RU2523349C1 - Method for rehabilitation of patients suffered stroke - Google Patents

Abstract

FIELD: medicine.

SUBSTANCE: patient is trained by challenging to imagine a paretic limb movement and controlling the imagined movement. An electroencephalography is recorded; the recorded data are transferred to a computer for synchronous processing, and arousal reaction signals of a sensorimotor rate responsible for the imagined movement are isolated by means of an EEG pattern classifier according to the Bayes method. Identification results of the mental challenge are presented to the patient by a visual feedback in the form of a mark on a monitor. The mark re-position testifies to accuracy of the challenge. The imagined movement challenge is presented for 10 seconds. The training course makes 6-12 days, one training a day, of the length of 20-30 minutes every 1 to 4 days.

EFFECT: method provides more effective rehabilitation that is ensured by the feedback training in the conditions enabling the patient controlling the imagined paretic limb movement visually.

3 dwg, 4 tbl, 2 ex

Description

The invention relates to medicine, in particular to neurology, and can be used to restore motor function in patients who have suffered an ischemic stroke and have severe motor impairment.

It is known that repeated active purposeful movements significantly contribute to the restoration of motor functions. In patients with a stroke, in addition to physical therapy, this approach has been implemented over the past 10 years in methods such as training in virtual reality, when patients are trained in feedback movements, and training in conditions where movement is facilitated by a robot (Chernikova L.A. Brain plasticity and modern rehabilitation technologies, Annals of Clinical and Experimental Neurology, 2007, 1 (2): 40-47).

These methods require maintaining the ability to make active movements with a paralyzed limb and, therefore, require partial preservation of motor functions. In the case of gross paresis or plegia, a promising method of stimulating the plasticity of the brain is the imagination of movements. The physiological mechanisms underlying the process of motion imagination are largely similar to the mechanisms of motor control and therefore can stimulate the same plastic mechanisms of the brain as the actual execution of the movement (Neuper C, Scherer R., Reiner M., Pfurtscheller G. Imagery of motor actions: differential effects of kinesthetic and visual-motor mode of imagery in single-trial EEG. Cogn. Brain Res. 2005.25 (3): 668-677).

To date, there is a known method for the rehabilitation of stroke patients by teaching the patient "precision grip" using biofeedback with feedback on an electromyogram (EMG). For precision grip training use the Boslab hardware and software complex (Novosibirsk). Electromyographic electrodes are applied to the muscles of the elevation of the thumb. The patient squeezes a small object with his thumb and forefinger and learns to maintain a given level of muscle activity at 20, 40 or 60% of the maximum contraction level. On the monitor screen, the patient sees a given level of EMG in the form of a corridor, as well as changes in his integrated EMG. The task is to combine the level of EMG with a given corridor and maintain the level of muscle tension for 30 s. After each test, the patient relaxes the muscles as much as possible under visual control (training to overcome spasticity), and then tries to reproduce the required level of EMG from memory, without seeing the EMG on the screen, i.e. without visual control. The duration of the session is 20-25 minutes. The course consists of 10 daily sessions. (Chernikova L.A., Ioffe M.E., Busheneva S.N., Shestakova M.V., Bilimenko A.E. Electromyographic biofeedback and functional magnetic resonance imaging in post-stroke rehabilitation (using the example of precision grasp training). Bulletin Siberian Medicine, 2010, No. 2, pp. 12-16).

This allows you to actively include biocontrol in motor rehabilitation programs, organized by an electromyogram and aimed at developing accurate movements of the fingers in patients with mild and moderate paresis and impaired fine motor skills, but does not provide full rehabilitation in patients with severe motor impairments.

The technical result consists in increasing the effectiveness of rehabilitation in patients who have had a stroke and have severe motor impairments due to the influence of training imagination of movement on the neurophysiology of cortical motor representations, which ensures an increase in the excitability of the motor cortex.

The technical result is achieved by the fact that the rehabilitation of stroke patients is carried out by training the paretic limb, while the training is carried out by presenting the patient with a task of imagining movement of the paretic limb, followed by monitoring the patient’s imagination of movement based on the analysis of electroencephalogram patterns (EEG) arising from imagination of movement with a paretic limb, and EEG registration of brain signals is carried out from the system of active electrodes of the encephalograph, data on the surface of the head, and transferring this data to a computer for synchronous processing and extracting signals responsible for the imagination of movement using the Bayesian EEG pattern classifier, the results of recognition of the mental task being performed are presented to the patient by visual feedback confirming the correctness of the task on the screen monitor, and the task of imagining the movement is presented within 10 seconds, the training course of the paretic limb is 6-12 days, one training session per day ny, lasting 20-30 minutes with intervals between workouts from 1 to 4 days.

The method is as follows.

For the rehabilitation of a patient who has had a stroke, they use technology with feedback - the brain-computer interface (IMC). The patient receives real-time visual information about the activity of his brain.

IMC consists of the following elements (Fig. 1): ActiCap active electrode systems from Brain Products (Germany) for recording an electroencephalogram (EEG) (1), NBL640 encephalographic analog-to-digital converter (ADC) manufactured by NeuroBioLab LLC (2), personal computer (Windows 7 operating system) for synchronous data transfer, separation of EEG performance and signal classification for real-time recognition of a control command (3) Bobrov P.D., Korshakov A.V., et al. Bayesian approach to interface implementation brain-computer p based on the representation of movements. Journal of Higher Nervous Activity and Neurophysiology, 2012, 62 (1): 89-100).

During the training session, the patient sits in a comfortable chair so that his head is 1 meter away from the computer monitor, on which he is presented with visual tasks and feedback (information about the correctness of the task). In the center of the monitor screen there is a circle, which serves to fix the gaze, and 3 rhomboid arrows located around it, which are indicated on the block diagram of the layout of the IMC used in this rehabilitation process (Fig. 1). During the training, the patient performs one of three tasks: relax (when the upper arrow lights up), imagine the movement of the left or right hand (when the corresponding left or right arrow lights up). As a representation of the movement of the arm, they suggest imagining a slow clenching of the hand into a fist. On a task to relax (rest), the patient should sit quietly and look at the center of the screen. Tasks are presented in random order, each for 10 seconds. Recognition results of the mental task being performed are presented to the patient by visual feedback: the mark in the middle of the screen that fixes the gaze turns green if the classifier recognizes the task corresponding to the presented motion task and remains white if the task was not recognized.

EEG registration is carried out using 30 electrodes located on the surface of the patient’s head, according to the scheme 10-20 (Zenkov L.R. Clinical electroencephalography (with elements of epileptology). Guidelines for doctors. 3rd ed., M: MEDpressinform, 2004, 368 from.). A special gel is applied under each electrode to improve contact with the surface of the patient’s head. EEG signals are filtered in the frequency band from 5 to 30 Hz. The EEG pattern classifier based on the Bayes method (Bobrov P.D., Korshakov A.V. et al. The Bayesian approach to the implementation of the brain-computer interface based on the representation of movements was used in this technique. Journal of Higher Nervous Activity and Neurophysiology, 2012, 62 (1): 89-100). As an indicator of classification accuracy, the Cohen’s kappa index is used (with ideal recognition, k = 1, with random recognition, k = 0) (Kohavi, R., Provost F. Glossary of terms. Machine Learning - Special Issue of Applications of Machine Learning and the Knowledge Discovery Process, 1998, 30: 271-274) and the percentage of correct answers of the classifier (recognition is higher than random at> 35%). The sources of activity most significant for the functioning of the BMI were determined using the Independent Component Analysis (ICA) method (Frolov A., Husek D. et al. Comparison of four classification methods for brain computer interface, Neural Network World, 2011, 21 (2): 101-111).

With each patient, rehabilitation is carried out for 6-12 days, one training a day, lasting 20-30 minutes. The intervals between trainings are from 1 to 4 days.

A study was conducted to compare the quality of management of BMI in patients with focal brain damage and in healthy people, as well as to evaluate the effectiveness of this technique for the rehabilitation of stroke patients. The study involved 36 patients with hemiparesis of central genesis, of which 26 were men and 10 women, aged 34 to 70 years, with the duration of the disease (stroke or brain injury) from 1 month to 8 years (median age of the disease was 11.5 [3; 21.5] months). There were 14 patients in the early recovery period, and 22 patients in the late and residual period. In all patients, a very rough paresis or plegia in the arm was observed in the clinical picture, the median on the ARAT scale = 0 [0; four]. 16 patients constituted the main group, the complex treatment of which included training of IMC, and a comparison group consisting of 20 patients who received only traditional complex therapy. The main group and the comparison group were comparable in age, stroke duration and degree of neurological deficit. In addition, the study involved 7 healthy volunteers aged 24 to 68 years, 6 men and 1 woman. All subjects were right-handed (according to the Edinburgh manual asymmetry questionnaire: R> 40%).

Thirteen patients of the main group (3 patients retired after the first session of their own free will) and all healthy subjects underwent a course of IMC training, in which the task of imagination of movement was presented for 10 seconds and carried out one training session per day, lasting 20-30 minutes intervals between workouts from 1 to 4 days. The achieved results of BCI management for trained patients of both groups are presented in table 1 (data are presented as median and 25%, 75% percentiles).

Table 1 The achieved values of recognition by the classifier of mental tasks (peace, imagination of movement with the left hand and imagination with the right hand) in subjects of two groups Kappa cohen % recognition The main group (patients with gross paresis or plegia in the hand) who underwent a course of IMC training 0.37 [0.19; 0.43] 55.5 [45; 59.5] A group of healthy subjects who underwent a course of IMC training 0.33 [0.22; 0.44] 54 [47; 63] U test P = 0.77 P = 0.98

According to the indicators of the interface control quality (Cohen's kappa and recognition percentage), no differences between the groups were revealed (P> 0.05). This allows us to conclude that patients with focal brain damage are able to master the management of BCI with the same quality as people without neurological disorders.

In fig. Figure 2 shows the distribution of the contribution of the components to the potentials recorded on the surface of the head (topographic distribution), and the spectral activity densities for these components (for one of the healthy subjects). Topographic distributions of components important for controlling the IMC and their spectral densities under three mental states are the “norm”. The abscissa is the frequency in hertz, the ordinate is the spectral density normalized to the maximum value. The peak spectral density for these components corresponds to the range of the sensorimotor rhythm (mu rhythm). During the imagination of movement, a desynchronization reaction of the sensorimotor rhythm takes place in the corresponding hemisphere (lower curve in the spectral density graph).

Such a distribution of potentials and a change in mu rhythm during the imagination of movement was observed in healthy subjects and was taken as the "norm".

In fig. Figure 3 shows the topographic distributions of significant components and the spectral densities of their intensity under three mental states (rest, imagination of movement with the left hand and imagination with the right hand): A) in a patient with subcortical brain damage; B) in a patient with damage to the motor cortex. Among the patients, the following variants of the distribution of the contribution of the components to the potentials depending on the location of the stroke were found: with subcortical localization of the lesion, a pattern similar to the “norm” was observed (Fig. 3A), and with damage to the motor cortex, an atypical distribution of the contribution of components in the damaged hemisphere was observed: In cases, the signal source was localized closer to the CZ projection zone with a standard arrangement of EEG electrodes in a 10-20 system (Fig. 3B). The components important for the management of BCI, as well as in healthy subjects, were associated with the desynchronization reaction in the mu rhythm range (10 patients), and beta rhythm (1 patient, Fig. 3B), or mu and partially beta rhythm (1 patient, Fig. 3A) during the imagination of movement.

Among the subjects of both groups, the following variants of changes in brain activity were found, which were associated with an improvement in the quality of management of IMC (mastery of the skill of managing IMC): a) the selection of significant components from the first days and increased response of desynchronization during training; b) the absence of significant components in the early days and their appearance during the training process with the subsequent intensification of the desynchronization reaction. The topographic distribution did not change for each of the two components as a result of the training.

To assess the clinical effectiveness of rehabilitation in patients with central paresis of the arm, a comparative analysis of the restoration of motor function in the main group (13 patients) and the comparison group (20 patients) was performed. It should be noted that in this study, both the main group and the comparison group included patients only with plegia or very rough paresis of the hand with a poor prognosis for restoration of movements.

After the rehabilitation measures, patients of the main group showed a statistically significant improvement in the motor function of the arm (on the ARAT scale). In the comparison group, no significant improvement in impaired motor function was observed. The difference in improving hand function between groups is statistically significant (p = 0.02). The results of the assessment of the motor function of the arm in dynamics in patients of the main group and the comparison group are presented in table 2.

table 2 Results of an assessment of arm motor function on the ARAT scale before and after rehabilitation Before After R Main group 10; 4,5] 5 [0; 16] 0.01 Comparison group 0 [0; four] 0 [0; 5] 0.48

Thus, the data obtained allow us to conclude that, not only in healthy, but also in patients with brain damage, with the imagination of the movement of both a healthy and paretic hand, a desynchronization reaction is detected in the corresponding hemisphere of the brain. Moreover, for the first time in the study, it was shown that the topographic distribution of the components in trained healthy subjects and patients with subcortical stroke corresponded to the location of the arm in the primary sensorimotor areas of the central sulcus, while at the same time, patients with extensive damage to the cortex under imagination of movement of the parietal arm showed a shift localization of the source of activity in the region of the Cz electrode

As a result, a pattern was revealed that indicates that patients with focal brain damage are able to master the management of BCI with the same quality as subjects without neurological impairment. In this case, no dependence of the quality of BCI control on the degree, localization, or prescription of brain damage was found.

In addition, 42% of patients with plegia and gross paresis of the hand after the training course had a positive effect when using IMC training, which was expressed in a significant improvement in the function of the paretic brush. Therefore, this rehabilitation method will accelerate the pace and make the restoration of impaired functions more complete, prevent the development of secondary complications (thrombophlebitis, contractures, pressure sores, congestive pneumonia, etc.) in patients with different severity of brain damage.

Examples of the method

Example 1

Patient A., 70 years old, ischemic stroke, disease duration 1.5 months, MRI stroke focus: in the left hemisphere of the brain in the posterolateral part of the lenticular nucleus with transition to the radiant crown and posterior thigh of the inner capsule. Rough paresis of the arm (7 points on the ARAT scale).

A course of trainings was conducted from 8 sessions of 20 minutes with intervals between training sessions of 1 day. The task of the imagination of the movement was presented for 10 seconds. Achieved indicators of the quality of management of IMC by the 8th day of training: Kappa Cohen = 0.42; recognition percentage = 70%. Management of the IMC was carried out due to the desynchronization reaction of the sensorimotor rhythm in the central groove of the translateral hemisphere with the imagination of movement. After the rehabilitation, a clinically significant improvement in hand function was observed (by 11 points on the ARAT scale, table 4). The patient independently takes items from the table, opens the door handle.

Table 4 Improving hand function in patient A. after a course of IMC training Section ARAT Before the course After the course Maximum score Ball grip four 8 eighteen Cylinder grip 2 5 12 Plucked grip one 5 eighteen Major movements 0 0 9 Total 7 eighteen 57

Example 2

Patient B., 40 years old, ischemic stroke, disease duration 21 months, MRI stroke focus: in the left hemisphere of the brain in the lenticular nucleus with extension to the posterior thigh of the internal capsule, upper sections of the hippocampal gyrus. Rough paresis of the hand (4 points on ARAT).

A course of trainings was conducted from 12 sessions, one training a day, lasting 30 minutes with intervals between training 4 days. The task of the imagination of the movement was presented for 10 seconds. Achieved indicators of the quality of management of IMC by the 12th day of training: Kappa Cohen = 0.30; recognition percentage = 50%. Management of the IMC was carried out due to the desynchronization reaction of the sensorimotor rhythm in the central groove of the translateral hemisphere with the imagination of movement. After the rehabilitation, an improvement was observed in the function of the hand, including the hand (by 13 points on the ARAT scale, table 5). The patient may take some items from the table.

Table 5 Improving the function of the hand in patient A. after a course of IMC training. Section ARAT Before the course After the course Maximum score Ball grip one 7 eighteen Cylinder grip 0 four 12 Plucked grip 0 one eighteen Major movements 3 5 9 Total four 17 57

Thus, the proposed method for the rehabilitation of stroke patients provides an increase in its effectiveness due to the influence of movement imagination training on the neurophysiology of cortical motor representations. This increases the excitability of the motor cortex. The IMC-training course that we developed improves the accuracy of voluntary muscle activity in the limbs in patients with severe motor impairment after training with visual feedback. This allows you to actively include in the program of motor rehabilitation a course of IMC training, organized by an electroencephalogram and aimed at developing recovery of motor function, especially in patients who have suffered an ischemic stroke and have severe motor impairments.

Claims (1)

A method for the rehabilitation of stroke patients by training paretic limb, characterized in that the training is carried out by presenting the patient with an imagination of movement of the paretic limb with subsequent control of the patient’s imagination of movement based on the analysis of electroencephalogram patterns (EEG) arising from the imagination of movement of the paretic limb , while the registration of EEG brain signals is carried out with a system of active electrodes of the encephalograph, placed on the surface of the head, and transmitting sensing this data to a computer for synchronous processing and extracting signals responsible for the imagination of movement using the Bayesian EEG classifier, the recognition results of the mental task being performed are presented to the patient by visual feedback in the form of a mark on the monitor screen, the change of which determines the correct execution tasks established by comparing the parameters of the sensorimotor rhythm desynchronization reaction with those of healthy subjects in the corresponding hemisphere during oobrazheniya motion, the task of placing imagination motion for 10 seconds, the training course paretic limb is 6-12 days, one day in the exercise, duration of 20-30 minutes with intervals between training sessions 1 to 4 days.

Images