Physical function improvement system

Abstract

The present invention addresses the problem of providing a system capable of performing stable neural network reinforcement training independent of motion image capability without actual operation.　In the present invention, it is found that it is possible to activate a neural network by constructing an NIRS-NFB system using VR in combination and executing a problem in a pseudo manner under a VR environment.

Description

Physical function improvement system

[0001] The present invention relates to a physical function improvement system and a program.

[0002] In the human brain, more than 10 billion nerve cells form a functional network to realize various thoughts and behaviors. Regarding motor control, the strengthening of the network through repeated learning leads to the maintenance and enhancement of functions. The decline of the brain network function due to diseases, aging, etc. results in the decline of motor function. Exercise therapy and rehabilitation are considered useful for the treatment and prevention of these functional declines.

[0003] The decline of brain function due to aging and diseases is involved in the decline of walking ability. Only about 50% of stroke patients can acquire independent walking after a stroke, and 10 - 25% of community-dwelling elderly people experience falls annually. Falls among the elderly lead to an increase in the burden of medical care and nursing.

[0004] Human standing and walking involve the central nervous system, especially the cerebral cortex. In particular, the network associated with a region called the supplementary motor area plays an important role in controlling the walking posture, and the activation of the supplementary motor area has been reported by the present inventors to bring about improvement in walking disorders after a stroke (Non-Patent Document 1).

[0005] In the field of neurorehabilitation, training using motor imagery (motor imaging), in which patients imagine themselves moving, is being carried out, and neurofeedback (Neurofeedback; NFB) training, which displays the measurement results of brain activity using a brain activity measurement device and provides feedback to the subject, has attracted attention.

[0006] The present inventors have developed an NFB system using near-infrared spectroscopy (Near infra-red spectroscopy; NIRS) that places less burden on the subject (Non-Patent Document 2 and Patent Document 1). However, training using a neurofeedback system in a virtual reality (Virtual reality; VR) environment has not been reported.

[0007] Fujimoto H et al., Neuroimage. 2014 Jan 15;85Pt 1:547-54Mihara M et al.,PLoS One. 2012;7(3):e32234

[0008] Patent No. 6502488

[0009] Neurofeedback involves measuring the brain activity of a subject in response to a set task, analyzing the measured data, and converting it into feedback information. The feedback information is presented to the subject, and by performing motor imagery to improve the brain activity indicated by the received feedback, the brain activity pattern is updated, and the subject learns how to voluntarily control their brain activity. Strengthening task-specific neural networks through neurofeedback requires activation of the neural networks associated with the task. For motor tasks, it is necessary to either actually perform the task or simulate the activation of task-specific networks through motor imagery. However, tasks such as walking and falling are difficult for trainees to perform in practice. Furthermore, there are individual differences in motor imagery ability, making stable training difficult. On the other hand, virtual reality (VR) environments allow for the recall of images close to real-world movement through 3D visual stimuli, and its application in rehabilitation is being attempted. However, VR-only intervention has low training efficiency and insufficient effect size, preventing widespread adoption.

[0010] The objective of this invention is to provide a system that enables stable neural network strengthening training that does not depend on motor imagery ability, without the need for actual movement.

[0011] As a result of diligent research to solve the above problems, the inventors of this invention constructed a NIRS-NFB system that uses VR and found that it is possible to activate neural networks by performing tasks in a simulated VR environment, thus completing the present invention.

[0012] In other words, the present invention comprises the following: 1. A system for improving physical function, comprising: a brain activity measurement device for measuring the brain activity of a subject; a data processing device for calculating a statistical quantity that shows the statistical significance of changes in measurement data reflecting brain activity based on measurement data acquired from the brain activity measurement device; and a display device for displaying VR video of an exercise task related to posture control and a real-time brain activity index corresponding to the statistical quantity on the same screen. 2. The system for improving physical function according to item 1, wherein the display device is a head-mounted display device, and the exercise task includes one or more tasks selected from the group consisting of standing, walking, falling, balance training, and high-intensity exercise. 3. The system for improving physical function according to item 1 or 2, wherein the data processing device generates VR video of the exercise task including the brain activity index, and the brain activity index is represented by virtual objects in the VR space of the exercise task. 4. The system for improving physical function according to any one of items 1 to 3, wherein the brain activity measurement device is an optical measuring device or an electroencephalogram measuring device. 5. A physical function improvement system according to any one of paragraphs 1 to 4, further comprising a sensory stimulation output device of 5.1 or more. 6. A physical function improvement system according to any one of paragraphs 1 to 5, wherein the playback time per instance of the VR video of the exercise task is 0.5 to 60 seconds, the playback interval of the VR video of the exercise task is 1 to 60 seconds, and the number of times the VR video of the exercise task is played is 2 or more. 7. A program that causes a computer to function as a physical function improvement system according to any one of paragraphs 1 to 6.

[0013] According to the present invention, it is possible to perform stable neural network strengthening training that does not depend on motor imagery ability, without actual movement.

[0014] This is a block diagram of the physical function improvement system of the present invention (hereinafter sometimes referred to as "the system of the present invention"). This is a flowchart showing the flow from the acquisition of measurement data by the data processing device in the present invention to the output of VR images. This is a schematic diagram showing an example of the system of the present invention. (Example 1) Results of safety evaluation of training using the system of the present invention. (Example 1) Results of evaluation of the subject's postural control ability before and after training using the system of the present invention. (Example 1) Results of task-related brain activity measurement of the subject during training using the system of the present invention. (Example 1)

[0015] The present invention will be described in detail below.

[0016] Figure 1 is a block diagram showing the overall configuration of a physical function improvement system according to one embodiment of the present invention. The physical function improvement system of the present invention comprises a brain activity measurement device, a data processing device, and a display device.

[0017] Brain activity measurement devices measure the brain activity of a subject and generate time-series brain activity measurement data. Examples of brain activity measurement devices include optical measurement devices and electroencephalography (EEG) measurement devices. Optical measurement devices measure brain activity using near-infrared spectroscopy (NIRS). In the brain, local changes in blood flow due to neural activity cause changes in the concentration of hemoglobin in the blood. The measurement principle of NIRS is to irradiate the brain with near-infrared light, which has high penetration into living organisms, and measure the reflected light due to scattering, thereby measuring changes in cerebral blood flow associated with brain activity as changes in hemoglobin concentration. Hemoglobin molecules have different absorption spectra for oxygenated hemoglobin and deoxygenated hemoglobin, and by measuring the absorbance changes using multiple near-infrared wavelengths with different wavelengths, the changes in oxygenated hemoglobin, deoxygenated hemoglobin, and total hemoglobin can be measured. An electroencephalogram (EEG) measuring device can measure changes in potential difference using bilateral lobe-connected electrodes as reference electrodes and active electrodes such as the Cz point (parietal region) and F3 / F4 points (frontal region). Examples of EEG measuring devices include the Neurofax EEG-1275 (Nihon Kohden Corporation) and the Starstim™ system (Neuroelectrics). Optical measurement is particularly advantageous as a brain activity measurement device in this invention because it is less prone to noise and artifacts due to body movement compared to EEG measurement, the brain activity analysis evaluation values ​​tend to be more stable over several trials, and the time and effort required for attaching measurement electrodes is reduced. By having multiple measurement channels or electrodes, the brain activity measurement device can measure brain activity in each brain region and include positional information of the measurement site in the measurement data.

[0018] The brain activity measurement device may be, for example, the brain activity measurement device described in Patent Document 1. The brain activity measurement site by the brain activity measurement device can be any site where task-related brain activity of the subject during training can be measured. The cerebral cortex is preferred, the prefrontal cortex and supplementary motor area are more preferred, the supplementary motor area with strong brain activity is even more preferred, and considering that the supplementary motor area is located on the medial surface of the cerebrum, it is more preferable to measure the activity of the upper supplementary motor area, which is closer to the surface of the head and has a particularly high activation effect, for ease of measurement. Therefore, the brain activity measurement data obtained from the brain activity measurement device is preferably cerebral cortex activity measurement data, preferably prefrontal cortex and / or supplementary motor area activity measurement data, and more preferably supplementary motor area activity measurement data.

[0019] The data processing device is connected to a brain activity measurement device and calculates a statistical measure (t-value) that indicates the statistical significance of changes in the measurement data reflecting brain activity, based on the measurement data acquired from the brain activity measurement device. In this case, the data processing device may be configured to calculate the statistical measure (t-value) by correcting the scalp blood flow component of the measurement data acquired from the brain activity measurement device in real time based on the scalp blood flow data.

[0020] The data processing device may be, for example, a device in which a VR application program (hereinafter sometimes referred to as a VR application) is further stored in the data processing device described in Patent Document 1. The data processing device consists of a control unit including a CPU and memory, and a PC (personal computer) equipped with a storage unit such as a hard disk drive or flash memory. The PC functions as a data processing device for the brain activity feedback system by the control unit (CPU) executing an analysis program stored in the storage unit. The control unit acquires measurement data by communicating with a brain activity measurement device. The analysis program is not particularly limited as long as it can analyze the measurement data and calculate statistics, but for example, a program built on MATLAB (MathWorks Inc.) can be used. The method for analyzing the measurement data may be part or all of the analysis methods described in various documents (for example, PLoS ONE. 2012;7(3):e32234 and J Neurophysiol. 2020; 124(6):1875-1884, etc.), and may be modified as appropriate. If the brain activity measurement device is an optical measurement device, for example, data from approximately 10 to 120 seconds, including the timing of the most recent 1 to 5 events, can be used from optically obtained cerebral blood flow data. Estimates obtained by statistical inference using a generalized linearization model from the cerebral blood flow response associated with the event, predicted by the cerebral blood flow response function (Hemodynamic response function), can be used as evaluation values ​​for brain activity related to the event. If the brain activity measurement device is an electroencephalogram (EEG) measurement device, for example, EEG data from 0.5 seconds before the start of an event such as a fall in a motor task to 3.5 seconds after the start can be processed using a 1-25 Hz bandpass filter. The largest negative potential occurring between 80 ms and 500 ms after the start of the event can be identified as a postural control-related brain potential, and the ratio of this potential to the average value of the action potentials over the 4-second period from 0.5 seconds before the start of the event to 3.5 seconds after the start can be used as an evaluation value for the magnitude of postural control-related brain activity.

[0021] The data processing device is communicatively connected to the display device and is configured to output a video signal to the display device that includes brain activity indicators corresponding to the calculated statistics and a motor task. The data processing device executes a VR application for utilizing and presenting control of the VR space of the motor task. In one example, the VR application is stored in the memory unit and executed by the control unit. The VR application may be configured to acquire statistics calculated by the analysis program and output the corresponding brain activity indicators within the VR space of the motor task. In one example, the data processing device is configured to output a VR video signal including the motor task and brain activity indicators to the display device. Figure 2 shows the flow from the acquisition of measurement data by the data processing device to the output of the VR video signal in an exemplary embodiment. When the analysis program acquires measurement data from the brain activity measurement device (step S1), it calculates a statistic (t-value) that shows the statistical significance of the change in the measurement data (step S2), the VR application acquires the statistic (t-value) (step S3), generates a VR video of the motor task that includes the display of brain activity indicators that are transformed according to the statistic (step S4), and outputs a VR video signal to the display device (step S5).

[0022] In this invention, the term "brain activity index" is a broad concept that includes not only the statistical quantity itself, but also feedback indices, symbols, graphs, and other representations that reflect the statistical quantity. The brain activity index may include information about the measurement site where brain activity was measured.

[0023] The display device is configured to display VR video of the motor task output from the data processing device and real-time brain activity indicators that change according to the statistics on the same screen. This allows subjects to receive real-time feedback on their own brain activity in response to the VR video of the motor task they are viewing, enabling them to efficiently learn how to voluntarily control the brain activity of brain regions related to posture control. In one example, the display device displays VR video including the motor task and brain activity indicators output from the data processing device.

[0024] The VR video of the motor task is represented using moving images (videos) that represent a virtual reality (VR) space in which virtual objects exist, preferably using video showing the scenery as seen from a virtual camera set up in the VR space. The virtual camera is a virtual viewpoint set up in the VR space, preferably a viewpoint corresponding to the gaze direction of the subject looking at the display device, and may be configured so that its position, direction, field of view range, etc., in the VR space can be changed as appropriate. The virtual object is an object that does not exist in the real world and is represented only on the computer system. The virtual object is represented by three-dimensional computer graphics (CG) using image materials independent of live-action images. For example, the virtual object may be represented using animation materials, or it may be represented to be close to the real thing based on live-action images. The VR space is a virtual three-dimensional space generated by a computer. Brain activity indicators may be represented by virtual objects in the VR space of the motor task. The virtual object of the brain activity indicator may be a graph, figure, etc., that deforms in real time according to the calculated statistics. The VR space for motor tasks may be represented using only visual stimuli, or it may be represented in combination with somatosensory stimuli such as hearing, touch, smell, and taste.

[0025] The exercise tasks are not particularly limited as long as they relate to postural control, but include, for example, standing, walking, falling, balance training (e.g., shifting center of gravity, standing on one leg, changing direction, responding to external disturbances, etc.), and high-intensity exercise (e.g., climbing stairs, jumping, running, various sports, etc.). In this specification, "postural control" is used as a concept that includes not only static postural control but also dynamic postural control such as walking. A suitable example of a standing exercise task is imagining the exercise of maintaining a standing position for a certain period of time. A suitable example of a walking exercise task is imagining the exercise of maintaining walking for a certain period of time. A suitable example of a falling exercise task is imagining the exercise of maintaining posture or taking a defensive stance during a fall. In one example, the VR video of the exercise task is a moving image showing scenes of standing, walking, falling, balance training, and high-intensity exercise as seen from a virtual camera set up in the VR space. The motor task is not particularly limited as long as it is a motor image video related to postural control, but tasks involving intentional (conscious) movements of the trunk and lower limbs are particularly effective. For example, tasks such as walking while avoiding obstacles, crossing a crosswalk by only stepping on the white parts, and going up and down stairs are examples of more effective tasks. Therefore, it is preferable that the VR video of the motor task includes virtual objects of obstacles that the subject should avoid when performing the motor image, and / or virtual objects of targets that the subject should contact in the motor image they perform. The virtual space may further include real objects that actually exist in the real world. The playback time per session of the VR video of the motor task is not particularly limited and may be, for example, 0.2 seconds to 10 minutes, but from the viewpoint of efficiently learning appropriate motor imagery and brain activity patterns, it is preferably 0.5 to 60 seconds, more preferably 0.6 to 30 seconds, more preferably 0.7 to 20 seconds, more preferably 0.8 to 10 seconds, more preferably 1 to 10 seconds, and more preferably 1 to 6 seconds.The number of times the VR video of the motor task is played in one training session is not particularly limited and may be, for example, 1 to 50 times. However, from the viewpoint of efficiently learning appropriate motor imagery and brain activity patterns, it is preferable to play it 2 or more times (e.g., 2 to 50 times), more preferably 5 or more times (e.g., 5 to 50 times), more preferably 8 or more times (e.g., 8 to 50 times), more preferably 10 or more times (e.g., 10 to 50 times, 10 to 40 times), and more preferably 15 or more times (e.g., 15 to 50 times, 15 to 35 times). When the number of times the VR video of the motor task is played in one training session is 2 or more times, the playback interval (rest period) is not particularly limited and may be, for example, 0 to 60 seconds. However, from the viewpoint of efficiently learning appropriate motor imagery and brain activity patterns, it is preferable for the VR video to be played intermittently. Specifically, the playback interval is preferably 1 to 60 seconds, more preferably 1 to 30 seconds, more preferably 1 to 20 seconds, and more preferably 5 to 10 seconds. The VR video during the rest period is not particularly limited and may, for example, be a still image of the VR video at the end of the previous exercise task, or it may be another display indicating that it is a rest period. Virtual objects such as obstacles, targets, and scenery in the VR space of the exercise task may be configured to change according to statistical quantities calculated by the analysis program. The VR application may be configured to allow operations such as pausing, playing, changing the speed of the VR video of the exercise task, and changing the position, direction, and field of view of the virtual camera in the VR space. The system of the present invention may further include an operating device, such as a keyboard or controller, that is communicably connected to the data processing device for a training assistant (e.g., a doctor, physical therapist, caregiver, etc.) or subject to perform such operations.

[0026] The display device is configured to display changes in statistical quantities as brain activity indicators in real time by changing the display representing statistical quantities displayed on the VR video of the motor task, preferably by changing at least one of the shape and color of the display representing statistical quantities displayed on the VR video of the motor task. A configuration that displays changes in statistical quantities as brain activity indicators in real time may be, for example, the configuration of the display device described in Patent Document 1.

[0027] The display device changes the display of brain activity indicators according to the activity level of brain regions related to the task, such as the cerebral cortex, preferably the prefrontal cortex and / or supplementary motor area. For example, the shape of the brain activity indicator may be changed so that a larger shape indicates greater activity in the task-related region and a smaller shape indicates less activity. The display color of the brain activity indicator may also be changed so that a warmer color indicates greater activity in the task-related region and a cooler color indicates less activity (for example, the colors may change from yellow, red, blue, green, and black in order of decreasing activity). These changes in shape and color may also be combined.

[0028] The display device is not particularly limited as long as it is capable of displaying VR images, but it is preferably a head-mounted display (HMD) device.

[0029] The physical function improvement system of the present invention may further comprise one or more sensory stimulus output devices. The sensory stimulus output devices are communicatively connected to a data processing device. The sensory stimuli are somatosensory stimuli such as hearing, touch, smell, and taste, and multiple types of sensory stimuli may be output from one sensory stimulus output device, or different sensory stimuli may be output from multiple sensory stimulus output devices. The somatosensory stimuli output from the sensory stimulus output devices are stimuli related to the content of the VR video of the motor task displayed on the display device. For example, if the motor task is walking, the stimuli may be auditory stimuli such as walking sounds and environmental sounds linked to the VR video, or tactile stimuli that reproduce vibrations when contacting obstacles and vibrations during walking. If the motor task is falling, the stimuli may be auditory stimuli such as falling sounds linked to the VR video, or tactile stimuli that reproduce vibrations during falling. By comprising one or more sensory stimulus output devices, the physical function improvement system of the present invention can output complex sensory stimuli that combine visual stimuli from the display device and somatosensory stimuli from the sensory stimulus output devices, thereby further enhancing the subject's sense of immersion and further improving the efficiency of brain activity activation. The display device and the sensory stimulation output device may be integrated into a single device, or one or more sensory stimulation output devices may be combined. Examples of such combinations include a head-mounted display with integrated headphones and a controller with vibration functionality.

[0030] A VR application program that generates video signals and sensory stimulation signals for an exercise task may be stored in the memory of a data processing device. In one example, the data processing device is configured to output the video signals and sensory stimulation signals for an exercise task generated by the stored VR application program to a display device and a sensory stimulation output device. The physical function improvement system of the present invention may further include a storage medium for storing the VR application program. The storage medium may be, for example, ROM (Read Only Memory), DRAM (Dynamic Random Access Memory), SRAM (Static Random Access Memory), tape, disk, card, semiconductor memory, server, cloud, etc. The data processing device may also be configured to communicate with an external server or cloud to acquire information.

[0031] The calculation of statistical quantities (t-values) can be performed using a general linear model (GLM) and a sliding window that changes the analysis time window over time, applied to measurement data within a predetermined time range. The time window is set to a range (time width) that includes at least one task period. As new measurement data (measured values) are acquired and the time window moves, the statistical quantities (t-values) are calculated, and the statistical quantities (brain activity indicators) are fed back in real time.

[0032] In a general linear model, time t ｓ Measured value Y ｓ This is expressed by the following equation (1). Here, each function f represents a pre-assumed measurement model, β is the coefficient (regression coefficient) of function f, and ε is the error.

[0033] Equation (1) above is used when the time t included in the time window is １ ~t Ｎ Expressed as a determinant consisting of N measured values, it is given by equation (2) below. That is, Y = X・β + ε, where Y is the vector of measured values ​​in the time series included in the time window, β is the vector of partial regression coefficients, and ε is the error component. X is a matrix constructed by a linear combination of the model function f, and is called the design matrix.

[0034] The design matrix (matrix X) is represented in a manner that expresses the magnitude of the values ​​contained in the matrix using grayscale. The design matrix is ​​composed of multiple model function vectors. Specifically, the design matrix includes a zero-order function term, a task model function term, a rest model function term, and a first-order function term. In one embodiment, the design matrix includes a scalp blood flow function term.

[0035] The zero-order function term is a constant term for offsetting the measured value. The task model function term and the rest model function term are functions assumed to represent the measured waveforms during the task phase and rest phase, respectively. The task model function term is obtained by convolving the hemodynamic response function (HRF) into the boxcar function (a box-shaped function that takes a value of 1 or 0) corresponding to the task phase, and the rest model function term is obtained by convolving the hemodynamic response function into the boxcar function corresponding to the rest phase. The task model function term and the rest model function term may each consist of multiple functions (multiple column vectors). The linear function term consists of a linear function for correcting the drift component of the measured value.

[0036] The scalp blood flow function term is a term (column vector) used to correct for the scalp blood flow component in the measurement data. By incorporating this scalp blood flow function term into the design matrix, the statistical value (t-value) is calculated after correcting for the scalp blood flow component.

[0037] In one embodiment, the scalp blood flow function term is calculated using scalp blood flow data from multiple second measurement channels SC and the average time-series change of measurement data from all first measurement channels LC. In calculating the scalp blood flow function term, the data processing device uses the average time-series change (average waveform) of the measurement data to Z-score (normalize) each scalp blood flow data, and then performs principal component analysis on each normalized scalp blood flow data.

[0038] The Z-score is a value obtained by converting (normalizing) a set of time-series measurement data so that the mean is 0 and the standard deviation is 1. Let A be the mean of the mean time-series change (mean waveform) of each measurement data and S be the standard deviation, and the Z-score z(t) is obtained by the following formula (3). z(t) = (x - A) / S...(3)

[0039] By the above formula (3), the magnitudes of the measurement data and the scalp blood flow data are normalized. Thus, in the present embodiment, the data processing device normalizes the magnitude of the component of the scalp blood flow function term (the first principal component described later) based on the scalp blood flow data of the second measurement channel in the time window (predetermined time range) and the measurement data (mean waveform of the measurement data) of the first measurement channel in the time window.

[0040] Further, the data processing device calculates the first principal component obtained by performing principal component analysis on the scalp blood flow data of a plurality of second measurement channels as the scalp blood flow function term. That is, the data processing device incorporates the obtained first principal component into the design matrix as the scalp blood flow function term. In other words, the data processing device uses the obtained first principal component as an index of the scalp blood flow component included in the measurement data to correct the scalp blood flow component of the measurement data. The first principal component is an example of the correction amount of the scalp blood flow component.

[0041] When the design matrix is determined, the data processing device calculates a statistic (t-value) by performing statistical analysis of the general linear model shown in the above formula (2).

[0042] Specifically, the data processing device estimates the coefficient vector β based on the measured value Y (measurement data for a predetermined time) included in the time window and the design matrix X. As the measured value Y, a value standardized by the Z-score is used as in the case of scalp blood flow data. The estimation of the coefficient vector β is performed, for example, by fitting to obtain the value of β that minimizes the error ε by the least squares method. When the task model function term and the rest model function term are each one column (one column vector), the coefficient vector in the above equation (2) includes β0 corresponding to the zero-order function term, β1 corresponding to the task model function term (one column), β2 corresponding to the rest model function term (one column), β3 corresponding to the first-order function term, and β4 corresponding to the scalp blood flow function term.

[0043] The data processing device calculates the t-value of the value obtained by subtracting the estimated β value (β2) corresponding to the first column in the rest period from the estimated β value (β1) corresponding to the first column in the task period among the obtained coefficient vectors β, and records the obtained t-value as a statistic indicating the statistical significance of the change in the measurement data reflecting brain activity.

[0044] The calculation of the statistic (t-value) starts when the number of data (measurement data and scalp blood flow data for a predetermined time) for the time width of the time window is accumulated in the data processing device, and then, each time new data (measurement data and scalp blood flow data) is acquired, the time window is moved (slid) and implemented in real time. The moving timing of the time window can be delayed, if necessary, after one or more task periods end.

[0045] To facilitate the understanding of the present invention, examples are shown below to specifically describe the present invention, but it is needless to say that the present invention is not limited thereto.

[0046] (Example 1) A system integrating a head-mounted display (trade name HTC-vive pro, manufactured by HTC) and the NIRS-NFB system "NIRS Neuro Rehab System" developed by the present inventors so far was constructed, and 22 healthy subjects (31.1 ± 8.7 years old) were used as subjects to perform training for activating the bilateral supplementary motor areas using walking and falling tasks. A schematic diagram of the system is shown in FIG. 3.

[0047] Participants wore a NIRS brain activity measurement device and a head-mounted display, and performed motor imagery in response to intermittent presentations of 3D visual stimuli simulating walking and falling in a VR space. Each task consisted of approximately 5 seconds for the walking task and approximately 1 second for the falling task, with a rest period of 6-8 seconds in between, and was repeated 20-30 times. Participants first underwent VR neurofeedback training including 31 falling tasks, followed by a 2-3 minute interval, and then VR neurofeedback training including 19 walking tasks. During the interval, the devices were not removed, and it was checked for any problems or discomfort with the fit. A fixation point was set in the VR space, and participants performed motor imagery in response to tasks presented with 3D visual stimuli while fixating on that point. The results of the brain activity evaluation associated with motor imagery were presented in the VR space as disk size and color information, and participants trained to voluntarily control their brain activity based on this information.

[0048] After training, participants (N=22) were asked to rate their safety on a four-point scale regarding nausea and dizziness. As shown in Figure 4, there was almost no nausea or dizziness, confirming that training could be conducted safely without causing VR sickness symptoms. The vertical axis in Figure 4 shows the percentage of participants for each response.

[0049] To evaluate the improvement in postural control function of subjects (N=22) using this system, a standing task was performed on a swaying platform before and after the aforementioned training, and the distance of center of gravity movement in response to the swaying was evaluated as an indicator of postural control ability. The platform swayed intermittently back and forth at a speed of 4 cm and a maximum of approximately 20 cm / second, with each back-and-forth sway lasting 0.5 to 1 second. Rapid back-and-forth sways occurred without warning at intervals of approximately 8 to 12 seconds, and the distance of center of gravity movement from 0.5 seconds before to 2.5 seconds after the swaying was used as an indicator of postural maintenance ability in response to the swaying. Subjects were instructed to maintain an upright position in response to the swaying, and maintained an upright position on a horizontal platform with their legs spread approximately shoulder-width apart, while looking at a fixation point approximately 1 m in front. The center of gravity during standing was measured using a foot pressure distribution measuring device on the platform. As a result, as shown in Figure 5, the distance of center of gravity movement in response to swaying significantly improved before and after training (p<0.01). The vertical axis in Figure 5 shows the center of gravity displacement distance (cm). In addition, the activity of the bilateral supplementary motor area associated with the walking / falling task from the start of training until the end of the first quarter of trials, and the activity of the bilateral supplementary motor area associated with the task in the latter three-quarters of trials, were analyzed using brain activity data obtained with a NIRS device. Figure 6 is a visualization on a three-dimensional model of which parts of the brain are affected by the brain activity activation effect produced by training using this system. The vertical and horizontal axes in Figure 6 show the location of brain regions, and the bar on the right is a color chart that evaluates the brain activity activation effect on a scale of 0 to 10 based on the difference in the average values ​​of brain activity measured in the latter three-quarters of trials and the first quarter of trials, with circles indicating the activity activation effect for each measurement site. As shown in Figure 6, in both the falling task and the walking task, more active brain activity was observed in the latter three-quarters of trials compared to the first quarter of trials, and the results showed that task-related brain activity in brain regions, mainly the supplementary motor area, was activated. These findings indicate that this system can activate brain regions related to postural control, including walking, while the subject remains seated, without explicit movement and without relying on the subject's motor imagination ability, and that it also has an effect of improving postural control ability.

[0050] All patents, published patent applications, and teachings disclosed in reference herein are incorporated herein by reference in their entirety.

[0051] According to the physical function improvement system of the present invention, it is possible to strengthen a stable neural network without actual movement or dependence on motor imagination ability, and it can be used for minimally invasive fall prevention rehabilitation and training to safely improve postural control ability while seated.

Claims

1. A system for improving physical function, comprising: a brain activity measurement device for measuring the brain activity of a subject; a data processing device that calculates a statistical quantity indicating the statistical significance of changes in measurement data reflecting brain activity based on the measurement data acquired from the brain activity measurement device; and a display device that displays VR images of a motor task related to posture control and real-time brain activity indicators corresponding to the statistical quantity on the same screen.

2. The physical function improvement system according to claim 1, wherein the display device is a head-mounted display device, and the exercise task includes one or more tasks selected from the group consisting of standing, walking, falling, balance training, and high-intensity exercise.

3. The physical function improvement system according to claim 1, wherein the data processing device generates a VR image of an exercise task including the brain activity indicator, and the brain activity indicator is represented by virtual objects in the VR space of the exercise task.

4. The physical function improvement system according to claim 1, wherein the brain activity measurement device is an optical measurement device or an electroencephalogram (EEG) measurement device.

5. The physical function improvement system according to claim 1, further comprising one or more sensory stimulation output devices.

6. The physical function improvement system according to claim 1, wherein the playback time per session of the VR video of the exercise task is 0.5 to 60 seconds, the playback interval of the VR video of the exercise task is 1 to 60 seconds, and the number of times the VR video of the exercise task is played is 2 or more.

7. A program that causes a computer to function as a physical function improvement system according to any one of claims 1 to 6.

Images