208000014674 injury Diseases 0.000 description 4
- 230000006855 networking Effects 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 210000003205 muscle Anatomy 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000013475 authorization Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 210000003414 extremity Anatomy 0.000 description 2
- 210000001624 hip Anatomy 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 230000004118 muscle contraction Effects 0.000 description 2
- 238000012015 optical character recognition Methods 0.000 description 2
- 235000021018 plums Nutrition 0.000 description 2
- 208000025978 Athletic injury Diseases 0.000 description 1
- 206010041738 Sports injury Diseases 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
- A61B5/1118—Determining activity level
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/389—Electromyography [EMG]
- A61B5/397—Analysis of electromyograms
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/48—Other medical applications
- A61B5/4836—Diagnosis combined with treatment in closed-loop systems or methods
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/6802—Sensor mounted on worn items
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
- G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/60—Protecting data
- G06F21/62—Protecting access to data via a platform, e.g. using keys or access control rules
- G06F21/6218—Protecting access to data via a platform, e.g. using keys or access control rules to a system of files or objects, e.g. local or distributed file system or database
- G06F21/6245—Protecting personal data, e.g. for financial or medical purposes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/092—Reinforcement learning
-
- 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
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/30—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to physical therapies or activities, e.g. physiotherapy, acupressure or exercising
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- General Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- Software Systems (AREA)
- Heart & Thoracic Surgery (AREA)
- Data Mining & Analysis (AREA)
- Surgery (AREA)
- Veterinary Medicine (AREA)
- Computer Security & Cryptography (AREA)
- General Engineering & Computer Science (AREA)
- Animal Behavior & Ethology (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Computing Systems (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Multimedia (AREA)
- Mathematical Physics (AREA)
- Bioethics (AREA)
- Physiology (AREA)
- Signal Processing (AREA)
- Dentistry (AREA)
- Human Computer Interaction (AREA)
- Physical Education & Sports Medicine (AREA)
- Psychiatry (AREA)
Abstract
The application provides a rehabilitation training safety assessment method based on wearable equipment, which comprises the following steps: constructing a user information system associated with the wearable equipment and identifying the user information; acquiring exercise function evaluation indexes and judging the optimal rehabilitation exercise effect of each injured part; acquiring a motion signal and a surface electromyographic signal of an injured part, and judging the association degree between the injured parts; judging the optimal training amplitude and duration in the process of mutual involvement of the parts; acquiring user body data and exercise function evaluation indexes, and generating and controlling training content and rhythm; acquiring user body data, and initializing training content and music playing content; real-time involvement detection is carried out according to the exercise function evaluation index and the rehabilitation effect; and automatically adjusting and generating play contents of rehabilitation training, including training contents and music play contents.
Description
Technical Field
The invention relates to the technical field of information, in particular to a rehabilitation training safety assessment method based on wearable equipment.
Background
In the case of falling or sports injury, people often do not have one part to be injured, and most of the situations exist in which different parts of one limb or a plurality of limb parts are injured, but most of the current rehabilitation training is aimed at training of one part, and the possible dragging between muscles is not considered. In the rehabilitation training process, the damage to another area is more serious possibly caused by training on one part. The current rehabilitation equipment can detect the movement data generally, but how to recommend reasonable movement amount according to the movement data according to the damage condition, so that the mutual involvement of different parts in training is avoided, and the rehabilitation equipment is an important difficult problem, especially when two damaged parts are in different muscle nerves and have certain relevance, and is difficult to train simultaneously. The training rhythm of one part is clear, but the training rhythms of different injury degrees of a plurality of parts are difficult to control. Especially when training is started and when training is performed for half an hour, the training of one part can cause discomfort of the other part, and how to adjust the training rhythm is also an unresolved problem. Therefore, how to perform rehabilitation training with contradictory sites is currently a very important topic.
Disclosure of Invention
The invention provides a rehabilitation training safety assessment method based on wearable equipment, which mainly comprises the following steps:
the user information system construction and user information identification associated with the wearable equipment specifically comprises the following steps: user information acquisition of wearable equipment, and automatically identifying user identity by a system; the method comprises the steps of obtaining exercise function evaluation indexes and judging the optimal rehabilitation exercise effect of each injured part, wherein the obtaining of the exercise function evaluation indexes and the judging of the optimal rehabilitation exercise effect of each injured part specifically comprise the following steps: the exercise and physiological data acquisition module is used for evaluating and analyzing the module, establishing a rehabilitation exercise effect judgment model of different parts and judging the rehabilitation exercise effect of the injured part; acquiring a motion signal and a surface electromyographic signal of an injured part, and judging the association degree between the injured parts; judging the optimal training amplitude and duration in the process of mutual involvement of the parts; the method comprises the steps of obtaining user body data and exercise function evaluation indexes, generating and controlling training contents and rhythms, wherein the step of obtaining the user body data and the exercise function evaluation indexes, and generating and controlling the training contents and the rhythms specifically comprises the following steps: the training content generation module and the training rhythm generation module; acquiring user body data, and initializing training content and music playing content; real-time involvement detection is carried out according to the exercise function evaluation index and the rehabilitation effect; and automatically adjusting and generating play contents of rehabilitation training, including training contents and music play contents.
Further optionally, the user information system construction associated with the wearable device includes:
constructing a user information system associated with the wearable equipment, and acquiring face information, identity card information and body data of a user after acquiring the user permission; the identity card information comprises a user name, an age, an identity card number and a photo, and the body data comprises height, weight, injured parts and damage degree; the damage degree is divided into 0 to 3 periods, wherein the 0 period is undamaged, the 1 period is mild pathological changes, the 2 period is moderate pathological changes, the 3 period is severe pathological changes, and the damage degree is obtained after clinical diagnosis of doctors; the method comprises the steps of performing privacy protection on face information, identity card information of a user and body data in the rehabilitation training process by using chaotic encryption; the wearable equipment is provided with a user side system and an external camera, wherein the user side system comprises a user side capable of networking and a user information system, and when a user uses the equipment for the first time, the user side capable of networking firstly uploads face information, identity card information and body data of the user to the user information system; when the user uses the equipment again, the wearable equipment firstly automatically invokes the external camera to collect the face information of the user and then compares the face information with the face information in the user information system; comprising the following steps: collecting user information of the wearable equipment; the system automatically recognizes the identity of the user;
The user information acquisition of the wearable equipment specifically comprises the following steps:
the user inputs the face information, the identity card information and the body data of the user into a user information system of the wearable device. And the user information system calculates the similarity between the face information in the user information system and the photo in the user identity card information in the system, sets a first threshold value, successfully matches when the similarity exceeds the first threshold value, and then extracts the text information of the identity card by using optical character recognition. After applying for the authorization of the user identity information to the payment bank or the WeChat, calling a payment bank interface or a WeChat interface to check the identity card information of the user, and comparing the extracted identity card text information with the identity card information of a third party interface to confirm the authenticity of the user identity; if the verification is successful, the user enters a system homepage, and if the verification is not passed, the system shall re-ask the user to confirm, change and upload the information.
The system automatically identifies the identity of the user, and specifically comprises the following steps:
before the user uses, the user firstly selects whether rehabilitation training is needed, and when the user selects rehabilitation training, the user information system associated with the wearable equipment acquires the face information of the user through the external camera. And the user information system carries out face recognition on the user according to the obtained face information and the user information existing in the system, judges the identity of the user and acquires the historical training data of the user. The system identifies the user identity according to the information of the current user, acquires the historical training data of the user, and uploads the physical data of the user in real time in the rehabilitation training process.
Further optionally, the acquiring the exercise function evaluation index, and determining the optimal rehabilitation exercise effect of each injured part includes:
establishing a rehabilitation effect judging model based on a BP neural network, wherein the rehabilitation effect judging model comprises a movement and physiological data acquisition module and an evaluation analysis module; the exercise and physiological data acquisition module acquires exercise signals and surface electromyographic signals of various injured parts of a user in rehabilitation training in real time through wearable equipment, and inputs the exercise signals and the surface electromyographic signals into the evaluation analysis module; the evaluation analysis module recognizes the motion function evaluation index of each injured part of the user in rehabilitation exercise according to the motion signal and the surface electromyographic signal transmitted by the motion and physiological data acquisition module; the exercise function evaluation indexes comprise exercise intensity, exercise amplitude and exercise quantity; acquiring a movement function evaluation index of a user in rehabilitation movement, inputting a rehabilitation effect judging model, and automatically identifying the optimal rehabilitation movement effect of each injured part; comprising the following steps: a movement and physiological data acquisition module; an evaluation analysis module; establishing a rehabilitation exercise effect discrimination model of different parts; judging the rehabilitation exercise effect of the injured part;
The motion and physiological data acquisition module specifically comprises:
the motion and physiological data acquisition module comprises an inertial sensor and a myoelectric sensor on the wearable device. The inertial sensor collects motion signals including acceleration signals and angular velocity signals of a user's motion part. After the data acquisition is completed, the motion and physiological data acquisition module transmits the motion signal and the surface electromyographic signal to the evaluation analysis module.
The evaluation analysis module specifically comprises:
and calculating the exercise function evaluation index of the user. The exercise function evaluation indexes comprise exercise intensity, exercise amplitude and exercise quantity. The motion intensity is obtained through amplitude information of the surface electromyographic signals, and the evaluation analysis module extracts the average amplitude of the surface electromyographic signals up to the current moment as a motion intensity index. The motion amplitude is obtained through the acceleration and the motion angle of the motion part of the user, the evaluation analysis module respectively calculates the acceleration root mean square value and the motion angle at the current moment according to the motion acceleration and the angular velocity at each moment, and the acceleration root mean square value and the motion angle are used as motion amplitude indexes. And the motion quantity is used for extracting the median frequency of the surface electromyographic signals to the current moment by the evaluation analysis module as a motion quantity index.
The building of the rehabilitation exercise effect discrimination model of different parts specifically comprises the following steps:
a prior knowledge base of rehabilitation exercise effects, which contains a great deal of user body data, exercise function evaluation indexes and mapping relations for judging the scores of the rehabilitation effects by doctors, is established in advance. The body data includes height, weight, wound site, and degree of damage. The damage degree is divided into 0 to 3 stages, wherein the 0 stage is undamaged, the 1 stage is mild pathological changes, the 2 stage is moderate pathological changes, the 3 stage is severe pathological changes, and the damage degree is obtained after clinical diagnosis of doctors. The priori knowledge base is obtained through rehabilitation exercise experiments and comprises data of normal healthy users and users to be rehabilitated. Building a rehabilitation effect recognition model based on a BP neural network, and using a priori knowledge base according to 8:2 are divided into training and testing sets. And inputting the training set into the BP neural network for training, and then inputting the testing set into the BP neural network after preliminary training, and outputting the loss function value and the predicted value corresponding to the testing set. And repeatedly and iteratively training to update model parameters, reducing a loss function, and finishing model training when the loss function value is not changed.
Judging the rehabilitation exercise effect of the injured part, comprising the following steps:
After the user takes the wearable equipment, the motion signal and the surface electromyographic signal of each injured part of the user in rehabilitation training are real-time through the motion and physiological data acquisition module, and then the motion intensity, the motion amplitude and the motion quantity are calculated in the evaluation analysis module; and inputting the physical data, the injured part, the injury degree, the exercise intensity, the exercise amplitude and the exercise quantity of the user into a rehabilitation effect identification model to obtain an exercise rehabilitation effect score of the user.
Further optionally, the acquiring the motion signal and the surface electromyographic signal of the injured part, and determining the association degree between injured parts includes:
calculating and extracting the association degree between injured parts based on the correlation analysis of the exercise function evaluation indexes; the method comprises the steps that through an inertial sensor of a wearable device and an electromyographic sensor electrode patch, motion signals and surface electromyographic signals of injured parts of a user in rehabilitation training are collected in real time, the motion signals and the surface electromyographic signals are input into an evaluation analysis module, and the evaluation analysis module can calculate and obtain motion function evaluation indexes including motion intensity, motion amplitude and motion quantity; calculating the Pearson coefficient among the exercise function evaluation indexes of each injured part of the user in rehabilitation exercise to obtain a Pearson matrix, wherein each value in the Pearson matrix can measure the association degree of different injured parts, namely judging whether the exercise of one part leads to mutual involvement; if the involvement exists, judging the magnitude of negative influence caused by the involvement according to the score given by the model.
Further optionally, the determining the optimal training amplitude and duration in the process of inter-connecting the parts includes:
constructing a rehabilitation effect optimization algorithm in the process of mutual involvement of parts based on a Markov model; firstly, establishing a rehabilitation optimization Markov process: abstracting the rehabilitation optimization agent into a learner, wherein the learner is in a state of a motion signal and a surface electromyographic signal of the injured part and the involved part at different moments; then, learning the optimal training amplitude and duration of the injured part in the traction process: modeling the environment, modeling the same or similar situation as the environment, and obtaining a Bellman equation in a recursion form based on reinforcement learning of intrusion detection on the basis of the environment, wherein the equation comprises two functions: "State value function" and "State-action value function", respectively, represent cumulative rewards on the designated "state" and on the designated "state-action", the cumulative rewards function using gamma discount cumulative rewards; finally, solving an optimal solution for a value function of an optimal motion strategy in the part involvement process: determining a four-element group of a Markov decision process of an optimal motion strategy in the part dragging process, accumulating rewards parameter gamma, setting a convergence threshold value, and then obtaining the optimal motion strategy in the part dragging process, namely the optimal training amplitude and duration based on a gamma discount accumulated rewards value iterative algorithm.
Further optionally, the step of obtaining the user body data and the sports function evaluation index, and the step of generating and controlling the training content and the rhythm include:
the generation control of the training content and the rhythm is divided into a training content generation module and a training rhythm generation module; the training content generation module is used for constructing a training content generation method based on the deep reinforcement learning model, inputting the obtained user body data and the exercise function evaluation index into the training content generation module, and automatically generating rehabilitation training content; the training rhythm generation module analyzes the user movement rhythm according to the movement amplitude in the movement function evaluation index, selects the music matched with the user movement state through the music matching unit and the music adjusting unit and adjusts the music rhythm according to the user state in real time; comprising the following steps: a training content generation module; a training rhythm generation module;
the training content generation module specifically comprises:
and constructing a training content generation method based on the deep reinforcement learning model. The training content comprises training actions, training frequency and training modes, wherein the training modes comprise active and passive. A training content knowledge base based on user body data, exercise function evaluation indexes, rehabilitation effects judged by doctors, training contents formulated by the doctors and corresponding Q values is established in advance, and the user body data, the exercise function evaluation indexes and the training contents are taken as priori knowledge. And the Q value is a quantized value corresponding to the quality of the training content. Firstly, constructing a deep-reinforcement Learning model by adopting a deep Q-Learning algorithm: taking physical data and exercise function evaluation indexes of a user as states, taking training content as actions, taking the rehabilitation effect after rehabilitation training according to the generated training content as rewards, training a deep reinforcement learning model, and introducing a training content knowledge base to accelerate training in the model training process. The exercise function evaluation indexes comprise exercise amplitude, exercise intensity and exercise quantity. In the rehabilitation training process of the user, the acceleration and the angular velocity acquired by the wearable equipment are utilized to calculate and obtain two motion amplitude indexes of an acceleration root mean square value and a motion angle, and the surface electromyographic signals are utilized to respectively obtain a motion intensity index and a motion quantity index. And then, inputting the obtained user body data, the exercise function evaluation index and the obtained user body data into a pre-established deep reinforcement learning model, outputting the Q values of different training contents by the model, combining the content items with the highest Q values, and automatically generating rehabilitation training contents.
The training rhythm generation module specifically comprises:
the movement rhythm generation module comprises a movement rhythm analysis unit, a music matching unit and a music adjusting unit. The motion rhythm analysis unit firstly acquires motion amplitude of a user in the rehabilitation training process, wherein the motion amplitude comprises an acceleration signal and an angular velocity signal, then calculates median frequencies of the acceleration signal and the angular velocity signal, which are cut off at the current moment, respectively recorded as A and B, and takes the average value of the A and the B as the motion rhythm of the user. The music matching unit searches the music with highest matching degree with the motion rhythm of the user in the local music library or the online music library. The music adjusting unit adjusts the music in real time according to the motion rhythm of the user, and performs speed change and tone change processing on the music.
Further optionally, the acquiring the user body data, and initializing training content and music playing content includes:
acquiring user body data including height, weight, injured part and damage degree; presetting a one-minute simple action library as initialization data; counting the number of injured parts, if the number of injured parts is one, matching the same actions as the injured parts from a one-minute simple action library, guiding the user to move, evaluating and analyzing the movement signals and the surface electromyographic signals acquired in the minute to obtain movement function evaluation indexes, inputting the movement function evaluation indexes into a training content generation module and a training rhythm generation module respectively, and outputting training content and music playing content; if the number of the injured parts is two or more, the user selects the parts to be trained, the user is guided to exercise, meanwhile, the exercise signals and the surface electromyographic signals of all the injured parts are obtained, the association degree between the injured parts is obtained, and the optimal training amplitude and duration are judged in the process of the mutual involvement of the parts; analyzing the surface electromyographic signals to obtain a motion intensity index and a motion quantity index, and replacing the motion amplitude index with the optimal training amplitude to obtain a new motion function evaluation index; and respectively inputting the modified sports function evaluation indexes into a training content generation module and a training rhythm generation module, and outputting training content and music playing content.
Further optionally, the real-time involvement detection according to the exercise function evaluation index and the rehabilitation effect includes:
the method comprises the steps that motion signals and electromyographic signals of injured parts of a user in rehabilitation training are collected in real time through wearable equipment, and motion function evaluation indexes are calculated; outputting the rehabilitation effect of each part of the user based on the rehabilitation effect judging model; calculating and extracting the involvement degree of each injured part in real time based on the correlation of the exercise function evaluation indexes, presetting a second threshold value, and judging the rehabilitation effect of each part at the moment and the rehabilitation effect of the previous moment when the involvement degree exceeds the second threshold value; if the rehabilitation effect of each part is larger than or equal to the rehabilitation effect of the previous moment, no operation is performed; otherwise, the wearable device will issue a warning to the user and feed back to the server.
Further optionally, the automatically adjusting and generating the playing content of the rehabilitation training, including the training content and the music playing content, includes:
the server acquires a motion signal and a myoelectric signal corresponding to the early warning feedback generation moment, and calculates a motion function evaluation index comprising motion intensity, motion amplitude and motion quantity; extracting training content corresponding to early warning feedback generation time, adding the training content into a knowledge base of a depth enhancement model, expanding the knowledge base, and then repeating training the model by using a new knowledge base; based on the motion function evaluation index corresponding to the early warning feedback generation moment, outputting the optimal training amplitude and duration; the motion amplitude is replaced by the optimal training amplitude, the training content generation module inputs user body data and updated motion function evaluation indexes, and updated training content is output; and analyzing the movement rhythm of the user according to the optimal training amplitude in the movement function evaluation index, selecting music matched with the movement state of the user through a training rhythm generation module, and adjusting the music rhythm according to the user state in real time.
The technical scheme provided by the embodiment of the invention can have the following beneficial effects:
the invention can detect the associated damage part and automatically control the training content during rehabilitation training, generally reduces the involvement actions, reduces the intensity of the involvement training and slows down the rhythm. And automatically generates and changes the content of the rehabilitation training play and adjusts the music play rhythm. The method can well decompose multi-part rehabilitation training, and can still maintain two rehabilitation training when the multi-part is damaged. Avoiding the involvement and damage of other parts due to rehabilitation training of one part or the overtraining of another involved part caused by the normal training of one part.
Drawings
Fig. 1 is a flowchart of a rehabilitation training safety evaluation method based on a wearable device.
Fig. 2 is a schematic diagram of a rehabilitation training safety assessment method based on a wearable device.
Fig. 3 is a further schematic diagram of a rehabilitation training safety assessment method based on a wearable device according to the present invention.
Detailed Description
For a further understanding of the present invention, the present invention will be described in detail with reference to the drawings and examples. The present application is described in further detail below with reference to the drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be noted that, for convenience of description, only the portions related to the invention are shown in the drawings.
The rehabilitation training safety assessment method based on the wearable device in the embodiment specifically comprises the following steps:
step 101, a user information system associated with a wearable device constructs a user information identification.
And constructing a user information system associated with the wearable equipment, and acquiring face information, identity card information and body data of the user after acquiring the user permission. The identity card information comprises a user name, an age, an identity card number and a photo, and the body data comprises a height, a weight, an injured part and a damage degree. The damage degree is divided into 0 to 3 stages, wherein the 0 stage is undamaged, the 1 stage is mild pathological changes, the 2 stage is moderate pathological changes, the 3 stage is severe pathological changes, and the damage degree is obtained after clinical diagnosis of doctors. And carrying out privacy protection on the face information, the identity card information and the body data in the rehabilitation training process by using chaotic encryption. The wearable device is provided with a user side system and an external camera, wherein the user side system comprises a user side capable of networking and a user information system, and when a user uses the device for the first time, the user side capable of networking uploads face information, identity card information and body data of the user to the user information system. When the user uses the device again, the wearable device automatically invokes the external camera to collect face information of the user, and then compares the face information with the face information in the user information system. For example, the user Li San purchases the wearable device, and needs to obtain various identity information of the face and the credentials, and various body data of the height, the weight, the injured part and the damage degree; the wearable equipment transmits the data to the server through the Internet to carry out chaotic encryption so as to carry out privacy protection on face information of the user and physical and rehabilitation data of the user in the middle of rehabilitation training.
User information collection of a wearable device.
The user inputs the face information, the identity card information and the body data of the user into a user information system of the wearable device. And the user information system calculates the similarity between the face information in the user information system and the photo in the user identity card information in the system, sets a first threshold value, successfully matches when the similarity exceeds the first threshold value, and then extracts the text information of the identity card by using optical character recognition. After applying for the authorization of the user identity information to the payment bank or the WeChat, calling a payment bank interface or a WeChat interface to check the identity card information of the user, and comparing the extracted identity card text information with the identity card information of a third party interface to confirm the authenticity of the user identity; if the verification is successful, the user enters a system homepage, and if the verification is not passed, the system shall re-ask the user to confirm, change and upload the information. For example, when the user uses the wearable device for the first time, face information, identity card information and body data of the user are collected first. Firstly, the identity card information needs to be uploaded, and then the face information of the third person is collected through the wearable equipment. And the user information system calculates the similarity of the acquired face information of Zhang three and the photos in the identity card information, and when the similarity exceeds a first threshold value, the matching is successful, so that the user is ensured to be the same person. Wherein the first threshold is set to 90% because the paper considers 90% similarity to be the same person: zhang Cuiping, su Guangda. Overview of face recognition techniques. Chinese graphic school newspaper, 2000 (11): 7-16. And then extracting text information of the identity card, and checking the extracted text information of the identity card and the identity information of the third party interface to ensure that the identity card of the user is effective. And finally, checking the success of the third party, entering a system homepage, and if the success of the third party does not pass, carrying out information confirmation, modification and uploading again.
The system automatically recognizes the identity of the user.
Before the user uses, the user firstly selects whether rehabilitation training is needed, and when the user selects rehabilitation training, the user information system associated with the wearable equipment acquires the face information of the user through the external camera. And the user information system carries out face recognition on the user according to the obtained face information and the user information existing in the system, judges the identity of the user and acquires the historical training data of the user. The system identifies the user identity according to the information of the current user, acquires the historical training data of the user, and uploads the physical data of the user in real time in the rehabilitation training process. For example: when a certain registered user intend to be rehabilitation training, before wearing the wearable device, the user and other unregistered users are assumed to stand together, an information system associated with the device automatically carries out face recognition on the user, the user is identified to be the registered user and the unknown user, the unknown user can select to register or ignore, if the unknown user selects to register, face information and height weight information are acquired, if the unknown user selects to abandon the registration, training data of the registered user are finally returned to be used for providing intelligent rehabilitation training contents for the user in real time.
Step 103, acquiring exercise function evaluation indexes, and judging the optimal rehabilitation exercise effect of each injured part.
And establishing a rehabilitation effect judging model based on the BP neural network, wherein the rehabilitation effect judging model comprises a movement and physiological data acquisition module and an evaluation analysis module. The exercise and physiological data acquisition module acquires exercise signals and surface electromyographic signals of various injured parts of a user in rehabilitation training in real time through wearable equipment, and inputs the exercise signals and the surface electromyographic signals into the evaluation analysis module. The evaluation analysis module recognizes the motion function evaluation index of each injured part of the user in rehabilitation exercise according to the motion signals and the surface electromyographic signals transmitted by the motion and physiological data acquisition module. The exercise function evaluation indexes comprise exercise intensity, exercise amplitude and exercise quantity. The exercise function evaluation index of the user in rehabilitation exercise is obtained, a rehabilitation effect judgment model is input, and the optimal rehabilitation exercise effect of each injured part is automatically identified. If the optimal rehabilitation exercise effect of different injured parts of a user in rehabilitation is to be identified, firstly, an exercise signal and a surface electromyographic signal of each injured part are identified and acquired, after the exercise signal and the surface electromyographic signal are acquired for a period of time, the exercise signal and the surface electromyographic signal are respectively evaluated and analyzed, and then exercise function evaluation indexes of the different injured parts of the user are identified based on statistics of the exercise signal and the surface electromyographic signal, wherein the exercise function evaluation indexes comprise exercise intensity, exercise amplitude and exercise amount. Inputting the motor function evaluation index into a rehabilitation effect discrimination model based on the BP neural network, and outputting the optimal rehabilitation movement effect of each injured part.
And the exercise and physiological data acquisition module.
The motion and physiological data acquisition module comprises an inertial sensor and a myoelectric sensor on the wearable device. The inertial sensor collects motion signals including acceleration signals and angular velocity signals of a user's motion part. After the data acquisition is completed, the motion and physiological data acquisition module transmits the motion signal and the surface electromyographic signal to the evaluation analysis module. For example, the wearable device is worn on an arm. When a user starts rehabilitation exercise, an inertial sensor on the wearable device acquires an exercise signal, namely an arm exercise acceleration signal and an angular velocity signal of the user. Wherein the acceleration signal contains components in three directions of x, y and z axes. Meanwhile, the myoelectric sensor also starts to collect surface myoelectric signals of arm muscles in the movement process of the user, and the surface myoelectric signals quantify electric signals inducing muscle contraction, so that the degree of muscle contraction can be reflected. Wherein the acceleration signal, the angular velocity signal and the surface electromyographic signal are all time series signals.
And (5) evaluating an analysis module.
And calculating the exercise function evaluation index of the user. The exercise function evaluation indexes comprise exercise intensity, exercise amplitude and exercise quantity. The motion intensity is obtained through amplitude information of the surface electromyographic signals, and the evaluation analysis module extracts the average amplitude of the surface electromyographic signals up to the current moment as a motion intensity index. The motion amplitude is obtained through the acceleration and the motion angle of the motion part of the user, the evaluation analysis module respectively calculates the acceleration root mean square value and the motion angle at the current moment according to the motion acceleration and the angular velocity at each moment, and the acceleration root mean square value and the motion angle are used as motion amplitude indexes. And the motion quantity is used for extracting the median frequency of the surface electromyographic signals to the current moment by the evaluation analysis module as a motion quantity index. And the evaluation analysis module calculates and obtains the movement function evaluation index of the user according to the acceleration signal, the angular velocity signal and the surface electromyographic signal of the user movement part transmitted by the movement and physiological data acquisition module. For example, the evaluation analysis module receives the acceleration signal, the angular velocity signal and the surface electromyographic signal of the user movement part transmitted by the movement and physiological data acquisition module, and the time span is 1 minute. The evaluation analysis module spreads the analysis of the acceleration signal, the angular velocity signal and the surface electromyographic signal. The average amplitude of the surface electromyographic signals in one minute is used as a movement intensity index, the root mean square value of acceleration and the movement angle are used as movement amplitude indexes, and the median frequency of the surface electromyographic signals in one minute is used as a movement quantity index. The root mean square value of the acceleration is calculated as
，
Wherein x, y and z respectively refer to components of acceleration in three axes, and i is time. The motion angle calculation formula is
In the method, in the process of the invention,is the angle of the starting moment; />The angular velocity at time t.
And establishing a rehabilitation exercise effect discrimination model of different parts.
A prior knowledge base of rehabilitation exercise effects, which contains a great deal of user body data, exercise function evaluation indexes and mapping relations for judging the scores of the rehabilitation effects by doctors, is established in advance. The body data includes height, weight, wound site, and degree of damage. The damage degree is divided into 0 to 3 stages, wherein the 0 stage is undamaged, the 1 stage is mild pathological changes, the 2 stage is moderate pathological changes, the 3 stage is severe pathological changes, and the damage degree is obtained after clinical diagnosis of doctors. The priori knowledge base is obtained through rehabilitation exercise experiments and comprises data of normal healthy users and users to be rehabilitated. Building a rehabilitation effect recognition model based on a BP neural network, and using a priori knowledge base according to 8:2 are divided into training and testing sets. And inputting the training set into the BP neural network for training, and then inputting the testing set into the BP neural network after preliminary training, and outputting the loss function value and the predicted value corresponding to the testing set. The iterative training is continuously repeated to update the model parameters, the loss function is reduced, and when the loss function value is not changed any more, the model is trained And (3) forming the finished product. For example, the data format of the a priori knowledge base is: zhang san, height 167, weight 60k g Shank, degree of damage: stage 1, (exercise intensity: 1.5mv, (acceleration 1 m/s) ^ 2. Motion angle: 35 °), amount of movement: 250 HZ), rehabilitation effect: 90 minutes. The prior knowledge base comprises motion data and body data of each part in health and motion data and body data of each part in damage, and the smaller the value of a loss function is, the closer the predicted value of the model is to a true value, so that model parameters can be optimized by using the function.
Judging the rehabilitation exercise effect of the injured part.
After the user takes the wearable equipment, the motion signal and the surface electromyographic signal of each injured part of the user in rehabilitation training are real-time through the motion and physiological data acquisition module, and then the motion intensity, the motion amplitude and the motion quantity are calculated in the evaluation analysis module; and inputting the physical data, the injured part, the injury degree, the exercise intensity, the exercise amplitude and the exercise quantity of the user into a rehabilitation effect identification model to obtain an exercise rehabilitation effect score of the user. For example, when a user is injured by stretching a three-knee joint, but the ankle is intact, after the user takes the wearable device, the movement signals and the surface electromyographic signals of the knee and the ankle are obtained in real time, the system inputs the movement signals and the surface electromyographic signals into an evaluation analysis module, and the signals are analyzed and counted to obtain the movement intensity, the movement amplitude and the movement quantity of the knee and the ankle respectively; the injury part, the damage degree, the exercise intensity, the exercise amplitude and the exercise amount are input into a rehabilitation effect identification model to respectively obtain scores of the knee and the ankle, wherein the knee is 30 minutes, the ankle is 100 minutes, the knee rehabilitation exercise effect is generally obtained, and the ankle is completely rehabilitated.
Step 104, acquiring a motion signal and a surface electromyographic signal of the injured parts, and judging the association degree between the injured parts.
Calculating and extracting the association degree between injured parts based on the correlation analysis of the exercise function evaluation indexes; the method comprises the steps that through an inertial sensor of a wearable device and an electromyographic sensor electrode patch, motion signals and surface electromyographic signals of injured parts of a user in rehabilitation training are collected in real time, the motion signals and the surface electromyographic signals are input into an evaluation analysis module, and the evaluation analysis module can calculate and obtain motion function evaluation indexes including motion intensity, motion amplitude and motion quantity; calculating the Pearson coefficient among the exercise function evaluation indexes of each injured part of the user in rehabilitation exercise to obtain a Pearson matrix, wherein each value in the Pearson matrix can measure the association degree of different injured parts, namely judging whether the exercise of one part leads to mutual involvement; if the involvement exists, judging the magnitude of negative influence caused by the involvement according to the score given by the model. For example, the user Li San is injured at both the knees and thighs; the wearable equipment respectively acquires the movement signals and the surface electromyographic signals of the knees and the thighs, and respectively calculates the movement function evaluation indexes of the two parts in the movement process. And calculating the Pearson coefficient of each injured part of the user in rehabilitation exercise based on the exercise function evaluation index to obtain a Pearson matrix, wherein each value in the Pearson matrix reflects the association degree of each injured part. Since the intensity of motion, the magnitude of motion and the amount of motion are both time series data, the pearson coefficient can measure how two consecutive signals change together over time, and the linear relationship between them is represented by the numbers-1 (negative correlation), 0 (uncorrelated), and 1 (full correlation).
Step 105, determining the optimal training amplitude and duration in the process of the mutual involvement of the parts.
Constructing a rehabilitation effect optimization algorithm in the process of mutual involvement of parts based on a Markov model; firstly, establishing a rehabilitation optimization Markov process: abstracting the rehabilitation optimization agent into a learner, wherein the learner is in a state of a motion signal and a surface electromyographic signal of the injured part and the involved part at different moments; then, learning the optimal training amplitude and duration of the injured part in the traction process: modeling the environment, modeling the same or similar situation as the environment, and obtaining a Bellman equation in a recursion form based on reinforcement learning of intrusion detection on the basis of the environment, wherein the equation comprises two functions: "State value function" and "State-action value function", respectively, represent cumulative rewards on the designated "state" and on the designated "state-action", the cumulative rewards function using gamma discount cumulative rewards; finally, solving an optimal solution for a value function of an optimal motion strategy in the part involvement process: determining a four-element group of a Markov decision process of an optimal motion strategy in the part dragging process, accumulating rewards parameter gamma, setting a convergence threshold value, and then obtaining the optimal motion strategy in the part dragging process, namely the optimal training amplitude and duration based on a gamma discount accumulated rewards value iterative algorithm. For example, the logical relationship of the electromyographic signals at different moments constitutes a Markov process, namely: the state at the next time is related to the state at the present time only, and is not related to the state at any previous time. The Markov model is a reinforcement learning model and comprises four elements: environment, learner, return function, policy. The environment is in the wearable equipment, the learner is a rehabilitation optimization agent, the electromyographic signal related information in the wearable equipment is analyzed, the return function is the effect score of rehabilitation training, namely the output of a rehabilitation effect judging model, and the strategy is the optimal training amplitude and training duration of rehabilitation optimization.
And 106, acquiring user body data and exercise function evaluation indexes, and generating and controlling training content and rhythm.
The generation control of the training content and the rhythm is divided into a training content generation module and a training rhythm generation module. The training content generation module constructs a training content generation method based on the deep reinforcement learning model, and inputs the obtained user body data and the exercise function evaluation index into the training content generation module to automatically generate rehabilitation training content. The training rhythm generation module analyzes the user movement rhythm according to the movement amplitude in the movement function evaluation index, selects the music matched with the user movement state through the music matching unit and the music adjusting unit and adjusts the music rhythm according to the user state in real time. For example, the training content can be adjusted in real time and adaptively according to the physical data and the exercise function evaluation index of the user in the rehabilitation exercise process through the training content generation module, so that the rehabilitation efficiency is improved. The training rhythm generation module can select effective excitation music according to the motion rhythm of the user, can adjust the music playing effect to be in time with the current motion state, assist the user in adjusting the motion state information to the optimal state, and promote the user experience.
And a training content generation module.
And constructing a training content generation method based on the deep reinforcement learning model. The training content comprises training actions, training frequency and training modes, wherein the training modes comprise active and passive. A training content knowledge base based on user body data, exercise function evaluation indexes, rehabilitation effects judged by doctors, training contents formulated by the doctors and corresponding Q values is established in advance, and the user body data, the exercise function evaluation indexes and the training contents are taken as priori knowledge. And the Q value is a quantized value corresponding to the quality of the training content. Firstly, constructing a deep-reinforcement Learning model by adopting a deep Q-Learning algorithm: taking physical data and exercise function evaluation indexes of a user as states, taking training content as actions, taking the rehabilitation effect after rehabilitation training according to the generated training content as rewards, training a deep reinforcement learning model, and introducing a training content knowledge base to accelerate training in the model training process. The exercise function evaluation indexes comprise exercise amplitude, exercise intensity and exercise quantity. In the rehabilitation training process of the user, the acceleration and the angular velocity acquired by the wearable equipment are utilized to calculate and obtain two motion amplitude indexes of an acceleration root mean square value and a motion angle, and the surface electromyographic signals are utilized to respectively obtain a motion intensity index and a motion quantity index. And then, inputting the obtained user body data, the exercise function evaluation index and the obtained user body data into a pre-established deep reinforcement learning model, outputting the Q values of different training contents by the model, combining the content items with the highest Q values, and automatically generating rehabilitation training contents. For example, zhang three, height 167, weight 60kg, lower leg, degree of damage: stage 1, exercise intensity: 1.5mv, acceleration 1m/s 2, motion angle: 35 °, amount of exercise: and (2) inputting 250HZ into a deep reinforcement learning model to obtain the Q values of three training contents of training action, training frequency and training mode, wherein the combination result with the highest Q value is leg lifting, 10 times and initiative.
And a training rhythm generation module.
The movement rhythm generation module comprises a movement rhythm analysis unit, a music matching unit and a music adjusting unit. The motion rhythm analysis unit firstly acquires motion amplitude of a user in the rehabilitation training process, wherein the motion amplitude comprises an acceleration signal and an angular velocity signal, then calculates median frequencies of the acceleration signal and the angular velocity signal, which are cut off at the current moment, respectively recorded as A and B, and takes the average value of the A and the B as the motion rhythm of the user. The music matching unit searches the music with highest matching degree with the motion rhythm of the user in the local music library or the online music library. The music adjusting unit adjusts the music in real time according to the motion rhythm of the user, and performs speed change and tone change processing on the music. For example, training the leg with three pieces, the amplitude of the leg's motion includes an acceleration signal and an angular velocity signal. Wherein the acceleration signal and the angular velocity signal are both time-series data. Firstly, a motion rhythm of Zhang III is obtained through a motion rhythm analysis unit, music with highest matching degree is searched by an input music matching unit, and if the rhythm changes in the motion process of a user, a music adjusting unit is used for carrying out speed change and tone change processing on the music.
Step 107, acquiring user body data, and initializing training content and music playing content.
User body data including height, weight, wound site, and degree of damage is obtained. A one-minute simple action library is preset as initialization data. And counting the number of the injured parts, if the number of the injured parts is one, matching the same actions as the injured parts from a one-minute simple action library, guiding the user to move, carrying out evaluation analysis on the movement signals and the surface electromyographic signals acquired in the minute to obtain movement function evaluation indexes, inputting the movement function evaluation indexes into a training content generation module and a training rhythm generation module respectively, and outputting training content and music playing content. If the number of the injured parts is two or more, the user selects the parts to be trained, the user is guided to exercise, meanwhile, the exercise signals and the surface electromyographic signals of all the injured parts are obtained, the association degree between the injured parts is obtained, and the optimal training amplitude and duration in the process of the mutual involvement of the parts are judged. Analyzing the surface electromyographic signals to obtain a motion intensity index and a motion quantity index, and replacing the motion amplitude index with the optimal training amplitude to obtain a new motion function evaluation index. And respectively inputting the modified sports function evaluation indexes into a training content generation module and a training rhythm generation module, and outputting training content and music playing content. For example, if three legs are injured, the user first guides the user to do one-minute simple leg movements, and initial training content and music playing content are generated according to the acquired sports function evaluation indexes. If the positions to be trained are thighs, the thighs and the waists of the plums are injured, the plums are guided to do one-minute thigh exercise, and the association degree between the injured positions is analyzed according to data collected by the thighs and the waists, so that the optimal training amplitude and duration are obtained. And then, according to the surface electromyographic signals of the thighs, obtaining a motion strength index and a motion quantity index, taking the optimal training amplitude as a motion amplitude index, inputting a training content generation module and a training rhythm generation module, and outputting training content and music playing content.
And step 108, carrying out real-time involvement detection according to the exercise function evaluation index and the rehabilitation effect.
The method comprises the steps that motion signals and electromyographic signals of injured parts of a user in rehabilitation training are collected in real time through wearable equipment, and motion function evaluation indexes are calculated; outputting the rehabilitation effect of each part of the user based on the rehabilitation effect judging model; and calculating and extracting the involvement degree of each injured part in real time based on the correlation of the exercise function evaluation indexes, presetting a second threshold, and judging the rehabilitation effect of each part at the moment and the rehabilitation effect of the previous moment when the involvement degree exceeds the second threshold. If the rehabilitation effect of each part is larger than or equal to the rehabilitation effect of the previous moment, no operation is performed. Otherwise, the wearable device will issue a warning to the user and feed back to the server. For example, zhang Sanpei wears wearable devices at the ankle, knee and thigh root, the wearable devices will collect personal information of Zhang three, upload the electromechanical signals of each part to a server; the server outputs the optimal rehabilitation effect of each part of the user based on the rehabilitation effect judging model, wherein the optimal rehabilitation effect is respectively ankle, knee and thigh root; and calculating the involvement degree of each injured part in real time, and when the involvement degree exceeds a second threshold value, indicating that the injured part is greatly affected at the moment, and further, comparing the rehabilitation effect. If the rehabilitation effect of each part is larger than or equal to the rehabilitation effect of the previous moment, the training content is indicated to be beneficial to rehabilitation of each part. Otherwise, the training content is indicated to damage another part, and the system sends out early warning feedback. The second threshold is determined by an average value of the elements of the pearson matrix.
Step 109, automatically adjusting and generating play content of rehabilitation training, including training content and music play content.
The server acquires a motion signal and an electromyographic signal corresponding to the early warning feedback generation moment, and calculates a motion function evaluation index comprising motion intensity, motion amplitude and motion quantity. And extracting training content corresponding to the early warning feedback generation moment, adding the training content into a knowledge base of the depth enhancement model, expanding the knowledge base, and then repeating the training of the model by using a new knowledge base. Based on the motion function evaluation index corresponding to the early warning feedback generation moment, outputting the optimal training amplitude and duration. And replacing the motion amplitude with the optimal training amplitude, inputting user body data and updated motion function evaluation indexes in the training content generation module, and outputting updated training content. And analyzing the movement rhythm of the user according to the optimal training amplitude in the movement function evaluation index, selecting music matched with the movement state of the user through a training rhythm generation module, and adjusting the music rhythm according to the user state in real time. For example, after receiving the early warning feedback of the user Li San, the server gives the best training amplitude and duration in real time based on the trained markov model; and reproducing training contents and adapted music according to the given training amplitude and the calculated movement intensity and movement amount.
It will be evident to those skilled in the art that the present application is not limited to the details of the foregoing illustrative embodiments, and that the present application may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Claims (9)
1. A wearable device-based rehabilitation training security assessment method, the method comprising:
constructing a user information system associated with the wearable equipment and identifying the user information; acquiring exercise function evaluation indexes and judging the optimal rehabilitation exercise effect of each injured part; acquiring a motion signal and a surface electromyographic signal of an injured part, and judging the association degree between the injured parts; judging the optimal training amplitude and duration in the process of mutual involvement of the parts; acquiring user body data and exercise function evaluation indexes, and generating and controlling training content and rhythm; acquiring user body data, and initializing training content and music playing content; real-time involvement detection is carried out according to the exercise function evaluation index and the rehabilitation effect; and automatically adjusting and generating play contents of rehabilitation training, including training contents and music play contents.
2. The method of claim 1, wherein the user information system associated with the wearable device builds a user information identification, comprising:
constructing a user information system associated with the wearable equipment; acquiring face information, identity card information and body data of a user; the identity card information and the body data specifically comprise the name, age, identity card number, photo, height, weight, injured part and the damage degree of the division of the injured part of the user; privacy protection using chaotic encryption; uploading user information to a user information system through equipment; the device automatically invokes the camera to collect information and compares the information with the system information.
3. The method of claim 1, wherein the acquiring the motor function evaluation index to determine the optimal rehabilitation exercise effect for each injured site comprises:
establishing a rehabilitation effect judging model, wherein the model consists of a motion and physiological data acquisition module and an evaluation analysis module; collecting a motion signal and a surface electromyographic signal of a user in real time; identifying a motion function evaluation index according to the acquired signals, wherein the motion function evaluation index specifically comprises motion intensity, motion amplitude and motion quantity; and identifying the optimal rehabilitation exercise effect through the model.
4. The method of claim 1, wherein the acquiring the motion signal and the surface electromyographic signal of the injured site, determining the degree of association between injured sites, comprises:
performing correlation analysis of the exercise function evaluation index; collecting a motion signal and a surface electromyographic signal in real time through an equipment sensor; calculating to obtain motion function evaluation indexes, wherein the motion function evaluation indexes comprise motion strength, motion amplitude and motion quantity; and calculating the Pearson coefficient to obtain a Pearson matrix, and judging the association degree of different injured parts.
5. The method of claim 1, wherein the determining the optimal training amplitude and duration during the region-to-region involvement comprises:
constructing a rehabilitation effect optimization algorithm based on a Markov model; abstracting the recovered agent into a learner, wherein the learner is in the state of motion signals and surface electromyographic signals of the injured part and the affected part; modeling the environment to simulate the same or similar conditions, and obtaining a Bellman equation, wherein the equation comprises two functions: "state value function" and "state-action value function"; and obtaining an optimal motion strategy in the part involvement process based on a value iteration algorithm.
6. The method of claim 1, wherein the acquiring the user body data, the motor function evaluation index, generating and controlling the training content and rhythm, comprises:
constructing a training content generation method based on the deep reinforcement learning model, and inputting user body data and exercise function evaluation indexes into a training content generation module; the training rhythm generation module analyzes the movement rhythm of the user according to the movement function evaluation index, and selects the music matched with the movement state of the user and adjusts the music rhythm through the music matching unit and the music adjusting unit.
7. The method of claim 1, wherein the acquiring user body data, initializing training content and music playing content comprises:
acquiring the height, weight, injured part and damage degree of a user; counting the number of injured parts, performing different action matching according to the number of the parts, guiding a user to move, and acquiring a movement signal and a surface electromyographic signal to obtain a movement function evaluation index; the training content generation module and the training rhythm generation module are input, and the training content and the music playing content are output.
8. The method of claim 1, wherein the real-time involvement detection based on the motor function evaluation index and the rehabilitation effect comprises:
Acquiring a motion signal and an electromyographic signal of an injured part of a user through wearable equipment and obtaining a motion function evaluation index; outputting rehabilitation effects of all parts of a user; calculating the traction degree of each injured part, and comparing the rehabilitation effect at the moment with the previous moment; if the rehabilitation effect of each part is lower than the previous moment, the wearable device gives out a warning and feeds back the warning to the server.
9. The method of claim 1, wherein the automatically adjusting and generating the play content of the rehabilitation training, including the training content and the music play content, includes:
acquiring a motion signal and an electromyographic signal through a server, calculating a motion function evaluation index, extracting training content, and adding the training content into a knowledge base to perform model training; outputting the optimal training amplitude and duration, and inputting data in a training content generating module to obtain updated training content; music adapted to the user's motion state is selected and the music tempo is adjusted in real time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311343794.7A CN117612669A (en) | 2023-10-17 | 2023-10-17 | Rehabilitation training safety assessment method based on wearable equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311343794.7A CN117612669A (en) | 2023-10-17 | 2023-10-17 | Rehabilitation training safety assessment method based on wearable equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117612669A true CN117612669A (en) | 2024-02-27 |
Family
ID=89954974
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311343794.7A Pending CN117612669A (en) | 2023-10-17 | 2023-10-17 | Rehabilitation training safety assessment method based on wearable equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117612669A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105446484A (en) * | 2015-11-19 | 2016-03-30 | 浙江大学 | Electromyographic signal gesture recognition method based on hidden markov model |
| CN108310632A (en) * | 2018-01-30 | 2018-07-24 | 深圳市前海未来无限投资管理有限公司 | A kind of muscle electric stimulation method, apparatus and wearable device |
| CN110473603A (en) * | 2019-07-09 | 2019-11-19 | 大连理工大学 | A kind of body-building householder method based on electromyography signal |
| CN111724879A (en) * | 2020-06-29 | 2020-09-29 | 中金育能教育科技集团有限公司 | Rehabilitation training evaluation processing method, device and equipment |
| CN111752394A (en) * | 2020-08-18 | 2020-10-09 | 长春大学 | Non-implanted wearable double-layer electrode tongue touch force feedback control system and device |
| CN111957024A (en) * | 2019-05-19 | 2020-11-20 | 郑州大学 | Wearable Taiji motion gait evaluation and training system based on cloud platform |
| CN113571153A (en) * | 2021-08-19 | 2021-10-29 | 中国科学院计算技术研究所 | Passive training sensing system for limb rehabilitation of cerebral palsy patient and client side thereof |
| CN116312947A (en) * | 2023-03-13 | 2023-06-23 | 北京航空航天大学 | Immersive ankle and foot rehabilitation training method based on upper limb movement signals and electronic equipment |
-
2023
- 2023-10-17 CN CN202311343794.7A patent/CN117612669A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105446484A (en) * | 2015-11-19 | 2016-03-30 | 浙江大学 | Electromyographic signal gesture recognition method based on hidden markov model |
| CN108310632A (en) * | 2018-01-30 | 2018-07-24 | 深圳市前海未来无限投资管理有限公司 | A kind of muscle electric stimulation method, apparatus and wearable device |
| CN111957024A (en) * | 2019-05-19 | 2020-11-20 | 郑州大学 | Wearable Taiji motion gait evaluation and training system based on cloud platform |
| CN110473603A (en) * | 2019-07-09 | 2019-11-19 | 大连理工大学 | A kind of body-building householder method based on electromyography signal |
| CN111724879A (en) * | 2020-06-29 | 2020-09-29 | 中金育能教育科技集团有限公司 | Rehabilitation training evaluation processing method, device and equipment |
| CN111752394A (en) * | 2020-08-18 | 2020-10-09 | 长春大学 | Non-implanted wearable double-layer electrode tongue touch force feedback control system and device |
| CN113571153A (en) * | 2021-08-19 | 2021-10-29 | 中国科学院计算技术研究所 | Passive training sensing system for limb rehabilitation of cerebral palsy patient and client side thereof |
| CN116312947A (en) * | 2023-03-13 | 2023-06-23 | 北京航空航天大学 | Immersive ankle and foot rehabilitation training method based on upper limb movement signals and electronic equipment |
Non-Patent Citations (3)
| Title |
|---|
| 叶雯;李智军;: "可穿戴便携式康复外骨骼机器人的开发", 集成技术, no. 04, 15 July 2013 (2013-07-15), pages 29 - 33 * |
| 谢地;陈卉芳;亓建洪;刘海斌;高锋;朱振华;杨文宁;: "膝前交叉韧带损伤后垂直起跳能力评价:运动力学指标的有效性", 中国组织工程研究, no. 51, 9 December 2016 (2016-12-09), pages 50 - 55 * |
| 陈玲玲;李珊珊;刘作军;张燕;: "基于表面肌电的下肢肌肉功能网络构建及其应用研究", 自动化学报, no. 03, 15 March 2017 (2017-03-15), pages 90 - 100 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Al-Angari et al. | Distance and mutual information methods for EMG feature and channel subset selection for classification of hand movements | |
| CN108244744B (en) | Motion state identification method, sole and shoe | |
| CN107007257B (en) | The automatic measure grading method and apparatus of the unnatural degree of face | |
| CN104573458A (en) | Identity recognition method, device and system based on electrocardiogram signals | |
| CN103049741A (en) | Foot-to-ground acting force-based gait feature extraction method and gait identification system | |
| CN113663312B (en) | Micro-inertia-based non-apparatus body-building action quality evaluation method | |
| Gao et al. | Human gait recognition based on multiple feature combination and parameter optimization algorithms | |
| Senanayake et al. | A knowledge-based intelligent framework for anterior cruciate ligament rehabilitation monitoring | |
| CN115631532A (en) | Training action classification and completion degree scoring method in rehabilitation training system | |
| CN111931568A (en) | Human body falling detection method and system based on enhanced learning | |
| Attenberger et al. | Modeling and visualization of classification-based control schemes for upper limb prostheses | |
| Karimi et al. | A novel prosthetic hand control approach based on genetic algorithm and wavelet transform features | |
| CN113974612B (en) | Automatic evaluation method and system for upper limb movement function of stroke patient | |
| Ling et al. | Lower limb exercise rehabilitation assessment based on artificial intelligence and medical big data | |
| CN110910984A (en) | System and method for processing state adjustment information of sportsman | |
| CN117612669A (en) | Rehabilitation training safety assessment method based on wearable equipment | |
| CN105354468A (en) | User identification method based on multi-axis force platform gait analysis | |
| Zhang et al. | Activity recognition of the torso based on surface electromyography for exoskeleton control | |
| Asemi et al. | Handwritten signatures verification based on arm and hand muscles synergy | |
| Majumder et al. | A novel wireless system to monitor gait using smartshoe-worn sensors | |
| CN115762708A (en) | Lower limb rehabilitation scheme design method and device using multi-source information and case reasoning | |
| Veer et al. | Analysis and recognition of operations using SEMG from upper arm muscles | |
| CN109325428A (en) | Mankind's activity gesture recognition method based on multi-level end-to-end neural network | |
| González-Villanueva et al. | A tool for linguistic assessment of rehabilitation exercises | |
| KR102630840B1 (en) | EMG signal-based recognition information extraction system and EMG signal-based recognition information extraction method using the same |
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 |