JP2018139975A - Walking training device, walking diagnostic device, body weight relieving device, walking training method, and walking diagnostic method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable walking training that promotes gain of function by actively drawing patient's exercise recovery capability in robot therapy of stroke patients, etc.SOLUTION: A walking training device 1 includes a saddle 13 for supporting a subject in a crotch part. The walking training device also includes a body weight relieving part 10 employing a seesaw type mechanism for relieving a portion of a body weight of the subject, a treadmill 20 positioned below the saddle 13 and having a floor on which the subject walks, and an electrical stimulation part 50 for performing electrical stimulation to an ankle joint plantar flexion muscle of the subject at a start timing of a swing phase.SELECTED DRAWING: Figure 4

Description

  The present invention relates to a gait training device, a gait diagnosis device, a weight-bearing device, a gait training method, and a gait diagnosis method that support recovery of walking ability.

  In recent years, various gait training devices using robot therapy for the lower limbs have been proposed for the purpose of restoring the walking ability of stroke patients and spinal cord injury patients. Among them, in particular, Body-weight-supported treadmilltraining (BWSTT), in which walking training is performed on a treadmill while releasing a part of the body weight, has attracted attention (Patent Document 1, Non-Patent Document 1).

  By the way, with the rapid aging of the population today, there are many expectations for science and technology that support people's health promotion. Proper support requires the right theory and the scientific evidence to support it. In the control of a body movement system that is extremely redundant and multi-degrees of dynamics, how the central nervous system manipulates the degrees of freedom and determines the muscle activity pattern, and what are the control rules? This is a classic open problem in the motion control domain. The present inventors have analyzed human voluntary movement by paying attention to physical coordination based on the co-activity of muscle pairs. More specifically, the present inventors have so far extracted, as a motion analysis device, a motion analysis method, and a motion analysis program, the coordination of body motion from the myoelectric potential information, and the motion control of the body extremity, for example, the tip of the upper limb. Has invented a motion analysis technology that estimates the features (muscle synergy, equilibrium point, impedance (rigid ellipse)) related to humans, and has contributed to elucidation of the motion generation mechanism of human voluntary motion and human motion support by robots (patents) Document 2, Non-Patent Document 2). These technologies are extremely useful tools for exploring effective interventions while properly identifying correct and incorrect interventions and establishing meaningful methods.

US 8684890 B2, Apr. 16, 2010. (Application date) JP2015-112453A

T. Susko, K. Swaminathan, and HI Krebs, “MIT-Skywalker: A Novel Gait Neurorehabilitation Robot for Strokeand Cerebral Palsy,” IEEE Transactions on NeuralSystems and Rehabilitation Engineering, vol. 24, no.10, pp. 1089-1099, 2016. H. Hirai, F. Miyazaki, H. Naritomi, K. Koba, T. Oku, K. Uno, M. Uemura, T. Nishi, M. Kageyama and HIKrebs, “On the Origin of Muscle Synergies: InvariantBalance in the Co-activation of Agonist and Antagonist Muscle Pairs, ”Frontiers in Bioengineering and Biotechnology, vol.3, no.192, pp. 1-16, 2015.

  However, the current lower-limb robot therapy has been reported to suggest good training effects (restoration of balance function and motor function, walking speed, durability, improvement of symmetry, etc.), but end effector type and exoskeleton type In any of these forms, the therapeutic effect over the usual manual treatment has not been clarified, and it is required to search for effective intervention methods and accumulate scientific evidence.

  The present invention has been made in view of the above, and has developed the above-described new motion analysis technology to robot therapy for stroke patients and the like (currently in the early stages of robot therapy for the lower limbs). It is an object of the present invention to provide a walking motion diagnosis device for elucidating the above, a walking training device that actively extracts the motion recovery ability of a patient and promotes the acquisition of the function, and a weight-bearing device used therefor.

  The present invention proposes a gait training device that combines weight relief for manipulating the interaction between the floor and the end effector (foot) as an intervention and functional electrical stimulation acting on the ankle joint adjacent to the end effector. That is, the walking training apparatus according to the present invention includes a saddle that supports the subject at the crotch, and is located below the saddle, a weight-loading portion that unloads part of the subject's weight, and the subject A floor portion having a floor surface on which the person walks and an electrical stimulation unit that performs electrical stimulation to the ankle plantar flexor muscle of the subject are provided.

  Further, the walking training method according to the present invention is the floor of the floor portion in a state in which the subject is supported on the crotch by the saddle of the weight-loading portion on the floor surface of the floor portion and a part of the weight is unloaded. Electrical stimulation is performed on the ankle plantar flexor muscle of the subject while moving the surface from the front to the rear of the subject indicated by the saddle.

  In human gait, the hip, knee and ankle work account for more than 80% of the hip and ankle joints (DJ Farris and GS Sawicki, “Themechanics and energetics of human walking and running: a joint levelperspective, ”Journal of the Royal Society Interface, vol. 9, no. 66, pp. 110-118, 2012)). In particular, it is known to perform large tasks during hip extension and ankle push-off. According to the present invention, the combination of the weight-bearing portion serving as the first intervention and the floor portion having the floor surface for walking allows balance control and the function of hip joint extension, and the ankle joint sole serving as the second intervention. The electrical stimulation of the flexor muscles can support and train the push-off function by the ankle joint. Note that the floor surface is not limited to a horizontal plane, and may be set to an inclined surface, which makes it possible to support training for inclined walking.

  The electrical stimulation unit includes an electrode attached to the surface of the subject's ankle plantar flexor muscle, a stimulation signal generation unit that applies a stimulation signal to the electrode, and an instruction unit that instructs the application of the stimulation signal. It is equipped with. According to this configuration, since the stimulation signal is applied to the surface of the subject's ankle plantar flexor muscle in accordance with an instruction from the instruction unit, the push-off function by the ankle joint is executed. Note that the instruction from the instruction unit may be performed by the subject himself or by detecting that the leg to be trained is the start timing of the swing phase and automatically instructing according to the detection result. .

  In addition, the present invention provides a detection unit that detects a myoelectric potential of each muscle in the lower limb of the subject and a movement of each joint of the lower limb of the subject, and a feature amount related to motion control of the lower limb tip from the detection result by the detection unit And an arithmetic unit for calculating According to this configuration, by detecting the myoelectric potential of each muscle of the subject's lower limbs and the movement of each joint of the subject's lower limbs, a feature amount for gait diagnosis can be obtained for efficient walking training. Provided.

  The detection unit includes a myoelectric potential detection unit that measures myoelectric potential as the muscle activity, and a movement detection unit that detects joint movement as the lower limb movement, and the calculation unit includes: At least one of muscle synergy, equilibrium point, and rigidity is calculated as the feature amount from the detection result and the detection result of the motion detection unit. According to this configuration, at least one of muscle synergy, equilibrium point and stiffness ellipse related to the translational motion of the ankle joint, and equilibrium point and stiffness related to the rotational motion of the ankle joint are calculated as effective feature quantities for the determination of the gait diagnosis. Is done. Therefore, it is possible to determine the recovery status of the lower limb from the information at the time of walking by referring to the feature amount.

  Further, the floor portion is arranged in parallel with each other and is laid around a horizontally supported rotating body so as to be able to circulate, and the upper surface side constitutes the endless belt, and the floor surface is provided with the endless belt. A load detection unit that detects the stepping position of the sole of the subject and the floor reaction force is provided. According to this configuration, the target person can realize relative walking in a state where a part of the body weight is released from the body weight-unloading unit. Further, the stepping position and the floor reaction force can be obtained as information on the walking situation, and can be used for diagnostic information and other purposes (games that also serve as training).

  Further, the present invention is a video projection unit that projects a training video on the floor surface of the floor unit, a load detection unit that detects a stepping position and a floor reaction force of the subject's foot on the floor surface, A training evaluation unit that evaluates training from the display position of the training video on the floor and the stepping position detected by one of the motion detection unit and the load detection unit is provided. According to this configuration, the training video is projected on the floor to be walked, and the evaluation is performed based on the correlation with the training video (such as the degree of coincidence). .

  The weight loader includes a base, an arm that is pivotally supported by the base and swings around the pivot within a vertical plane, and a weight, and the saddle is attached to a tip of the arm. The weight is suspended from the base end side of the arm. According to this configuration, by adopting a so-called seesaw type mechanism and counterweighting, it becomes possible to balance with a part of the weight of the subject straddling the saddle with a simple configuration.

  Further, the present invention includes a weight engaging portion that vertically suspends the weight on the arm, and the weight is formed by stacking a plurality of weights on the base. A plurality of weights are engaged with the base end of the arm. According to this configuration, the counterweight weight is easily set in the seesaw type mechanism.

  The gait diagnostic apparatus according to the present invention includes a saddle that supports the subject at the crotch, is located below the saddle, and a weight-bearing portion that unloads a part of the subject's weight. Motion diagnosis information from the detection result of the floor part having the floor surface on which the person walks, the detection unit for detecting the myoelectric potential of each muscle in the lower limb of the subject and the movement of each joint of the lower limb of the subject A calculating unit that calculates a myoelectric potential as the muscle activity, and a movement detecting unit that detects a joint movement as the lower limb movement, the calculating unit comprising: From the detection result of the myoelectric potential detection unit and the detection result of the motion detection unit, at least one of muscle synergy, equilibrium point, and rigidity, which is a feature amount related to the motion control of the lower limb tip, is calculated as a feature amount. .

  The gait diagnosis method according to the present invention includes a saddle that supports the subject at the crotch, and is located below the saddle, and a weight relief portion that unloads part of the subject's weight. Motion diagnosis information from the detection result of the floor part having the floor surface on which the person walks, the detection unit for detecting the myoelectric potential of each muscle in the lower limb of the subject and the movement of each joint of the lower limb of the subject A calculation unit that calculates a muscle potential as the muscle activity, and detects joint movement as the lower limb movement, and the calculation unit detects the myoelectric potential detection result, and As a lower limb movement, at least one of muscle synergy, equilibrium point, and rigidity, which is a characteristic amount related to the motion control of the lower limb tip, is calculated as a feature amount from the detection result of joint movement.

  The effects of walking training according to these inventions are quantified based on at least one of muscle synergy, equilibrium point, and rigidity (which may include both translational and rotational motions), which are characteristic amounts related to motion control of the lower limb tip. Evaluated and fed back to patients and therapists.

  In the present invention, rehabilitation such as stroke is regarded as functional recovery of the central nervous system based on physical cooperation, and robot training for gait reconstruction that promotes reacquisition and maintenance of physical cooperation of the patient is performed. In particular, (1) Diagnose the patient's physical coordination using a series of muscle synergy technologies revealed by the inventors' previous research (Patent Document 2, Non-Patent Document 2, etc.), and (2) the diagnosis Based on the results, it is characterized by several types of robot training that promotes motion improvement. The skill of coordinating the body is indispensable for realizing the smooth movement (adjustment of equilibrium point and rigidity (impedance)) that is apt to be lost in stroke patients, and the acquisition leads to the dramatic improvement of the patient's motor ability.

  In addition, a weight loader according to the present invention includes a saddle that supports a subject at the crotch, and a part of the subject's weight is unloaded. An arm that is pivotally supported by the base and swings in a vertical plane; a weight; and a weight engaging portion that suspends the weight from the arm; and the saddle is attached to a tip of the arm, A plurality of weights are stacked on the base, and the weight engaging portion engages an arbitrary plurality of weights with the base end of the arm from above. According to the present invention, by adopting a so-called seesaw type mechanism and counterweighting, it is possible to balance with a part of the weight of the subject straddling the saddle with a simple configuration. In addition, it is easy to set the weight for the counterweight.

  According to the present invention, a new motion analysis technique is developed for robot therapy for stroke patients and the like (currently the robot therapy for the lower limbs in the early stages), and the walking motion toward the elucidation of the motor learning mechanism after the disorder is developed. Diagnosis, enabling gait training to actively acquire the ability of the patient to recover and promote function acquisition.

BRIEF DESCRIPTION OF THE DRAWINGS It is a whole schematic diagram which shows one Embodiment of the walking training apparatus which concerns on this invention, (a) is the external appearance schematic diagram seen from back diagonally, (b), (c) is the training image | video projected on the floor surface of a floor part It is a top view which shows each example of. FIG. 2 is an overall schematic diagram of the walking training apparatus shown in FIG. 1, where (a) is an external schematic diagram viewed from an oblique front, (b) is a side schematic diagram of the floor, and (c) is a plan view of the floor. It is an external view of a saddle, (a) is a perspective view explaining a structure, (b) is the schematic of the planar view explaining a yawing. It is a functional lineblock diagram showing one embodiment of a walking training device. It is explanatory drawing of the member attached to a leg, (a) is a test muscle, (b) is a figure which shows an example of the attachment position of a marker, a detection electrode, and a stimulation electrode. It is a figure which shows the coordinate system (The hip joint angle | corner (theta) h, the knee joint angle | corner (theta) k, the foot joint angle | corner (theta) a, the radial radius R, the deflection angle (PHI)) employ | adopted with the musculoskeletal model of a leg. It is a flowchart which shows the walk analysis process in the walk training which a control part performs. It is a figure which shows the example of a display which displays a motion analysis result on a monitor in real time, (a) is a display example of a muscle synergy, (b) is a display example of the rigid ellipse and the equilibrium point regarding the translational movement of the ankle joint in a gait phase, (C) is a display example of the rigidity related to the rotational motion of the ankle joint. It is a figure which shows the change of the gait by the difference in an unloading system and the presence or absence of an electrical stimulus. It is a figure which shows the measurement result of the vertical treading force in 1 walk by the difference in an unloading system, and an ankle joint angle. 6 is a chart showing a list of experimental conditions in Experiment 2. It is a table | surface which shows the change of the walking speed felt comfortable according to the amount of loads of a harness support type body weight load apparatus and a saddle support type body weight load apparatus. Shows the gait phase (a) for one gait of subject A (horizontal axis 0-100%), and hip joint (b), knee (c), and ankle joint (d) joint angle changes due to different unloading methods. FIG. It is a figure which shows the perpendicular | vertical component (a) of the floor reaction force in 1 walk of the test subject A by the difference in an unloading system, the front-back shear component (b), and the left-right shear component (c). It is a figure which shows the perpendicular | vertical component (a) of the floor reaction force in one walk of the test subject A by the difference in walking speed in non-weight-free walking, the front-rear shear component (b), and the left-right shear component (c). The figure which shows each joint angle change of the crotch (a), the knee (b), and the ankle joint (c) in one walk of the test subject A by the walking speed in the saddle support type weight-free walking (medium weight-free). It is. Vertical component (a), front-rear shear component (b), left-right shear component (c) of floor reaction force in one walk of subject A based on walking speed in saddle-supported weight-free walking (medium weight-bearing) It is a figure which shows the change of. It is a figure which shows each joint angle change of the crotch (a), the knee (b), and the ankle joint (c) in 1 walk of the test subject A by the load-carrying amount in saddle support type body weight-free walking. It is a figure which shows the perpendicular | vertical component (a) of the floor reaction force in one walk of the test subject A by the amount of load in the saddle support type body weight-free walk, the front-rear shear component (b), and the left-right shear component (c). Harness-supported and saddle-supported weight-free walking (moderate weight-loading) and non-weight-free walking, and the vertical component of floor reaction force in one walk of subjects A, B, and C with or without FES intervention ( It is a figure which shows a) front-back shear component (b) and right-and-left shear component (c).

  FIG. 1 is an overall schematic diagram showing an embodiment of a walking training apparatus 1 according to the present invention, and FIG. 2 is an external schematic diagram showing a part thereof. FIG. 3 is an external view of the saddle. FIG. 4 is a functional configuration diagram showing an embodiment of the walking training apparatus 1.

  The walking training device 1 according to the present invention is a device for robot therapy of the lower limbs for the purpose of recovering the walking ability of, for example, a stroke patient or a spinal cord injury patient (hereinafter referred to as a subject). The gait training device 1 includes a weight-loading unit 10 that unloads a part of the subject's weight, a treadmill 20 that is an example of a floor portion on which the subject walks, a motion detection unit 30, and a myoelectric potential detection unit. 40, an electrical stimulation unit 50 that applies electrical stimulation to muscles acting on the ankle joint, a control unit 60 (see FIG. 4), a motion observation unit 70 that captures and displays a subject, and a leg motion is induced. Such a game or the like is provided with a training processing unit 80 that displays a training video and evaluates the training result as necessary.

  In this embodiment, the weight loader 10 is a “saddle-supported weight loader device” ((Yuma Nagakawa, Fumiyoshi Yoshikawa, Hiroaki Hirai, Satoshi Kuroiwa, Hidetomo Watanabe, Mitsunori Uemura, Fumio Miyazaki, “Analysis of Saddle-supported Weight-Unloading Treadmill Walking: Visualization of Lower Limb Equilibrium Trajectory and Toe Stiffness”, 10th Motor Control Study Group Abstract, A36, (2016.), and (Fumiaki Yoshikawa, Yuma Nagakawa, Hiroaki Hirai, Satoshi Kuroiwa, Hidenori Watanabe, Mitsunori Uemura, Fumio Miyazaki, “Effects of weight-loading methods on walking: lifting and saddle-supporting “Comparison”, 10th Motor Control Study Group Abstract, A37, 2016.)).

  More specifically, as shown in FIGS. 1 (a) and 2 (a), the weight loader 10 has a base 11 installed on the floor and a left-right direction (see FIG. 1) on the top of the base 11. (A) an arm 12 that is supported by a horizontal shaft 121 parallel to the arrow XX direction and can swing within a vertical plane, a saddle 13 attached to the tip of the arm 12, and a base end of the arm 12 And a plurality of weights 16 stacked and mounted on the lower portion of the frame portion 15. The direction from the base end to the tip end of the arm 12 is the forward direction.

  The frame part 15 has a pair of upright stays 151 parallel to each other. The weight 16 has a flat plate shape in the present embodiment, and is provided with an insertion hole into which the weight engaging portion 14 is inserted and a pair of holes in which the upright stay 151 is loosely fitted. The weight 16 can be moved up and down along the upright stay 151 and is moved up and down in accordance with the swing of the arm 12 by being engaged with the weight engaging portion 14.

  Various forms of the engagement structure between the weight engaging portion 14 and the weight 16 can be adopted. In the present embodiment, the following forms are adopted. That is, the weight engaging portion 14 is configured by connecting a deformable string-like object having a predetermined length, for example, a chain portion 141 and a rod-like body 142 below the chain portion 141. The rod-like body 142 is formed in the length direction. A long hole (not shown) is formed. A pin 143 is inserted into the long hole. For example, a lateral groove 161 is formed on the lower surface side of each weight 16 so that the pin 143 can be inserted from the lateral direction. The pin 143 is inserted along the lateral groove and is inserted into the long hole. The upper weight 16 including the weight 16 into which the pin 143 is inserted is lifted by the pin 143 to be used as a counterweight. By adjusting the number of weights 16 to be counterweighted, the amount of unloading with respect to the subject who straddles the saddle 13 via the seesaw type mechanism is adjusted. In addition, as another example of the weight engaging portion, a configuration in which a bottom plate is provided and weights are stacked one by one on the top may be used, or each counterweight prepared in advance may be replaced and used. Alternatively, a mode in which a weight balance is achieved by suspending a weight on the rear end side of the arm 12 and sliding the arm 12 in the longitudinal direction may be used.

  2 and 3, the saddle 13 includes a saddle body 13A having a seat shape, and a skin 13B made of an elastic material that covers the saddle body 13A provided as necessary. The saddle 13 includes a horizontal shaft 131 parallel to the front-rear direction provided on the fixture 122 at the tip of the arm 12, and a vertical shaft 132 attached via the horizontal shaft 131. The saddle body 13A includes the vertical shaft 132. Is attached through. With this configuration, the saddle 13 rolls around the horizontal axis 131 as shown by an arrow Roll in FIG. 3A and vertically as shown by an arrow Yaw in FIG. Yawing around axis 132. As the saddle 13 rolls and yaws, a more natural walking motion can be provided to the subject.

  In this embodiment, the treadmill 20 employs a split belt treadmill (ITR5018, Bertec Corp., USA) incorporating a floor reaction force meter. More specifically, the treadmill 20 includes a base 21 having a rectangular parallelepiped shape. In the front-rear direction of the base 21, as shown in FIGS. 2 (b) and 2 (c), a pair of rollers 22 having a horizontal rotation axis parallel to each other, and a belt 23 stretched between the pair of rollers 22. And a driving unit 24 such as a motor connected to one roller 22 via a driving force transmission structure 241 such as a belt. By driving the drive unit 24, the belt 23 rotates in the direction indicated by the arrows in FIGS. Conditions for the rotation operation of the belt 23 are appropriately set by the setting unit 240 (see FIG. 4). As shown in FIG. 2 (b), the subject walks on a belt 23 that is rotating or intermittently operated along with walking. The upper surface of the belt 23 constitutes a walking surface and is horizontal in this embodiment.

  The load detection unit 25 includes a sensor main body 251 and a floor reaction force calculation unit 252. The sensor main body 251 is disposed between the front and rear rollers 22 and in the vicinity of the upper part on the inner side of the belt 23 (see FIG. 2B). 2B and 2C, the sensor main body 251 includes a rectangular parallelepiped force plate 2511 and load sensors (mechanical-electrical devices) arranged at a plurality of predetermined positions, for example, four corners of the force plate 2511. A load cell 2512 which is a conversion element), and a change in the resistance value of the strain gauge corresponding to the load applied on the force plate 2511 is detected as a change in the current value by each load sensor 2512. The three-way component force (Fx, Fy, Fz) is measured. Further, the floor reaction force calculation unit 252 calculates a floor reaction force action point (COP: Center of Pressure). Accordingly, the floor detection force and the center of gravity position of the foot pedal force of the subject who has received a part of his / her weight by applying the crotch portion to the saddle 13 can be obtained by the load detection unit 25 via the belt 23.

  In addition, the load detection part 25 may be arrange | positioned corresponding to each belt 23 in the case of a split belt treadmill type. In this structure, it becomes possible to perform the walking training of the subject at a method that matches the degree of recovery of each leg, for example, at different lap speeds with the left and right belts. Moreover, the aspect which can obtain the action point of floor reaction force at least may be sufficient as the load detection part 25. FIG. Further, a method of measuring the load position by using another pressure-sensitive member may be used. The treadmill 20 may not be a split belt treadmill type.

  By combining the treadmill 20 and the weight loader 10 in this manner, while interfering with the end effector (foot) of the subject while reducing the influence of gravity due to body weight, the balance control and hip joint extension are suitably supported. You can train while doing.

  As shown in FIG. 5 (a), the test muscles of the lower limbs that are the target site of the motion analysis are, in this example, the 3-6 muscles of the lower limbs related to translation of the ankle position, Modeled as a 4 to 8 musculoskeletal system with 1 to 2 muscles added.

  In the present embodiment, an optical tracking system such as a motion capture in which cameras 31 as imaging means are arranged on the left and right sides is employed as the motion detection unit 30. The left and right cameras 31 image the subject's lower limbs from the left and right sides. In this embodiment, as shown in FIG. 5 (b), markers 301 are attached to a total of four positions on each side of the subject's hip joint, knee joint, ankle joint, and toe as shown in FIG. 5B. Or it is installed. The marker 301 may employ a specific color, for example, a member that emits far infrared light, or a light emitting element. The camera 31 images each marker 301 as a bright spot representing position information within the camera field of view. The markers 301 corresponding to the subject's hip joint and ankle joint are used to measure radius R information and declination Φ information (see FIG. 6), which will be described later. The number of cameras 31 is not limited to two, and may be arranged corresponding to each marker 301, for example.

  As shown in FIG. 5B, the myoelectric potential detection unit 40 detects myoelectric potentials of eight muscles per one foot. A plurality of electrodes 401 are attached on the skin of the lower limbs of the subject. As shown in FIG. 5B, each electrode 401 is affixed to each surface side of eight muscles and detects myoelectric potentials generated in the corresponding muscles. The myoelectric potential is obtained by EMG (electromyography that directly measures muscle activity). Each electrode 401 is connected to the myoelectric potential detection circuit 41 by wire or wirelessly. In the wirelessly connected mode, the electrode 401 includes at least an antenna unit that performs near field communication with the antenna unit of the myoelectric potential detection circuit 41.

  The electrical stimulation unit 50 acts on the ankle joint of the subject, and specifically performs functional electrical stimulation for changing the ankle joint angle at the start of the swing leg in the plantar flexion direction. The electrical stimulation unit 50 includes a stimulation signal generation unit 51 and a switch 52 that instructs output of the stimulation signal. The stimulation signal electrically stimulates the ankle plantar flexor muscle group based on an optional cue (pressing of the switch 52) of the subject for the purpose of supporting push-off by the ankle joint. The stimulation signal is a pulse carrier wave of 60 Hz, for example, and is set to an output time of, for example, 0.35 seconds at a level of, for example, 7 mA to the outer head of the gastrocnemius, for example, 8 mA to the inner head. This can be expected to lead to increased push-off and voluntary movement. The switch 52 uses the load detection information from the load detection unit 25, the motion detection content from the motion detection unit 30, or the motion observation content from the motion observation unit 70 described below, in addition to a mode of being pressed by the subject himself / herself. The swing leg start timing may be automatically recognized and switched on.

  The motion observation unit 70 includes a Kinect sensor (Kinect for Windows v2, Microsoft Corp., USA) 71 that captures the walking motion of the subject from the front side in front of the treadmill 20, and further displays a captured image. 711 is provided forward. Moreover, you may provide the left and right imaging parts 72 and 73 and the left and right monitors 721 and 731 which display a left and right image as needed. As is well known, the Kinect sensor 71 obtains the three-dimensional information of the image of the subject imaged by using the captured image and the laser beam together, and displays the result together with the captured image on the monitor 711, thereby walking. The kinematics information about the left-right balance of the subject is presented. Instead of the Kinect sensor 71, a normal imaging unit may be used.

  The training processing unit 80 is provided as necessary, and generates and displays an image that supports the walking of the subject. The training processing unit 80 includes a training image generation unit 81 and a projector 82 that projects the generated training image onto the belt 23. Further, the training processing unit 80 may be provided in front of or in front of the belt 23 as necessary so that the subject can easily see the training image. The training image is displayed on the monitor 83 arranged obliquely forward. The training image generation unit 81 may be configured to generate the training image so as to move rearward of the belt in synchronization with the moving speed of the belt 23 or to display the training image at a fixed position on the belt 23.

  The training processing unit 80 includes a responsiveness determination unit 84. The responsiveness determination unit 84 determines that the subject in a state straddling the saddle 13 acts on the training image from the training image, the circular motion information from the driving unit 24, and the action point information from the floor reaction force calculation unit 252. The results are also evaluated. Timekeeping information until a response may be included in the evaluation item. Further, the motion detection unit 30 may detect the action on the training image.

  FIGS. 1B and 1C show examples of training images on the belt 23 projected from the projector 82. FIG. 1B shows, for example, a plurality of circular images 801 of a predetermined shape arranged on the surface of the belt 23 slightly forward of the position just below the saddle 13, and one of them can be identified, for example, a color different from the others. A difference circular image 802 is displayed to indicate the next step position. In this state, the evaluation is set high or low according to the degree to which the action point obtained by the floor reaction force calculation unit 252 matches the display area of the training image, and is used for training of discrete motion (stepping). In FIG. 1 (c), for example, a bar image 803 is displayed alternately on the left and right sides of the belt 23 surface slightly below the saddle 13 and the front bar image 803 indicates the next step position. Yes. In this state, a higher evaluation is generated such that the action point obtained by the floor reaction force calculation unit 252 matches the display area of the bar image 803, and is used for periodic exercise training. In the evaluation, the score is integrated, and the integrated value or a score in time unit is calculated and displayed. By introducing a game sensation, the willingness to train can be increased.

  Next, the functional configuration unit in FIG. 4 will be described. The control unit 60 is connected to each unit of the gait training apparatus 1 to control and control the operation of the gait training apparatus 1 and calculates feature points including muscle synergy information and performs diagnostic processing. The control unit 60 includes a microcomputer having a CPU. The ROM 61 stores a processing program for training operation and motion analysis processing according to the present invention, a RAM 62 temporarily stores data during processing, An operation unit 63 including a numeric keypad and a mouse for giving instructions and a storage unit 64 are provided. The storage unit 64 stores information on the human musculoskeletal model (see FIG. 5A), each arithmetic expression for motion diagnosis, and other data. The length of each part of the human musculoskeletal model can be calculated by motion capture. The human musculoskeletal model information and each arithmetic expression may be stored in the ROM 61 or the RAM 62.

  The myoelectric potential detection unit 40 includes a myoelectric potential detection circuit 41 that detects electrical signals generated on the eight electrodes 401 per leg, and a preprocessing circuit 42 that performs predetermined preprocessing on the detected electrical signals. The myoelectric signal obtained on the body surface is an AC signal having a level of several tens of μV to several hundreds of μV and a frequency of about 5 Hz to 500 Hz. Therefore, the preprocessing circuit 42 includes an amplifier that amplifies the myoelectric potential to a level that can be processed (several thousand times), a band-pass filter that passes only signals in the main frequency band of the myoelectric potential, and a full-wave rectifier circuit. . The preprocessing circuit 42 includes an AD conversion unit on the output side in order to digitally process the myoelectric potential signal.

  The motion detection unit 30 includes each camera 31 that detects each marker 301 and a preprocessing circuit 32 that detects a bright spot from an imaging signal from the camera 31, and measures the position coordinate information of the detected bright spot. The predetermined pretreatment is performed. In addition to the optical position measurement, the motion detection unit 30 may employ a known measuring device that can detect a three-dimensional position and orientation including a magnetic generator and a magnetic sensor.

  The images created by the display processing unit 605 are displayed on the monitors 711 to 731. The display processing unit 605 outputs the input information from the operation unit 63, the processing result, the captured image by the Kinect sensor 71, and the motion diagnosis information processed by the control unit 60 to the monitor 711. The video imaged at 73 is output to the monitors 721 and 731. In addition, it is good also as an aspect which displays each information on a separate monitor as needed.

  The control unit 60 performs exercise diagnosis of the subject. By executing a processing program read from the ROM 61 to the RAM 62, the control unit 60 periodically reads the measurement result from the electrode 401, and the marker 301. The position measurement processing unit 602 that periodically captures the measurement results, the muscle synergy calculation unit 603, the equilibrium point / rigidity calculation unit 604, and the display processing unit 605 described above.

  According to the motion diagnosis according to the present invention, human body motion is generated by a cooperative action (muscle synergy) of a plurality of muscles, and the human body motion is controlled by a command related to a joint equilibrium point and stiffness. Is based on thinking.

  The myoelectric potential measurement processing unit 601 obtains EMG (myoelectric potential) of each antagonistic muscle pair obtained by the preprocessing circuit 42 as% MVC, which is EMG normalized by EMG at the time of maximum voluntary contraction. Myoelectric potential introduces, as a concept, a ratio of activities of antagonistic muscles (muscle antagonistic ratio r) considered as a minimum unit of muscle coordination and a sum of activities of antagonistic muscles (muscle antagonistic sum s). .

The muscle antagonist ratio r is expressed as% MVC of the main muscle / (% MVC of the main muscle +% MVC of the antagonist muscle), and the muscle antagonistic sum s is expressed as% MVC of the main muscle +% MVC of the antagonist muscle. As shown in FIG. 5 (a), there are antagonistic muscle pairs (4 to 8 muscles) around the hip (h), knee (k), and ankle joint (a) from the waist to the ankle joint. muscle antagonistic ratio _{(r h, r hk, r} k, r a) and muscle antagonizing sum _{(s h, s hk, s} k, s a) is calculated from each myoelectric potential. Equation (1) represents a muscle antagonist ratio and a muscle antagonist sum which are explanatory variables for translational movement of the ankle joint.

  Further, the muscle antagonist ratio and the muscle antagonist sum, which are explanatory variables for the rotational motion of the ankle joint, are shown in Equation (2).

The calculation method of the muscle synergy vector using these explanatory variables is the synergy u _R for adjusting the toe equilibrium point in the radial direction, the synergy u _Φ for adjusting the toe equilibrium point in the declination direction, and the toe rigidity. Use synergy u _{R × Φ} to adjust. Equation (3) shows a calculation formula for each synergy u _R , u _Φ , and u _{R × Φ} .

Incidentally, _{q h} in Equation (3) _{(s trans),} and q _{k (s trans)} is expressed by Equation (4).

Furthermore, the muscle synergy activity coefficient Δw _R (r _trans , s _trans ) and Δw _Φ (r _trans , s _trans ) are obtained from the equation (5). By multiplying the obtained muscle synergy activity coefficient Δw _R (r _trans , s _trans ) and Δw _Φ (r _trans , s _trans ) by the gain k _{R, trans} , k _{Φ, trans} , equation (6) is obtained. As shown, the amount of change ΔR _EP (r _trans , s _trans ) and ΔΦ _EP (r _trans , s _trans ) of the equilibrium point position related to the translational motion of the ankle joint were obtained.

Further, the stiffness matrix K _x (s _trans ) for the translational motion of the ankle joint is calculated using s _trans , the Jacobian matrix J (θ), and the gains k _{s and trans} as shown in the equation (7).

Further, estimation (analysis) regarding the rotational motion of the ankle joint is executed. That is, from the muscle antagonist ratio r _rot (= r _a ) of the antagonist muscle pair around the ankle joint, the displacement of the equilibrium angle of the ankle joint is calculated as in Expression (8). Further, the rotational stiffness of the ankle joint is calculated as in equation (9) from the muscle antagonist sum s _rot (= s _a ) of the antagonistic muscle pair around the ankle joint.

  Each information of muscle synergy, equilibrium point, and rigidity (rigid ellipse, rotational rigidity), which is the calculated feature amount, is displayed on the monitor 711 or another monitor. FIG. 8 shows an example of the display, in which (a) shows muscle synergy, (b) shows the rigidity ellipse 91 and the equilibrium point 92, and (c) shows the rotational rigidity 93 of the ankle joint. The analysis contents of walking training can be visually confirmed and confirmed in real time, and recovery training is effectively performed.

  FIG. 7 is a flowchart showing the procedure of the motion diagnosis process. With the subject straddling the saddle 13, the belt 23 starts rotating, and this flow is started. First, the position signal of the marker 301 and the myoelectric potential signal of the electrode 401 are fetched from the motion detection unit 30 and the myoelectric potential detection unit 40 (step S11), and the measurement process of the movement position and the myoelectric potential measurement process are performed from these signals. It is executed (step S13).

  When the detection signals from the respective measurement points are taken in this way, the muscle synergy vector is calculated by the equation (3) (step S15), and further, the translational motion of the ankle joint is calculated by the equations (6) and (7). The equilibrium point and stiffness related to the rotational motion of the ankle joint are calculated from the equilibrium point, the stiffness ellipse, and the equations (8) and (9) (step S17). These calculated feature values are output to the display processing unit 605 (step S19) and displayed on the monitor 711. Then, it is determined whether or not the training is finished (step S21). Until the training is finished, the process returns to step S11 and the process is repeated.

  Next, Experiment 1 and Experiment 2 for confirming the effect, and the results thereof will be described.

<Experiment 1>
(1-1) Equipment (weight free treadmill)
As a weight-loading device, (A) a lifting type weight-loading device (New Assist, Interliha Corporation, Japan) as a comparative example that lifts the upper body of the subject by a pneumatic cylinder and harness installed on the ceiling, (B) Two types of “saddle-supporting weight-bearing device” acting on the subject's crotch by a counterweight and a seesaw type mechanism shown in FIGS. 1 to 3 were used. As the treadmill, a split belt treadmill (ITR5018, Bertec Corp., USA) incorporating a floor reaction force meter was used. By combining these weight-bearing devices and a split belt treadmill, while interfering with the end effector (foot) of the subject while reducing the influence of gravity due to weight, "balance control" and "hip extension" are safe I started to support training. Furthermore, Kinect Sensor 71 (Kinect for Windows v2, Microsoft Corp., USA) and a large display (REGZA 65J7, Toshiba, Japan) as a monitor 711 are deployed in front of the subject, and kinematics regarding left-right balance during walking. Information was presented (see FIG. 1A).

(1-2) Equipment (functional electrical stimulation)
As another intervention, functional electrical stimulation acting on the ankle joint adjacent to the end effector (foot) was employed. As the electrical stimulation unit 50, STG4008 (Multi Channel Systems, Inc., Germany) was used. Here, for the purpose of supporting / training “Push-off by ankle joint”, the ankle plantar flexor muscles were electrically stimulated based on an optional cue of the subject (pressing the switch 52). Stimulation was performed by applying a pulse carrier with a frequency of 60 [Hz] for 0.35 seconds and 7 [mA] to the lateral head of the gastrocnemius muscle and 8 [mA] to the medial head using a pair of electrodes, respectively.

(1-3) Experimental procedure The test subject (subject) was one healthy adult female (157 [cm], 45.9 [kg]). The experiment was conducted in accordance with procedures approved by the Ethics Committee of the Graduate School of Engineering Science, Osaka University. (A) Lifting type and (B) Saddle support type weight-loading device unloads the subject's weight moderately, and each of them at a speed (2.5 [km / h]) that the subject feels natural A weight-free walk was performed for 2 seconds. In each trial, the first half was weight-free only, and the second half was weight-free in addition to functional electrical stimulation.

  The four types of load-free walking with different conditions are as follows: (a) lifting-type weight-loading, (b) lifting-type weight-loading + functional electrical stimulation, (c) saddle-supporting weight-loading, (d) saddle Supporting body weight relief + functional electrical stimulation, kinematics of lower limbs by motion capture system (Optitrack, Natural Point, Inc., USA), and floor reaction force by treadmill system (ITR5018, Bertec Corp. , USA). In addition, an exercise period was set in advance before the experiment, and exercise measurement was performed after the subject was sufficiently used to each environment.

(1-4) Results / Discussion Figures 9 and 10 show the effects on the treadmill walking due to the difference between the two types of weight-loading methods: (A) Lifting type and (B) Saddle support type. The ankle push-off support effect by functional electrical stimulation under unloading is shown. In the saddle-supported weight-free walking in FIG. 9C, a larger stride is realized as compared with the lifting-type weight-free walking in FIG. 9A. Even when functional electrical stimulation was applied, effective exercise support was confirmed without losing the advantage in the saddle-supported weight-free walking (FIGS. 9B and 9D).

  Further, the difference between the two unloading methods was significantly observed in the floor reaction force (FIGS. 10A and 10B). That is, as shown in FIG. 10 (b), the vertical pedaling force of the saddle-supported weight-free walking has a bimodal waveform similar to that of natural walking, whereas the lifting-type weight-free walking in FIG. 10 (a). So, even at the end of the stance, it shows a high vertical pedaling force, making it an unnatural walk. This difference is further marked by functional electrical stimulation, and it can be seen that lifting support does not support exercise correctly. Saddle-supported support can achieve a natural pedaling pattern even with weight-unloading alone, while maintaining the pedaling force pattern with functional electrical stimulation, the push-off pedaling force can be increased to support exercise in a form close to nature. I understood.

<Experiment 2>
Next, in order to search for effective interventions in gait rehabilitation, features of an end effector type gait training device that combines a saddle-supported weight-bearing device and functional electrical stimulation (FES) To verify. This is a task-oriented approach aimed at reconstructing voluntary motor control in patients with neuromuscular disease. The difference between the lift-type weight-free walk and the saddle-supported weight-free walk is evaluated by the effects of comfortable walking speed, kinematics, floor reaction force, and FES.

(2-1) Subjects Subjects (subjects) were three healthy individuals (one male, two females, 24 years old). The experiment was conducted in accordance with procedures approved by the Ethics Committee of the Graduate School of Engineering Science, Osaka University. In addition, subjects A, B, and C were given sufficient explanations about the purpose and content of the experiment in advance, and consent was obtained from the subject.

(2-2) Device (2-2-1) As a device to relieve the body weight of the treadmill test subject, as in <Experiment 1>, lift the upper body of the subject with the installed pneumatic cylinder and harness. Type weight-relief device (New Assist, Interliha Co., Ltd., Japan) and “saddle-supported weight-relief device” (see FIGS. 1 to 3) acting on the subject's crotch by a seesaw-type mechanism Interventions were performed using types. Both weight-bearing systems include a split belt treadmill (ITR5018, Bertec Corp., USA) with a built-in floor reaction force meter. Hereinafter, the lifting type weight-bearing device is referred to as a harness-supporting weight-weighting device.

(2-2-2) Electrical stimulation device (FES)
As another intervention, functional electrical stimulation acting on the ankle joint adjacent to the end effector was employed. As the stimulation system, the same electrical stimulation apparatus as in the case of <Experiment 1> was used. In both the gastrocnemius lateral and medial heads, subjects A and B were stimulated for 0.15 seconds with a pulse carrier of 16 [mA] and subject C with an intensity of 12 [mA] and a frequency of 200 [Hz].

(2-2-3) Experimental procedure The subject walked for 110 seconds under the experimental conditions shown in FIG. FES was applied after 50 seconds. In each trial, the first 50 seconds (walking with or without exemption) and the last 50 seconds (walking with or without exemption under FES intervention) were analyzed. Lower limb kinematics (left and right hips, knees, legs, toes) were all measured at 100 [Hz] using a motion capture system (OptiTrack, NaturalPoint, Inc., USA) including 11 cameras. The floor reaction force was measured at 1000 [Hz] with a floor reaction force meter mounted on the split belt treadmill, and down-sampled to 100 [Hz]. Kinematics and floor reaction force were each adopted as an index for evaluation of the training system. The kinematics is whether or not the intervention subject can walk in a natural posture, and the floor reaction force is the interaction with the floor surface required for the three gait functions (balance control, hip extension, ankle push-off) Shows how the weight-bearing device, treadmill and FES support. During the subject's voluntary gait experiment, kinematics, floor reaction force and FES stimulation timing were recorded synchronously and the recorded data was smoothed and then normalized for one cycle of gait. The first grounding of the foot on the FES control side is 0%, and the next grounding is 100% (1 walk). Before the experiment, a walking practice period was provided in advance, and exercise measurements were made after the subjects were fully accustomed to each environment.

(2-3) Experimental results (2-3-1) Comfortable walking speed In FIG. 12, the walking speed of the comfortable walking according to the unloading amount of the harness-supporting weight-loading device and the saddle-supporting weight-loading device. Showing change. These subjective preference data were obtained from responses to questionnaire surveys on subjects.

(2-3-2) Change in weight-free walking by load-carrying method FIG. 13 shows that subject A walked with a moderate weight-free weight (33% BWS) in a harness-supported and saddle-supported weight-loaded system. FIG. 14 shows a vertical component, a longitudinal shear component, and a lateral shear component of the floor reaction force under the same conditions. These exercise data were also measured at the same speed in non-BWS gait.

(2-3-3) Change in Non-Weight-Free Walking with Walking Speed FIG. 15 shows non-weight-free walking with a different speed (1.5, 2.5, 3). 0.0, 3.5, 4.0, and 4.5 [km / h]), the three components of the floor reaction force are shown. Here, non-weight-free walking is 2 in the middle stance phase (10-30% gait phase) and the final stance phase (30-50% gait phase) at the comfortable walking speed (4.0 [km / h]). Two prominent peaks are seen. The first peak is mainly due to hip extension and the second peak is due to hip extension and ankle push-off.

  16 and 17 show that subject A at moderate weight relief (33% BWS) at various speeds (1.5, 2.5, 3.0, 3.5, and 4.5 [km / h]) shows three components of angle change and floor reaction force of each joint when saddle-supported weight-free walking is performed. In addition, the data of non-weight-free walking at a walking speed of 3.5 [km / h] comfortable under 33% BWS are included.

(2-3-4) Change in weight-free walking depending on the amount of load FIGS. 18 and 19 show the change in the angle of each joint and the floor reaction force when subject A performs a saddle-supported weight-free walking. Ingredients are shown. The parameter was the amount of unloading at a walking speed of 3.5 [km / h] (a comfortable walking speed under 33% BWS).

(2-3-5) Change in weight-free walking with and without FES FIG. 20 shows subject A (3.5 [km / h], 33% BWS), subject B (2.5 [km / h], 40% BWS) and subject C (4.5 [km / h], 40% BWS) performed a harness-supported weight-bearing walk and a saddle-supported weight-free walk at comfortable walking speeds under each condition. 3 shows the difference between the three components of the floor reaction force and the presence or absence of the FES.

(2-4) Consideration (2-4-1) Partial weight relief and treadmill intervention From the answers to the questionnaire of three subjects, as shown in FIG. 12, the comfortable walking speed increases as the weight relief amount increases. It turns out that there is a consistent trend of decline. This tendency was confirmed with both the harness support type and the saddle support type. Based on this data, we will consider changes in kinematics and floor reaction force depending on walking speed and unloading amount.

(2-4-2) Characteristic of kinematics FIG. 13 shows the joint angles of walking (harness-supporting weight-free walking, saddle-supporting weight-free walking, and non-weight-free walking) using different load-carrying methods. Show. The unloading amount was fixed at 33% BWS (medium weight unloading), and the walking speed was fixed at 3.5 [km / h] (the comfortable walking speed at this unloading amount). There was no significant difference in each joint angle between the two types of weight support systems, the harness support type and the saddle support type (r> 0.93, p <0.05). However, in these weight-free walking, the range of motion of the hip joint angle tended to decrease compared to non-weight-free walking.

  FIG. 16 shows each joint angle in saddle-supported weight-free walking at different walking speeds. The amount of unloading was fixed at 33% BWS (medium weight unloading). The joint angle in weight-free walking at a comfortable walking speed (3.5 [km / h]) or slightly slower (2.5, 3.0 [km / h]) It almost coincides with the joint angle in non-weight-free walking at 5 [km / h] (r> 0.93, p <0.05). However, the joint angle in the weight-free walking at the slowest speed or the fast speed tends to be different from that in the non-weight-free walking (r> 0.52, p <0.05).

  FIG. 18 shows saddle-supported weight-free walking with four levels of unloading. The walking speed was fixed at a comfortable walking speed (3.5 [km / h]) at a medium weight-free (33% BWS). The ankle joint angle (dashed line) in the non-weight-free walking is gradually bent to about 10 [deg] from the initial angle and then bent back to about 20 [deg]. However, regarding the joint angle in 50% BWS weight-free walking, the change in the ankle joint angle is small (23 to 53% of the walking cycle, ± 1.7 [deg]), and correlation with non-weight-free walking. Is small (r = 0.83, p <0.05). That is, if the weight-unloading amount is too large, the ankle joint push-off becomes a gait that hardly functions, which is inconvenient for rehabilitation. From these, in subject A, the saddle-supported weight-free weight is approximately 1.0% [km / h] slower than the comfortable walking speed to a comfortable walking speed, and up to about 33% BWS. It was found that walking support can be performed while maintaining a natural posture with the weight-unloading amount.

(2-4-3) Characteristics of floor reaction force In clinical walking training, many patients have difficulty in supporting their bodies with their legs, so a weight-free device is necessary. For this reason, effective weight-free walking must be considered. As described above, the comfortable walking speed decreases as the weight-unloading amount increases. For this reason, attention was paid to whether or not there was no load for low-speed walking. As shown in FIG. 15A, the vertical component of the floor reaction force during non-weight-free walking at low speed (2.5 [km / h]) does not show bimodality. On the other hand, from FIG. 17 (a), the vertical component of the floor reaction force in saddle-supported weight-free walking at low speed (2.5 [km / h]) and moderate weight-free (33% BWS) is prominent. Naive bimodality is observed. That is, in the saddle-supported weight-free walking, it was recognized that the vertical component of the floor reaction force showed two peaks not only at the comfortable walking speed but also at a lower speed. However, bimodality is observed in the vertical component of floor reaction force during saddle-supported weight-free walking under this characteristic (low speed, moderate weight-free (2.5 [km / h], 33% BWS). Although it is not always clear whether it is effective for support at present, it is considered suitable for walking training, assuming that stroke patients walk slowly.

  From FIG. 14 (c), the left and right shear components of the floor reaction force in the saddle-supported weight-free walking also show two peaks similar to those in the non-weight-free walking as compared with the harness-supported weight-free walking. . From FIG. 14 (b), the front-rear shear component of the floor reaction force showed a significant difference between the harness-supported weight-free walk, the saddle-supported weight-free walk, and the non-weight-free walk. Due to the shear component of floor reaction force, the impulse (normalized by weight and walking cycle) during one cycle (1 walk) is +46.7 [-], saddle for harness-supported weight-free walking It was -207.3 [-] in the support-type body weight-free walking, whereas it was -26.2 [-] in the non-body weight-free walking. The reason for this difference is thought to be that the harness and saddle that are in contact with the subject give momentum in the positive and negative directions, respectively. Therefore, it is considered that the harness-supporting weight-free walking and the saddle-supporting weight-free walking are obviously different actions. Such a result showed a similar tendency for the subjects B and C.

(2-4-4) Intervention by FES FIG. 20 shows that in non-weight-free walking, subjects A, B, and C do not show any significant change in floor reaction force due to the presence or absence of FES (r> 0.97). , p <0.05). This is because the treading force exerted by electrical stimulation of the muscles around the ankle joint is insufficient to support the total weight (without load relief) of the subject. Subject C is saddle-supported weight-free walking (Fig. 20 (a), subject C characteristic line (6)), FES increases the vertical component of floor reaction force by 3.0 [% BW], and supports harness During the body weight-free walking (characteristic line (5) in the figure), the vertical component of the floor reaction force increased by 2.8 [% BW] due to FES during the front swing phase. Whether or not this increase in pedaling force is sufficient to assist the patient's walking depends on the severity of the symptoms, but if the amplitude, frequency, and stimulation time of the stimulation current are adjusted appropriately, the generated pedaling force will be greater. (CL Lynch, and MR Popovic, “Functional electrical stimulation: closed-loop control of induced muscle contractions,” IEEE Control System Magazine, vol. 28, no. 2, pp. 40-50, 2008.).

  When subject A performs saddle-supported weight-free walking using FES, the vertical component of the floor reaction force (FIG. 20 (a), the characteristic line (6) of Subject A) is 5 by FES during the front swing leg period. It can be seen that there was an increase of 0 [% BW]. When subject B performs saddle-supported body weight-free walking using FES, the front / rear shear component of the floor reaction force (FIG. 20 (b), characteristic line (6) of Subject B) is 5 by FES at the initial stage of the free leg. It can be seen that there was an increase of 0 [% BW].

  In the present embodiment, the walking surface is moved by the rotation of the belt 23. However, the weight-loading unit 10 may be moved relative to the floor surface, and may reciprocate and move forward. It is good also as a structure which carries out walking training during.

  The feature of the technology according to the present embodiment is that the muscle coordination is extracted from the myoelectric potential signal of the lower limbs during voluntary exercise, and the exercise strategy (encoded high-order variable) selected by the central nervous system for realizing the exercise. The operation can be visualized in real time. Extracting feature information on muscle synergies, impedance (rigidity) and equilibrium points of body lower limbs from myoelectric signals that reflect motor commands of the central nervous system to visualize muscle coordination of body lower limbs, and stroke patients, etc. Robotic training will be conducted to regain the ideal body coordination while analyzing the plasticity of the central nervous system in the recovery of lower limb movements. Such research is unparalleled in Japan and overseas, and has both academic aspects approaching the essence of the functional recovery mechanism of the central nervous system and practical aspects based on clinical evidence. This technology has the potential to significantly change the current state of lower limb robotic therapy.

DESCRIPTION OF SYMBOLS 1 Walking training apparatus 10 Weight-bearing part 12 Arm 13 Saddle 14 Weight engaging part 16 Weight 20 Treadmill (floor part)
23 belt (endless belt)
25 Load detector 30 Motion detector (motion detector)
40 EMG detection unit (EMG detection unit)
DESCRIPTION OF SYMBOLS 50 Electrical stimulation part 501 and 502 Electrode 51 Stimulation signal generation part 52 Switch (instruction | indication part)
60 Control unit (calculation unit)
603 Muscle synergy calculation part (calculation means)
604 Equilibrium point / rigidity calculation unit 70 Motion observation unit (position detection unit)
80 Training processing unit 82 Projector (video projection unit)
84 Response determination unit (training evaluation unit)

Claims (12)

A saddle for supporting the subject at the crotch, and a weight-bearing portion for unloading part of the subject's weight;
A floor portion located below the saddle and having a floor surface on which the subject walks;
An gait training apparatus comprising: an electrical stimulation unit that performs electrical stimulation to an ankle plantar flexor muscle of a subject. The electrical stimulation unit is an electrode attached to the surface of the subject's ankle plantar flexor muscles;
A stimulation signal generator for applying a stimulation signal to the electrodes;
The walking training apparatus according to claim 1, further comprising: an instruction unit that instructs application of the stimulation signal. A detection unit for detecting the myoelectric potential of each muscle in the lower limb of the subject and the movement of each joint of the lower limb of the subject;
The gait training device according to claim 1, further comprising: a calculation unit that calculates a feature amount related to motion control of the lower limb tip from a detection result by the detection unit. The detection unit includes a myoelectric potential detection unit that measures myoelectric potential as the muscle activity, and a motion detection unit that detects joint motion as the lower limb motion,
The gait training apparatus according to claim 3, wherein the calculation unit calculates at least one of muscle synergy, stiffness, and equilibrium point as the feature amount from the detection result of the myoelectric potential detection unit and the detection result of the motion detection unit. . The floor portion is arranged in parallel with each other, is looped around a horizontally supported rotating body, and an endless belt whose upper surface constitutes the floor surface,
The walking training apparatus according to any one of claims 1 to 4, further comprising a load detection unit that detects a stepping position of the sole of the subject to the floor and a floor reaction force. A video projection unit for projecting a training video on the floor of the floor,
A load detection unit for detecting a stepping position and a floor reaction force of the sole of the subject to the floor; and
The training evaluation part of Claim 4 provided with the training evaluation part which evaluates training from the display position on the floor surface of the said training image | video and the said stepping position detected by one of the said motion detection part and the said load detection part. Walking training device. The weight-carrying portion includes a base, an arm that is pivotally supported by the base, and swings around the spindle within a vertical plane, and a weight.
The walking training apparatus according to claim 1, wherein the saddle is attached to a distal end of the arm, and the weight is suspended from a proximal end side of the arm. A weight engaging portion for vertically hanging the weight on the arm;
A plurality of weights are stacked on the base,
The gait training device according to claim 7, wherein the weight engaging portion engages an arbitrary plurality of weights with a base end of the arm from above. A saddle for supporting the subject at the crotch, and a weight-bearing portion for unloading part of the subject's weight;
A floor portion located below the saddle and having a floor surface on which the subject walks;
A detection unit for detecting the myoelectric potential of each muscle in the lower limb of the subject and the movement of each joint of the lower limb of the subject;
A calculation unit that calculates motion diagnosis information from the detection result of the detection unit,
The detection unit includes a myoelectric potential detection unit that measures myoelectric potential as the muscle activity, and a motion detection unit that detects joint motion as the lower limb motion,
The calculation unit calculates, as a feature amount, at least one of muscle synergy, stiffness, and equilibrium point, which are feature amounts related to motion control of the lower limb tip, from the detection result of the myoelectric potential detection unit and the detection result of the motion detection unit. Gait diagnostic device. In a weight-bearing device for a gait training device, comprising a saddle that supports the subject at the crotch, and unloading part of the subject's weight,
A base, an arm that is pivotally supported by the base and swings in a vertical plane, a weight, and a weight engaging portion that hangs the weight on the arm;
The saddle is attached to the tip of the arm;
A plurality of weights are stacked on the base,
The weight engaging portion is a weight-bearing device that engages an arbitrary plurality of weights with the base end of the arm from above. On the floor surface of the floor portion, the subject is supported by the saddle of the weight-carrying portion and the part of the body weight is released by supporting the subject at the crotch portion, and the floor surface of the floor portion of the subject indicated by the saddle A walking training method characterized by performing electrical stimulation to the ankle plantar flexor muscle of the subject while moving from the front toward the rear. A saddle for supporting the subject at the crotch, and a weight-bearing portion for unloading part of the subject's weight;
A floor portion located below the saddle and having a floor surface on which the subject walks;
A detection unit for detecting the myoelectric potential of each muscle in the lower limb of the subject and the movement of each joint of the lower limb of the subject;
A calculation unit that calculates motion diagnosis information from the detection result of the detection unit,
The detection unit measures myoelectric potential as the muscle activity and detects joint movement as the lower limb movement,
The calculation unit calculates at least one of muscle synergy, equilibrium point, and stiffness, which is a feature amount related to motion control of the lower limb tip, from the detection result of the myoelectric potential and the detection result of joint motion as the lower limb motion. Gait diagnosis method to calculate as