An intelligent rehabilitation guidance system based on machine learning

Abstract

The application discloses an intelligent rehabilitation guidance system based on machine learning, and relates to the technical field of virtual rehabilitation guidance, which comprises the following steps: integrating multiple source sensors, collecting physiological and motion data of a user in real time, and constructing health data; fusing multi-modal data, adjusting resistance difficulty according to the condition of a patient through virtual reality, calculating a comprehensive health score and a motion deviation early warning index; setting a rehabilitation collaborative particle swarm algorithm, combining the comprehensive health score and the risk early warning index, combining the resistance difficulty and the muscle-joint collaborative coefficient to compensate for motion compensation, optimizing rehabilitation motion parameters, and returning to the previous layer; and analyzing abnormal roots based on a structural causal model to generate targeted rehabilitation suggestions. The application solves the problems that traditional rehabilitation training lacks dynamic regulation and control of training intensity and muscle compensation behavior of patients, thereby increasing the risk of secondary injury.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual rehabilitation guidance, and in particular to an intelligent rehabilitation guidance system based on machine learning. Background Art

[0002] Traditional rehabilitation training relies on the guidance of physical therapists, and rehabilitation plans are mainly based on experience, making it difficult to achieve personalized and precise adjustments. At the same time, patients may have problems with movement compensation, overtraining, or undertraining during the rehabilitation process, which affects the rehabilitation effect.

[0003] The intelligent rehabilitation guidance system based on machine learning can collect the patient's physiological data in real time and analyze the rehabilitation progress to improve rehabilitation efficiency and safety, but the traditional machine learning algorithm cannot dynamically coordinate the patient's rehabilitation degree and exercise risks; in addition, although the introduction of virtual reality technology can enhance the patient's sense of immersion and improve the compliance of rehabilitation training, traditional training uses fixed resistance, which can easily lead to insufficient training or excessive fatigue. During the patient's rehabilitation training, it is impossible to judge the standard of the patient's movements in real time. As the patient's rehabilitation process and physical recovery progress, the existing rehabilitation guidance program cannot adjust the rehabilitation difficulty according to the patient's real-time physical condition. In the rehabilitation process, patients often use compensatory movements due to insufficient muscle strength or pain, which will lead to joint wear or muscle atrophy in the long run. Traditional methods cannot monitor the safety of the patient's movements.

[0004] Therefore, the present invention proposes an intelligent rehabilitation guidance system based on machine learning. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent rehabilitation guidance system based on machine learning to solve the existing problems raised in the above background technology.

[0006] To achieve the above objectives, the present invention provides the following technical solutions: an intelligent rehabilitation guidance system based on machine learning, comprising:

[0007] Health data collection module, used to integrate multi-source sensors, collect user physiological and motion data in real time, and build health data;

[0008] The health status scoring module is used to integrate multimodal data, adjust the resistance difficulty according to the patient's condition through virtual reality, and calculate the comprehensive health score and movement deviation warning indicators;

[0009] The guidance parameter optimization module is used to set the rehabilitation collaborative particle swarm algorithm, combine the comprehensive health score and risk warning indicators, combine the resistance difficulty and muscle-joint synergy coefficient to compensate for the movement, optimize the rehabilitation movement parameters, and return them to the health status scoring module;

[0010] The cause-effect reasoning and guidance generation module analyzes the root cause of the anomaly based on a structural cause-effect model and generates targeted rehabilitation suggestions.

[0011] The health data acquisition module is further improved, wherein the health data acquisition module is integrated with a physiological sensor, a motion sensor, a bioelectricity sensor and a virtual reality interaction module, and a multi-modal spatio-temporal data set D={x t |x t =(h t ,m t ,e t ,v t ) is constructed, wherein a physiological data vector h t =(HR t ,BP t ,SpO2 t ) is constructed, HR t represents a heart rate, BP t represents a blood pressure, and SpO2 t represents a blood oxygen level; a motion data vector m t =(θ joint ,a acc ,f step ) is constructed, θ joint represents an angle of a joint, a acc represents accelerometer data, and f step represents a gait frequency; a bioelectricity vector e t =(EMG amp ,EMG frep ) is constructed, EMG amp represents an electromyography signal amplitude feature, and EMG frep represents an electromyography signal frequency feature; and v t represents a virtual reality feedback.

[0012] The health state scoring module further comprises a health score unit and a motion deviation early warning index unit, the health score unit extracts individual health state change features through a hierarchical physiological-motion attention network, processes time series data through an LSTM model to extract health state change features in a physiological data vector, calculates importance weights of different data modalities by using an attention mechanism, and obtains a comprehensive health score S h .

[0013] The motion deviation early warning index unit calculates a motion deviation early warning index by combining the physiological data vector and the motion data vector, and the calculation formula is as follows:

[0014]

[0015] wherein α1, α2 and α3 represent weight coefficients, sig( ) represents a Sigmoid function, Rk represents an action offset early warning index, represents a fluctuation threshold of the i-th data in the physiological data vector, represents a fluctuation threshold of the i-th data in the motion data vector, represents a physiological data vector time difference, represents a motion data vector time difference, F vir represents a difficulty resistance coefficient.

[0016] The application further improves that the difficulty resistance coefficient is based on the Unity engine to build a dynamic rehabilitation environment, and adjusts the difficulty resistance coefficient in the scene according to the patient's ability, sets the target action sequence represents a target action vector at time point t, extracts the patient t∈[1,2,...,T], extracts the patient actual action sequence represents a patient actual action vector at time point t, by calculating the time sequence similarity of the target action sequence and the patient actual action sequence, and combining the bioelectricity vector, the difficulty resistance coefficient is obtained Wherein, sim() represents similarity calculation, sim max represents a maximum similarity, represents a patient muscle strength coefficient.

[0017] The application further improves that the known parameter optimization module adopts a rehabilitation collaborative particle swarm algorithm to optimize the rehabilitation action, and the specific steps include:

[0018] Step one, define an action target, and encode the action target as (Am, fr, du, η), wherein Am represents amplitude, fr represents frequency, du represents joint angle, and η represents motion intensity;

[0019] Step two, randomly generate a set of action target combinations for each group of targets;

[0020] Step three, set an action target function;

[0021] Step four, dynamically balance the safety and training intensity of the target function according to the real-time comprehensive health score and the action offset early warning index;

[0022] Step five, repeat steps two to four, select the minimum region of the target function through iteration, and output the rehabilitation action.

[0023] The application further improves that the action target function is represented as:

[0024] SF=λ1S h +λ2sim(Tac [1,T] ,Aac [1,T])-λ3R k +λ4||mjs act -mjs opt ||2;

[0025] Among them, λ1, λ2, λ3, and λ4 represent the weights of rehabilitation needs, mjs act represents the muscle-joint synergy coefficient, mjs opt represents the ideal collaboration threshold.

[0026] A further improvement of the present invention is that the rehabilitation demand weight is dynamically coordinated by the real-time comprehensive health score and the action deviation warning index, wherein λ3=log(1+R k ).

[0027] The present invention is further improved in that the muscle-joint synergy coefficient mjs is obtained by combining the muscle activation intensity represented by the amplitude characteristics of the electromyographic signal and the Euclidean norm of the joint angle at each time node. act , the calculation formula is

[0028] A further improvement of the present invention is that the structural causal model in the causal reasoning and guidance generation module represents the causal relationship of multimodal spatiotemporal data sets using a directed graph, and provides rehabilitation guidance suggestions for muscle fatigue or abnormal heart rate through independence testing.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] 1. The present invention first calculates the motion deviation warning index by combining physiological data vectors with motion data vectors through virtual reality, and dynamically adjusts the warning threshold. This can effectively detect the patient's compensatory behavior and abnormal movement patterns, and judge the accuracy of the patient's rehabilitation movements in real time. When the patient's movement deviation is large due to fatigue or abnormal physiological state, the system can issue a real-time warning, thereby improving the safety and effectiveness of rehabilitation training.

[0031] 2. Based on real-time comprehensive health scores and movement deviation warning indicators, the rehabilitation collaborative particle swarm algorithm dynamically adjusts the resistance level to match the training intensity with the patient's muscle strength status. By adjusting the virtual rehabilitation environment, it ensures that the patient achieves the best rehabilitation effect while adapting to their own abilities.

[0032] 3. By calculating the muscle-joint synergy coefficient, the degree of matching between muscle activation and joint movement is analyzed in real time. When the synergy index is low, the system can dynamically adjust the training parameters, guide the patient to adopt the correct force mode, reduce compensatory movements, detect compensatory behaviors in real time, optimize the rehabilitation training effect, and improve the quality of rehabilitation. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a framework diagram of an intelligent rehabilitation guidance system based on machine learning in the present invention;

[0034] Figure 2 This is a flowchart of the rehabilitation collaborative particle swarm algorithm of the intelligent rehabilitation guidance system based on machine learning in the present invention. DETAILED DESCRIPTION

[0035] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0036] The term "and / or" simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can mean: A exists alone, A and B exist simultaneously, or B exists alone. Additionally, the character " / " generally indicates an "or" relationship between the related objects.

[0037] Example 1

[0038] Figure 1 The framework diagram of a machine learning-based intelligent rehabilitation guidance system disclosed in this embodiment is shown, including:

[0039] Health data collection module, used to integrate multi-source sensors, collect user physiological and motion data in real time, and build health data;

[0040] The health status scoring module is used to integrate multimodal data, adjust the resistance difficulty according to the patient's condition through virtual reality, and calculate the comprehensive health score and movement deviation warning indicators;

[0041] The guidance parameter optimization module is used to set the rehabilitation collaborative particle swarm algorithm, combine the comprehensive health score and risk warning indicators, combine the resistance difficulty and muscle-joint synergy coefficient to compensate for the movement, optimize the rehabilitation movement parameters, and return them to the health status scoring module;

[0042] The causal reasoning and guidance generation module analyzes the root causes of abnormalities based on structural causal models and generates targeted rehabilitation suggestions.

[0043] The health data acquisition module integrates physiological sensors, motion sensors, bioelectric sensors and virtual reality interaction modules to construct a multimodal spatiotemporal data set D = {x t |x t =(h t ,m t ,e t,v t )}, where the physiological data vector h t =(HR t ,BP t ,SpO2 t ), HR t Indicates heart rate, BP t Indicates blood pressure, SpO2 t Represents blood oxygen; motion data vector m t =(θ joint ,a acc ,f step ),θ joint represents the joint angle, a acc represents the accelerometer data, f step represents the gait frequency; the bioelectric vector e t =(EMG amp ,EMG frep ), EMG amp Represents the amplitude characteristics of electromyographic signals, EMG frep Represents the frequency characteristics of the electromyographic signal; v t Represents virtual reality feedback.

[0044] The health status scoring module includes a health scoring unit and a motion deviation warning indicator unit. The health scoring unit extracts individual health status change characteristics through a hierarchical physiological-motor attention network, processes time series data through an LSTM model, extracts health status change characteristics in physiological data vectors, and uses an attention mechanism to calculate the importance weights of different data modalities. After weighted calculation, a comprehensive health score S is obtained. h , the calculation formula is: Where L represents the data dimension;

[0045] The motion deviation warning indicator unit calculates the motion deviation warning indicator by combining the physiological data vector and the motion data vector. The calculation formula is: Among them, α1, α2 and α3 represent weight coefficients, sig( ) represents Sigmoid function, R k Indicates the action deviation warning indicator, represents the fluctuation threshold of the i-th item in the physiological data vector, represents the fluctuation threshold of the i-th data in the motion data vector, represents the temporal difference of physiological data vector, Represents the temporal difference of motion data vector, F vir Indicates the difficulty resistance coefficient.

[0046] When S h When it is low, it may indicate that the patient is fatigued, compensatory movements or physiological abnormalities. In this case, excessive simulated resistance may cause Rk Increase (such as too fast heart rate, muscle fatigue), so the resistance F needs to be adjusted dynamically vir , to balance training effect and safety;

[0047] In traditional training, fixed resistance can easily lead to excessive fatigue or insufficient training. vir , ensuring that the resistance matches the patient's muscle strength; the difficulty resistance coefficient is based on the Unity engine to build a dynamic rehabilitation environment, and the difficulty resistance coefficient in the scene is adjusted according to the patient's ability to set the target action sequence Represents the target action vector at time point t, extracts the patient t∈[1,2,...,T], and extracts the patient's actual action sequence The actual motion vector of the patient at time point t is represented. The difficulty resistance coefficient is obtained by calculating the time series similarity between the target motion sequence and the patient's actual motion sequence and combining it with the bioelectric vector. Among them, sim() represents similarity calculation, sim max Indicates the maximum similarity, Represents the patient's muscle strength coefficient.

[0048] The patient's muscle strength coefficient is set by setting the patient's muscle strength ms st Standard value and standard coefficient The muscle strength coefficient changes with the patient's actual muscle strength ms ac The improvement gradually increases,

[0049] The calculation formula of the function sim() is:

[0050]

[0051] Among them, π represents the most aligned path, Indicates the distance between the actual action and the ideal action in the feature space. The output of the function sim() indicates the degree of matching between the actual action and the ideal action. The smaller the value, the closer the match.

[0052] By calculating sim(), the system can quantify the accuracy of the patient's movements; for example, if sim(Tac [1,T] ,Aac [1,T] ) is large, indicating that the patient's movements deviate greatly from the standard template, and there may be compensatory behavior or movement errors; if sim(Tac [1,T] ,Aac [1,T] ) gradually decreases, indicating that the patient's movement quality is improving.

[0053] Example 2

[0054] Figure 2The present invention shows a flowchart of a rehabilitation collaborative particle swarm algorithm for an intelligent rehabilitation guidance system based on machine learning. Based on the inventive concept of Example 1, this embodiment proposes a rehabilitation collaborative particle swarm algorithm to optimize rehabilitation movements. The specific steps include:

[0055] Step 1: Define the action target and encode it as (Am, fr, du, η), including amplitude, frequency, joint angle and movement intensity;

[0056] Step 2: Randomly generate a set of action target combinations for each group of targets;

[0057] Step 3: Set the action objective function:

[0058] SF=λ1S h +λ2sim(Tac [1,T] ,Aac [1,T] )-λ3R k +λ4||mjs act -mjs opt ||2;

[0059] Among them, λ1, λ2, λ3, and λ4 represent the weights of rehabilitation needs, mjs act represents the muscle-joint synergy coefficient, mjs opt represents the ideal collaboration threshold.

[0060] Step 4: Dynamically balance the safety and training intensity of the objective function based on the real-time comprehensive health score and movement deviation warning indicators;

[0061] The rehabilitation demand weight is dynamically coordinated by the real-time comprehensive health score and the movement deviation warning indicator, among which λ3=log(1+R k ).

[0062] For the rehabilitation demand weight of the comprehensive health score, when S h Big, R k Hour, It will be close to 1, otherwise, it will tend to 0. h When the system detects the risk value R k As it increases, it will automatically reduce the h (positive return part) weight, and focus more on risk control; when the risk is very low, more emphasis is placed on improving or maintaining the health score; adaptive adjustment Through such a fractional structure, λ1 will adaptively adjust according to the current state S h With R k By adjusting the relative size of the health score, there is no need to manually define multiple interval rules, which can achieve the effect of lowering the priority of the health score when the risk is high and increasing the priority of the health score when the risk is low.

[0063] For the rehabilitation demand weight of the similarity between the actual action and the ideal target action, the “difference” is converted into “similarity” by the method of “1-relative difference”. The larger λ2 is, the closer the action is to the standard action.

[0064] For the rehabilitation demand weight of the movement deviation warning indicator, when R k = 0, λ3 = 0, indicating that there is no risk and no additional penalty is required. k When the value of λ3 increases, λ3 also increases monotonically. However, due to the "slow increase" characteristic of the logarithmic function, the weight will not grow too fast, which will cause the system to be too sensitive or unstable. k =λ3” form, which may cause extreme changes in the objective function when the risk is slightly higher.

[0065] Step 5: Repeat steps 2 to 4, select the minimum area of ​​the objective function through iteration, and output the rehabilitation action.

[0066] The muscle-joint synergy coefficient mjs is obtained by combining the muscle activation intensity represented by the amplitude characteristics of the electromyographic signal and the Euclidean norm of the joint angle at each time node. act , the calculation formula is

[0067] EMG amp (t) represents the myoelectric amplitude measured at the tth time point, that is, the degree of muscle activation. The ratio will fall in the interval [-1, 1] or [0, 1]. Usually, if the two signals are positively correlated and synchronized in their changing trends, the synergy index approaches 1; if the two change in opposite directions or are not synchronized, the index will become smaller, or even approach 0 and become negative. At this time, the larger the value, the higher the degree of synergy or matching between muscle activation and joint movement; the smaller the value, the weaker their coupling, which may mean that the muscle force and joint movement stages are not synchronized. The present invention can adjust the virtual resistance during the rehabilitation guidance process, detect compensatory behavior in real time, and guide patients to adopt the correct force mode.

[0068] Because patients find it difficult to understand abstract suggestions such as "why the range of motion needs to be adjusted," the structural causal model in the causal reasoning and guidance generation module represents the causal relationship between multimodal spatiotemporal datasets using a directed graph and provides rehabilitation guidance for muscle fatigue and abnormal heart rate through independence tests. Specifically, it includes:

[0069] Assuming that there is a potential relationship between all variables in the multimodal spatiotemporal dataset, we construct a completely undirected graph in which there is an edge between each pair of variables.

[0070] Test that under the condition that no other variables are included, that is, under the empty set condition, each pair of variables x in the graph κ and x τ Are they independent? If the statistical test shows that x κ and x τ If the conditions are independent, remove the edge between them; at this time, extract any data x that is still connected in the multimodal spatiotemporal dataset again σ and x γ , then extract any σ and x γ Other connected data to verify x σ and x γ Is it independent? If so, remove x σ and x γ While recording the edges between them, the conditional data that makes them independent are sent to separate data sets;

[0071] The size of the condition set is gradually increased in this way, starting with a single variable, until a preset upper limit is reached or all variable combinations have been tested.

[0072] In the undirected graph obtained in the first stage, if there are two variables x κ and x τ Both with variables Connected, but x κ and There is no direct edge between them, and if x τ Does not belong to x κ and Separate data sets, then their structure is determined to be Represents x τ Maybe x κ and Common consequences, x κ and There is no direct relationship;

[0073] By transferring rules, the undirected edges are further directed to obtain a directed acyclic graph and establish a causal model.

[0074] For example, if it is found that "heart rate" and "electromyographic signal amplitude characteristics" are correlated when other variables are not considered, but become independent when given "joint angle", it can be inferred that "joint angle" may be a mediating variable between the two.

[0075] Similarly, if the relationship between Acceleration and Heart Rate is strong without any other variables, but remains significant after adding variables other than Gait Frequency, this suggests that Acceleration may have a direct causal effect on Heart Rate.

[0076] Ultimately, through this systematic test, the automatically inferred causal graph will show the possible causal edges between the variables. This graph can be used to provide patients with visual and easy-to-understand suggested explanations, such as generating text or voice for patients to indicate that intervention should be targeted at the key node of "gait frequency" to improve problems such as "acceleration" and "heart rate".

[0077] The threshold and weight can be set by default according to the present invention, or can be set by the operator.

[0078] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0079] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0080] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0081] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0082] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which are all protected by the present invention.

Claims

1. An intelligent rehabilitation guidance system based on machine learning, characterized by: include: Health data collection module, used to integrate multi-source sensors, collect user physiological and motion data in real time, and build health data; The health status scoring module is used to integrate multimodal data, adjust the resistance difficulty according to the patient's condition through virtual reality, and calculate the comprehensive health score and movement deviation warning indicators; The guidance parameter optimization module is used to set the rehabilitation collaborative particle swarm algorithm, combine the comprehensive health score and risk warning indicators, combine the resistance difficulty and muscle-joint synergy coefficient to compensate for the movement, optimize the rehabilitation movement parameters, and return them to the health status scoring module; The causal reasoning and guidance generation module analyzes the root causes of abnormalities based on structural causal models and generates targeted rehabilitation recommendations; The health status scoring module includes a health scoring unit and a motion deviation warning indicator unit. The health scoring unit extracts individual health status change characteristics through a hierarchical physiological-motor attention network, processes time series data through an LSTM model, extracts health status change characteristics in physiological data vectors, and uses an attention mechanism to calculate the importance weights of different data modalities. After weighted calculation, a comprehensive health score S is obtained. h The motion deviation warning indicator unit calculates the motion deviation warning indicator by combining the physiological data vector and the motion data vector. The calculation formula is: Among them, α1, α2 and α3 represent weight coefficients, sig() represents Sigmoid function, R k Indicates the action deviation warning indicator, represents the fluctuation threshold of the i-th item in the physiological data vector, represents the fluctuation threshold of the i-th data in the motion data vector, represents the temporal difference of physiological data vector, represents the temporal difference of motion data vector, F vir Indicates the difficulty resistance coefficient.

2. The intelligent rehabilitation guidance system based on machine learning according to claim 1, characterized in that: The health data acquisition module integrates physiological sensors, motion sensors, bioelectric sensors and virtual reality interaction modules to construct a multimodal spatiotemporal data set D = {x t |x t =(h t ,m t ,e t ,v t )}, where the physiological data vector h t =(HR t ,BP t ,SpO2 t ), HR t Indicates heart rate, BP t Indicates blood pressure, SpO2 t Represents blood oxygen; motion data vector m t =(θ joint ,a acc ,f step ),θ joint represents the joint angle, a acc represents the accelerometer data, f step represents the gait frequency; the bioelectric vector e t =(EMG amp ,EMG frep ), EMG amp Represents the amplitude characteristics of electromyographic signals, EMG frep Represents the frequency characteristics of the electromyographic signal; v t Represents virtual reality feedback.

3. The intelligent rehabilitation guidance system based on machine learning according to claim 2, characterized in that: The difficulty resistance coefficient is based on the Unity engine to build a dynamic rehabilitation environment, and the difficulty resistance coefficient in the scene is adjusted according to the patient's ability to set the target action sequence. Represents the target action vector at time point t, extracts the patient t∈[1,2,...,T], and extracts the patient's actual action sequence The actual motion vector of the patient at time point t is represented. The difficulty resistance coefficient is obtained by calculating the time series similarity between the target motion sequence and the patient's actual motion sequence and combining it with the bioelectric vector. Among them, EMG amp Represents the amplitude characteristics of the electromyographic signal, sim() represents the similarity calculation, sim max Indicates the maximum similarity, Represents the patient's muscle strength coefficient.

4. The intelligent rehabilitation guidance system based on machine learning according to claim 1, characterized in that: The guidance parameter optimization module uses the rehabilitation collaborative particle swarm algorithm to optimize rehabilitation movements. The specific steps include: Step 1: Define the action target and encode it as (Am, fr, du, η), where Am represents amplitude, fr represents frequency, du represents joint angle, and η represents movement intensity. Step 2: Randomly generate a set of action target combinations for each group of targets; Step 3: Set the action objective function; Step 4: Dynamically balance the safety and training intensity of the objective function based on the real-time comprehensive health score and movement deviation warning indicators; Step 5: Repeat steps 2 to 4, select the minimum area of ​​the objective function through iteration, and output the rehabilitation action.

5. The intelligent rehabilitation guidance system based on machine learning according to claim 4, characterized in that: The action objective function is expressed as: SF=λ1S h +λ2sim(Tac [1,T] ,Aac [1,T] )-λ3R k +λ4||mjs act -mjs opt ||2; Among them, λ1, λ2, λ3, and λ4 represent the weights of rehabilitation needs, mjs act represents the muscle-joint synergy coefficient, mjs opt represents the ideal collaborative threshold, Tac [1,T] represents the target action sequence, Aac [1,T] represents the patient's actual action sequence, S h represents the comprehensive health score, R k Indicates the motion deviation warning indicator.

6. The intelligent rehabilitation guidance system based on machine learning according to claim 5, characterized in that: The rehabilitation demand weight is dynamically coordinated by the real-time comprehensive health score and the movement deviation warning index, where λ3=log(1+R k ).

7. The intelligent rehabilitation guidance system based on machine learning according to claim 5, characterized in that: The muscle-joint synergy coefficient mjs is obtained by combining the muscle activation intensity represented by the amplitude characteristics of the electromyographic signal and the Euclidean norm of the joint angle at each time node. act , the calculation formula is EMG amp Represents the amplitude characteristics of the electromyographic signal.

8. The intelligent rehabilitation guidance system based on machine learning according to claim 1, characterized in that: The structural causal model in the causal reasoning and guidance generation module represents the causal relationship between multimodal spatiotemporal data sets using a directed graph, and provides rehabilitation guidance suggestions for muscle fatigue or abnormal heart rate through independence testing.

Images