Ddyskinesia rehabilitation training method and system based on electroencephalogram recognition

Abstract

The invention relates to the technical field of medical rehabilitation, in particular to a dyskinesia rehabilitation training method and system based on electroencephalogram signal recognition, and the system comprises a rehabilitation training management center, a test feedback unit, a motor imagery evaluation unit, a rehabilitation division unit, a force application monitoring unit and a training management unit. According to the method, analysis is preliminarily carried out from the perspective of electroencephalogram signal recognition to improve the electroencephalogram signal recognition precision, electroencephalogram and action evaluation analysis is carried out in a virtual training mode so as to visually know whether each training action of a user is qualified or not, analysis is carried out from the two perspectives of training overall completion degree and single action completion degree, and the accuracy of training is improved. And on one hand, the functional electrical stimulation is dynamically adjusted according to the information feedback condition to improve the rehabilitation training effect, and on the other hand, the user is reminded to adjust the power generation decision according to the feedback information, so that the rehabilitation training period of the user is shortened.

Description

Technical Field

[0001] The present invention relates to the field of medical rehabilitation technology, and in particular to a movement disorder rehabilitation training method and system based on electroencephalogram signal recognition. Background Art

[0002] Limb movement disorder is a symptom of limited or uncontrollable body muscle movement due to neurological diseases or injuries. It affects the daily life of patients and brings a lot of trouble to their families. In traditional limb rehabilitation, patients passively accept mechanical training and lack active brain participation. The effect on motor nerve repair is limited and long-term training is required to achieve the ideal effect. With the continuous development of modern science and technology, brain wave monitoring and analysis technology has been widely used in rehabilitation training. Brain waves are a kind of bioelectric signal generated by neurons in the cerebral cortex and can reflect the functional state of the brain. By monitoring and analyzing brain waves, a scientific basis can be provided for rehabilitation training to help patients better restore their functions. However, in the existing technology, there are problems of low recognition speed and accuracy in the process of recognizing brain wave signals, which leads to a decrease in the recognition accuracy of brain wave signals. At the same time, it is impossible to carry out targeted training on short board training movements according to the completion of training movements. In addition, in the traditional muscle strength analysis process, there is a lack of monitoring of the user's force decision, and the user's force decision cannot be adjusted in time, resulting in a prolonged user rehabilitation training cycle. In view of the above technical defects, a solution is now proposed. Summary of the invention

[0003] The purpose of the present invention is to provide a movement disorder rehabilitation training method and system based on EEG signal recognition to solve the technical defects mentioned above. The present invention initially analyzes from the perspective of EEG signal recognition to improve the accuracy of EEG signal recognition, and performs EEG and movement dual evaluation analysis through virtual training to intuitively understand whether each training movement of the user is qualified. At the same time, it analyzes from the two perspectives of overall training completion and single movement completion to reasonably adjust the training intensity, and at the same time, carries out targeted movement disorder rehabilitation training plans for the user's shortcomings, which helps to assist the next round of rehabilitation training tasks based on this. It analyzes from the perspective of the user's muscle strength deviation. On the one hand, it dynamically adjusts the functional electrical stimulation according to the information feedback to improve the rehabilitation training effect. On the other hand, it reminds the user to adjust the force decision based on the feedback information, which helps to shorten the user's rehabilitation training cycle.

[0004] The object of the present invention can be achieved by the following technical solutions: A movement disorder rehabilitation training system based on EEG signal recognition includes a rehabilitation training management center, a test feedback unit, a motor imagery evaluation unit, a rehabilitation division unit, a force monitoring unit and a training management unit; The rehabilitation training management center is used to collect the recognition information of the EEG signal recognition model established during the user's most recent training, and send the recognition information to the test feedback unit for recognition test evaluation feedback analysis to obtain a usable signal or an adjustment signal; When a usable signal is generated, the motor imagery evaluation unit is used to perform EEG and action dual evaluation analysis on the collected imagined EEG signals, and compare and analyze the obtained imagined EEG signals and imagined action information with the actual EEG signals and standard action information to obtain qualified signals or unqualified signals; the rehabilitation classification unit is used to perform rehabilitation action classification management analysis on the collected classified action labels, and perform discrimination processing on the obtained local completion degree Jg to obtain normal training actions and short board training actions; The force monitoring unit is used to perform a muscle force deviation risk analysis on the collected current muscle force value and target muscle force value of the user, and to perform judgment processing on the maximum number of consecutive A's to obtain a normal signal or a warning signal.

[0005] Preferably, the recognition test evaluation feedback analysis process is as follows: After collecting the user's training period, retrieve the EEG signal recognition model established during the user's most recent training from the server; The training period is divided into a test period and a rehabilitation period, and test EEG signals corresponding to the user's running imagination of n known limb movements in the test period are obtained, where n is a natural number greater than 3. The n test EEG signals are set as test training data, and the test training data are recognized through an EEG signal recognition model. Based on the EEG signal recognition model, recognition information of each test EEG signal is obtained, and the recognition information includes recognition time and recognition result. The recognition result includes qualified control and unqualified control. The mean of the recognition time is obtained, and the mean of the recognition time is set as the test recognition score, and the number of test EEG signals with qualified control recognition results is set as the accurate recognition rate. The test recognition score and the accurate recognition rate are discriminated and processed to obtain a usable signal or an adjusted signal.

[0006] Preferably, the EEG and movement dual evaluation analysis process is as follows: Limb rehabilitation training is performed based on a virtual training scenario, and limb movements are set. At the same time, actual EEG signals corresponding to the set limb movements are obtained, and the actual EEG signals corresponding to the set limb movements are identified. At the same time, control instructions of the actual EEG signals are obtained, and standard movement information of the limb rehabilitation robot is obtained based on the control instructions of the actual EEG signals. The standard movement information includes the limb rotation angle and the limb maintenance time.

[0007] Preferably, the user is prompted to perform corresponding limb movement imagination based on the set limb movement during the rehabilitation period, and then the imagined EEG signal of the user performing the corresponding limb movement imagination is obtained based on the signal acquisition device, and the imagined EEG signal of the user performing the corresponding limb movement imagination is identified, and the control instruction corresponding to the imagined EEG signal of the limb movement imagination is obtained at the same time, and the imagined movement information of the limb rehabilitation robot is obtained based on the control instruction corresponding to the imagined EEG signal; The imagined EEG signal and imagined movement information are compared and analyzed with the actual EEG signal and standard movement information: if the imagined EEG signal is consistent with the actual EEG signal, and the imagined movement information corresponds one-to-one with the standard movement information, a qualified signal is generated; if the imagined EEG signal is inconsistent with the actual EEG signal, or the imagined movement information does not correspond one-to-one with the standard movement information, an unqualified signal is generated.

[0008] Preferably, the rehabilitation action division management and analysis process is as follows: The set limb movements are divided into g classified action labels, where g is a natural number greater than zero. The classified action labels include left hand movements, right hand movements, and left leg movements. The total number of each classified action label is obtained, and the number of qualified signals generated by each classified action label is also obtained. The ratio between the number of qualified signals generated by the classified action label and the total number of classified action labels is set as the local completion degree Jg, and then the average value of the local completion degree Jg is obtained. The average value of the local completion degree Jg is set as the rehabilitation assessment completion degree.

[0009] Preferably, the rehabilitation assessment completion degree is subjected to a discrimination process: if the rehabilitation assessment completion degree is greater than or equal to a preset rehabilitation assessment completion degree threshold, a stable signal is generated; if the rehabilitation assessment completion degree is less than the preset rehabilitation assessment completion degree threshold, a feedback signal is generated; When the feedback signal is generated, the local completion degree Jg is judged: If the local completion degree Jg is greater than or equal to the local completion degree threshold, the corresponding classified action is judged to be a normal training action, and a standard-reaching signal is generated; if the local completion degree Jg is less than the local completion degree threshold, the corresponding classified action is judged to be a shortboard training action, and a deviation signal is generated.

[0010] Preferably, the muscle strength deviation risk analysis process is as follows: Based on the current muscle strength value and target muscle strength value of the user in the rehabilitation period obtained under the set limb movement, the value obtained by subtracting the target muscle strength value from the current muscle strength value and dividing it by the target muscle strength value is set as the muscle strength deviation value, and the muscle strength deviation value is judged and processed. If the muscle strength deviation value is less than the minimum value in the preset muscle strength deviation value range, an enhancement instruction is generated; if the muscle strength deviation value falls within the preset muscle strength deviation value range, a regular instruction is generated; if the muscle strength deviation value is greater than the maximum value in the preset muscle strength deviation value range, a reduction instruction is generated.

[0011] Preferably, the enhancement instruction or the reduction instruction are both set to A, and the regular instruction is set to B. A character string of A and B is constructed based on the generation order of A and B, and the character string constructed of A and B is set as a muscle strength label string. The maximum number of consecutive occurrences of A is obtained from the muscle strength label string, and the maximum number of consecutive occurrences of A is judged: if the maximum number of consecutive occurrences of A is greater than or equal to a preset threshold, a warning signal is generated; if the maximum number of consecutive occurrences of A is less than the preset threshold, a normal signal is generated.

[0012] The beneficial effects of the present invention are as follows: (1) The present invention initially analyzes from the perspective of EEG signal recognition to improve the accuracy of EEG signal recognition, and intuitively understands the recognition rate and accuracy of the EEG signal recognition model through information visualization feedback, thereby ensuring the rate of EEG signal recognition of motor imagery, realizing high-speed and efficient movement disorder rehabilitation training, and performing EEG and movement dual-item evaluation and analysis through virtual training, so as to intuitively understand whether each training movement of the user is qualified; (2) The present invention conducts a preliminary analysis from the perspective of overall training completion to understand whether the user's overall training completion has met the standard, so as to reasonably adjust the training intensity, and further develop a targeted movement disorder rehabilitation training plan based on the user's shortcomings, which can then help assist the next round of rehabilitation training tasks; (3) The present invention analyzes the user's muscle strength deviation. On the one hand, it dynamically adjusts the functional electrical stimulation according to the information feedback to improve the rehabilitation training effect. On the other hand, it reminds the user to adjust the force decision based on the feedback information, which helps to shorten the user's rehabilitation training cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The present invention will be further described below in conjunction with the accompanying drawings; Figure 1 It is a flowchart of the system of the present invention; Figure 2 It is a reference diagram of the method of embodiment 3 of the present invention. DETAILED DESCRIPTION

[0014] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0015] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive of other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments; Embodiment 1: See also Figure 1 to Figure 2 As shown, the present invention is a movement disorder rehabilitation training system based on EEG signal recognition, including a rehabilitation training management center, a test feedback unit, a motor imagery evaluation unit, a rehabilitation division unit, a force monitoring unit and a training management unit, the rehabilitation training management center is connected to the test feedback unit and the motor imagery evaluation unit in a one-way communication, the test feedback unit is connected to the motor imagery evaluation unit, the rehabilitation division unit and the training management unit in a one-way communication, the motor imagery evaluation unit is connected to the training management unit in a one-way communication, the rehabilitation division unit is connected to the force monitoring unit and the training management unit in a one-way communication, and the force monitoring unit is connected to the training management unit in a one-way communication; The rehabilitation training management center is used to collect the recognition information of the EEG signal recognition model established during the user's most recent training, and send the recognition information to the test feedback unit for recognition test evaluation feedback analysis. The recognition rate and accuracy of the EEG signal recognition model can be intuitively understood through information visualization feedback, thereby ensuring the recognition rate of the EEG signal of motor imagery. The specific recognition test evaluation feedback analysis process is as follows: After collecting the user's training period, retrieve the EEG signal recognition model established during the user's most recent training from the server; The training period is divided into a test period and a rehabilitation period, and the test EEG signals corresponding to the running imagination of n known limb movements of the user in the test period are obtained, where n is a natural number greater than 3. The n test EEG signals are set as test training data, and the test training data is recognized by the EEG signal recognition model. Based on the EEG signal recognition model, the recognition information of each test EEG signal is obtained, and the recognition information includes recognition time and recognition result. The recognition result includes control qualified and control unqualified. The mean of the recognition time is obtained, and the mean of the recognition time is set as the test recognition score, and the number of test EEG signals with recognition results of qualified control is set as the accurate recognition rate, and the test recognition score and the accurate recognition rate are discriminated: If the test recognition score is less than a preset test recognition score threshold, or the accurate recognition rate is less than a preset accurate recognition rate threshold, an adjustment signal is generated; If the test recognition score is greater than or equal to the preset test recognition score threshold, and the accurate recognition rate is greater than or equal to the preset accurate recognition rate threshold, an available signal is generated, and the available signal or the adjustment signal is sent to the training management unit. After receiving the available signal or the adjustment signal, the training management unit immediately displays the preset warning text corresponding to the available signal or the adjustment signal, that is, the preset warning text corresponding to the available signal is displayed as: model accurate, and the preset warning text corresponding to the adjustment signal is displayed as: model optimization. The recognition rate and accuracy of the EEG signal recognition model can be intuitively understood through information visualization feedback, thereby ensuring the rate of EEG signal recognition of motor imagery, and realizing high-speed and efficient movement disorder rehabilitation training; Among them, qualified control and unqualified control: if the action instruction obtained by identifying the test EEG signal based on the EEG signal recognition model is consistent with the action instruction corresponding to the known limb movement, the control is determined to be qualified; if the action instruction obtained by identifying the test EEG signal based on the EEG signal recognition model is inconsistent with the action instruction corresponding to the known limb movement, the control is determined to be unqualified; When a usable signal is generated, the motor imagery evaluation unit is used to perform EEG and action dual evaluation analysis on the collected imagination EEG signals, so as to intuitively understand whether each training action of the user is qualified. The specific EEG and action dual evaluation analysis process is as follows: Use VR technology to simulate life scenes to build virtual training scenes, conduct limb rehabilitation training based on the virtual training scenes, set limb movements, obtain actual EEG signals corresponding to the set limb movements, identify the actual EEG signals corresponding to the set limb movements, and obtain control instructions for the actual EEG signals. Based on the control instructions for the actual EEG signals, obtain standard movement information of the limb rehabilitation robot, which includes limb rotation angle, limb maintenance time, etc. Obtaining prompts for the user to perform corresponding limb movement imagination based on the set limb movement during the rehabilitation period, and then obtaining the imagined EEG signal of the user performing the corresponding limb movement imagination based on the signal acquisition device, and identifying the imagined EEG signal of the user performing the corresponding limb movement imagination, and at the same time obtaining the control instruction corresponding to the imagined EEG signal of the limb movement imagination, and obtaining the imagined movement information of the limb rehabilitation robot based on the control instruction corresponding to the imagined EEG signal; Among them, signal acquisition equipment includes but is not limited to: flexible scalp electrodes (non-invasive) or implantable ECoG electrodes (invasive); Compare and analyze the imagined EEG signals and imagined movement information with the actual EEG signals and standard movement information: If the imagined EEG signal is consistent with the actual EEG signal, and the imagined action information corresponds one-to-one with the standard action information, a qualified signal is generated; If the imagined EEG signal is inconsistent with the actual EEG signal, or the imagined action information is not one-to-one corresponding to the standard action information, an unqualified signal is generated, and the qualified signal or the unqualified signal is sent to the training management unit. After receiving the qualified signal or the unqualified signal, the training management unit immediately displays the preset warning text corresponding to the qualified signal or the unqualified signal, that is, the preset warning text corresponding to the qualified signal is displayed as: training qualified, and the preset warning text corresponding to the unqualified signal is displayed as: training unqualified. Through the information visualization feedback method, it is intuitively understood whether each training action of the user is qualified, so as to provide data support for subsequent analysis; The present invention can not only help patients achieve active rehabilitation through brain-muscle-limb coordinated training, but also achieve active rehabilitation through mutual integration, allowing the patient's brain to actively participate in closed-loop rehabilitation. This rehabilitation training based on motor imagery can be used for most patients with nerve damage. Modern medical rehabilitation theory proves that this type of functional recovery training can promote the repair of the nervous system of paralyzed patients, and active participation of patients will achieve better rehabilitation effects than passive training.

[0016] Embodiment 2: When an available signal is generated, the rehabilitation division unit is used to perform rehabilitation action division management and analysis on the collected classified action labels, so as to carry out targeted movement disorder rehabilitation training plans based on the user's shortcomings, and then help assist the next round of rehabilitation training tasks based on this. The specific rehabilitation action division management and analysis process is as follows: The set limb movement is divided into g classified action labels, where g is a natural number greater than zero. The classified action labels include left hand movement, right hand movement, left leg movement, etc. The total number of each classified action label is obtained, and the number of qualified signals generated by each classified action label is obtained. The ratio between the number of qualified signals generated by the classified action label and the total number of classified action labels is set as the local completion degree Jg, and then the mean value of the local completion degree Jg is obtained. The mean value of the local completion degree Jg is set as the rehabilitation assessment completion degree, and the rehabilitation assessment completion degree is discriminated: If the completion degree of the rehabilitation assessment is greater than or equal to the preset completion degree threshold of the rehabilitation assessment, a stable signal is generated and sent to the training management unit. After receiving the stable signal, the training management unit immediately displays the preset warning text corresponding to the stable signal, that is, the preset warning text corresponding to the stable signal is displayed as: the training expectation has been met, which helps to understand whether the overall completion degree of the user's training has met the standard, so as to reasonably adjust the training intensity; If the rehabilitation assessment completion degree is less than a preset rehabilitation assessment completion degree threshold, a feedback signal is generated; When the feedback signal is generated, the local completion degree Jg is judged: If the local completion degree Jg is greater than or equal to the local completion degree threshold, the corresponding classified action is determined to be a normal training action, and a standard-reaching signal is generated at the same time; If the local completion degree Jg is less than the local completion degree threshold, the corresponding classified action is determined to be a shortboard training action, and a deviation signal is generated at the same time, and the standard-reaching signal or the deviation signal is sent to the training management unit. After receiving the standard-reaching signal or the deviation signal, the training management unit immediately displays the preset warning text corresponding to the standard-reaching signal or the deviation signal, that is, the preset warning text corresponding to the standard-reaching signal is displayed as: the classified action label corresponding to the normal training action, and the preset warning text corresponding to the deviation signal is displayed as: the classified action label corresponding to the shortboard training action, so as to carry out a targeted movement disorder rehabilitation training plan for the user's shortcomings, which is helpful to assist the next round of rehabilitation training tasks based on this; When a stable signal is generated, the force monitoring unit is used to perform a force deviation risk analysis on the current and target force values ​​of the user so as to dynamically adjust the functional electrical stimulation according to the information feedback to improve the rehabilitation training effect. The specific force deviation risk analysis process is as follows: Based on the current muscle strength value and target muscle strength value of the user in the rehabilitation period obtained under the set limb movement, the value obtained by subtracting the target muscle strength value from the current muscle strength value and dividing it by the target muscle strength value is set as the muscle strength deviation value, and the muscle strength deviation value is judged and processed. If the muscle strength deviation value is less than the minimum value in the preset muscle strength deviation value range, an enhancement instruction is generated; if the muscle strength deviation value belongs to the preset muscle strength deviation value range, a conventional instruction is generated; if the muscle strength deviation value is greater than the maximum value in the preset muscle strength deviation value range, a reduction instruction is generated, that is, the training management unit dynamically adjusts the functional electrical stimulation according to the feedback enhancement instruction, conventional instruction and reduction instruction to improve the rehabilitation training effect; Set the enhancement command or the reduction command to A, set the general command to B, construct the strings of A and B based on the generation order of A and B, and set the constructed strings of A and B as the muscle strength label string, obtain the maximum number of consecutive A occurrences from the muscle strength label string, and perform discrimination processing on the maximum number of consecutive A occurrences: If the maximum number of consecutive A occurrences is greater than or equal to the preset threshold, a warning signal is generated; If the maximum number of consecutive A's is less than the preset threshold, a normal signal is generated, and the normal signal or warning signal is sent to the training management unit. After receiving the normal signal or the warning signal, the training management unit immediately displays the preset warning text corresponding to the normal signal or the warning signal, that is, the preset warning text corresponding to the normal signal is: normal training, and the preset warning text corresponding to the warning signal is: force reminder, and then reminds the user to adjust the force decision based on the feedback information, which helps to shorten the user's rehabilitation training cycle.

[0017] Embodiment three: The movement disorder rehabilitation training method based on EEG signal recognition includes the following steps: Step 1: Division of training period and recognition quality evaluation process of EEG signal recognition model, that is, dividing the training period into test period and rehabilitation period, performing recognition test evaluation feedback analysis through recognition information of EEG signal recognition model in the test period, and outputting the obtained available signal or adjustment signal as feedback; Step 2: Based on the information progression, the motor imagery action consistency evaluation process is to conduct EEG and action dual evaluation analysis on the collected imagery EEG signals, and output the qualified or unqualified signals as feedback; Step 3: Based on the training completion performance of motor imagery movements, the set limb movements are divided into normal training movements and short board training movements, that is, the collected classified movement labels are analyzed for rehabilitation movement division management, and the obtained local completion Jg is discriminated and processed to obtain normal training movements and short board training movements, and the obtained standard signal or deviation signal is output and fed back; Step 4: Dynamically adjust the functional electrical stimulation management process based on the discrimination analysis of the user's muscle strength deviation value, that is, conduct muscle strength deviation risk analysis on the collected current muscle strength value and target muscle strength value of the user, discriminate the obtained muscle strength deviation value, and output the obtained enhancement instructions, conventional instructions and relief instructions as feedback; Step 5: User force decision management process based on in-depth information combined analysis, that is, force decision regulation and early warning analysis of enhanced instructions, conventional instructions and reduced instructions, and output feedback of the obtained normal signal or early warning signal; In summary, the present invention preliminarily analyzes from the perspective of EEG signal recognition to improve the accuracy of EEG signal recognition, and intuitively understands the recognition rate and accuracy of the EEG signal recognition model through information visualization feedback, thereby ensuring the rate of EEG signal recognition of motor imagery, and realizing high-speed and efficient movement disorder rehabilitation training, and performs EEG and movement dual-item evaluation and analysis through virtual training, so as to intuitively understand whether each training action of the user is qualified, and at the same time, performs a preliminary analysis from the perspective of the overall completion of the training to understand whether the overall completion of the user's training is up to standard, so as to reasonably adjust the training intensity, and further carry out targeted movement disorder rehabilitation training plans for the user's shortcomings, which is helpful to assist the next round of rehabilitation training tasks based on this; From the perspective of the user's muscle strength deviation, on the one hand, the functional electrical stimulation is dynamically adjusted according to the information feedback to improve the rehabilitation training effect. On the other hand, the user is reminded to adjust the force decision based on the feedback information, which helps to shorten the user's rehabilitation training cycle.

[0018] The threshold is set to facilitate comparison. The size of the threshold depends on the amount of sample data and the number of bases set by technicians in this field for each group of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0019] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A movement disorder rehabilitation training system based on EEG signal recognition, characterized in that: It includes rehabilitation training management center, test feedback unit, motor imagery assessment unit, rehabilitation division unit, force monitoring unit and training management unit; The rehabilitation training management center is used to collect the recognition information of the EEG signal recognition model established during the user's most recent training, and send the recognition information to the test feedback unit for recognition test evaluation feedback analysis to obtain a usable signal or an adjustment signal; When a usable signal is generated, the motor imagery evaluation unit is used to perform EEG and action dual evaluation analysis on the collected imagined EEG signals, and compare and analyze the obtained imagined EEG signals and imagined action information with the actual EEG signals and standard action information to obtain qualified signals or unqualified signals; the rehabilitation classification unit is used to perform rehabilitation action classification management analysis on the collected classified action labels, and perform discrimination processing on the obtained local completion degree Jg to obtain normal training actions and short board training actions; The force monitoring unit is used to perform a muscle force deviation risk analysis on the collected current muscle force value and target muscle force value of the user, and to perform judgment processing on the maximum number of consecutive A's to obtain a normal signal or a warning signal.

2. The movement disorder rehabilitation training system based on EEG signal recognition according to claim 1 is characterized in that: The identification test evaluation feedback analysis process is as follows: After collecting the user's training period, retrieve the EEG signal recognition model established during the user's most recent training from the server; The training period is divided into a test period and a rehabilitation period, and test EEG signals corresponding to the user's running imagination of n known limb movements in the test period are obtained, where n is a natural number greater than 3. The n test EEG signals are set as test training data, and the test training data are recognized through an EEG signal recognition model. Based on the EEG signal recognition model, recognition information of each test EEG signal is obtained, and the recognition information includes recognition time and recognition result. The recognition result includes qualified control and unqualified control. The mean of the recognition time is obtained, and the mean of the recognition time is set as the test recognition score, and the number of test EEG signals with qualified control recognition results is set as the accurate recognition rate. The test recognition score and the accurate recognition rate are discriminated and processed to obtain a usable signal or an adjusted signal.

3. The movement disorder rehabilitation training system based on EEG signal recognition according to claim 2 is characterized in that: The EEG and action dual evaluation analysis process is as follows: Limb rehabilitation training is performed based on a virtual training scenario, and limb movements are set. At the same time, actual EEG signals corresponding to the set limb movements are obtained, and the actual EEG signals corresponding to the set limb movements are identified. At the same time, control instructions of the actual EEG signals are obtained, and standard movement information of the limb rehabilitation robot is obtained based on the control instructions of the actual EEG signals. The standard movement information includes the limb rotation angle and the limb maintenance time.

4. The movement disorder rehabilitation training system based on EEG signal recognition according to claim 3 is characterized in that: Obtaining prompts for the user to perform corresponding limb movement imagination based on the set limb movement during the rehabilitation period, and then obtaining the imagined EEG signal of the user performing the corresponding limb movement imagination based on the signal acquisition device, and identifying the imagined EEG signal of the user performing the corresponding limb movement imagination, and at the same time obtaining the control instruction corresponding to the imagined EEG signal of the limb movement imagination, and obtaining the imagined movement information of the limb rehabilitation robot based on the control instruction corresponding to the imagined EEG signal; Compare and analyze the imagined EEG signal and imagined action information with the actual EEG signal and standard action information: if the imagined EEG signal is consistent with the actual EEG signal, and the imagined action information corresponds one-to-one with the standard action information, a qualified signal is generated; If the imagined EEG signal is inconsistent with the actual EEG signal, or the imagined action information does not correspond one-to-one with the standard action information, an unqualified signal is generated.

5. The movement disorder rehabilitation training system based on EEG signal recognition according to claim 1 is characterized in that: The rehabilitation action division management and analysis process is as follows: The set limb movements are divided into g classified action labels, where g is a natural number greater than zero. The classified action labels include left hand movements, right hand movements, and left leg movements. The total number of each classified action label is obtained, and the number of qualified signals generated by each classified action label is also obtained. The ratio between the number of qualified signals generated by the classified action label and the total number of classified action labels is set as the local completion degree Jg, and then the average value of the local completion degree Jg is obtained. The average value of the local completion degree Jg is set as the rehabilitation assessment completion degree.

6. The movement disorder rehabilitation training system based on EEG signal recognition according to claim 5 is characterized in that: The rehabilitation assessment completion degree is judged: if the rehabilitation assessment completion degree is greater than or equal to the preset rehabilitation assessment completion degree threshold, a stable signal is generated; if the rehabilitation assessment completion degree is less than the preset rehabilitation assessment completion degree threshold, a feedback signal is generated. When the feedback signal is generated, the local completion degree Jg is judged: If the local completion degree Jg is greater than or equal to the local completion degree threshold, the corresponding classified action is judged to be a normal training action, and a standard-reaching signal is generated; if the local completion degree Jg is less than the local completion degree threshold, the corresponding classified action is judged to be a shortboard training action, and a deviation signal is generated.

7. The movement disorder rehabilitation training system based on EEG signal recognition according to claim 2 is characterized in that: The muscle strength deviation risk analysis process is as follows: Based on the current muscle strength value and target muscle strength value of the user in the rehabilitation period obtained under the set limb movement, the value obtained by subtracting the target muscle strength value from the current muscle strength value and dividing it by the target muscle strength value is set as the muscle strength deviation value, and the muscle strength deviation value is judged and processed. If the muscle strength deviation value is less than the minimum value in the preset muscle strength deviation value range, an enhancement instruction is generated; if the muscle strength deviation value falls within the preset muscle strength deviation value range, a regular instruction is generated; if the muscle strength deviation value is greater than the maximum value in the preset muscle strength deviation value range, a reduction instruction is generated.

8. The movement disorder rehabilitation training system based on EEG signal recognition according to claim 7 is characterized in that: The enhancement instruction or the reduction instruction is set as A, and the general instruction is set as B. The character strings of A and B are constructed based on the generation order of A and B, and the character strings constructed of A and B are set as the muscle strength label string. The maximum number of consecutive occurrences of A is obtained from the muscle strength label string, and the maximum number of consecutive occurrences of A is judged: if the maximum number of consecutive occurrences of A is greater than or equal to a preset threshold, a warning signal is generated; If the maximum number of consecutive A occurrences is less than the preset threshold, a normal signal is generated.

9. A method for rehabilitation training of movement disorders based on EEG signal recognition, the method being applied to the system for rehabilitation training of movement disorders based on EEG signal recognition according to any one of claims 1 to 8, characterized in that: The following steps are involved: Step 1: Division of training periods and recognition quality evaluation process of EEG signal recognition model; Step 2: The process of evaluating the consistency of motor imagery based on information progression; Step 3: Based on the training completion performance of the motor imagery action, the set limb movements are divided into normal training movements and short board training movement processes; Step 4: Dynamically adjust the functional electrical stimulation management process based on the user's muscle force deviation value discrimination analysis; Step 5: User-driven decision-making management process based on in-depth information analysis.

Images