Exercise assistance device, alternative exercise evaluation device, exercise assistance method, and exercise assistance program

Abstract

The objective is to provide an exercise assistance device, exercise assistance method, and exercise assistance program that can induce voluntary movement from users by generating muscle activity equivalent to that of the target exercise, even when users perform alternative exercises instead of the target exercise. [Solution] This device provides positive or negative assistive force to a user performing an alternative exercise instead of a target exercise, thereby generating muscle activity equivalent to that of the target exercise. It comprises a simulation unit 11, an electromyography data input unit 12, a muscle synergy analysis unit 13, an output assistive pattern generation unit 14, an output unit 15, a storage unit 16, and an exercise device 30. The simulation unit 11 inputs assistive patterns, which combine phase and assistive force in a time series, into a musculoskeletal model to generate electromyography data sets for each assistive pattern. It includes a body movement data measurement unit 11a, a muscle strength data input unit 11b, a musculoskeletal model generation unit 11c, and an assistive force adjustment unit 11d.

Description

[Technical Field]

[0001] This invention relates to technology that assists in the rehabilitation of patients with paralysis symptoms and in the transmission of specific skills and technical instruction. [Background technology]

[0002] In Japan, the population aged 65 and over exceeded 29.1% of the total population in 2023, and this number is expected to continue to increase. Generally, older people are at higher risk of stroke. A stroke is a disease in which the supply of blood and nerve cells to the brain is interrupted due to the rupture or blockage of blood vessels in the brain. One of the most common after-effects of a stroke is hemiplegia, where one side of the body is paralyzed. Depending on the degree of paralysis, stroke patients with hemiplegia may have difficulty walking normally. Therefore, rehabilitation is necessary to restore the walking function of stroke patients, and the demand for it is rapidly increasing.

[0003] There are various methods for gait training for stroke patients depending on the stage of the disease, but rehabilitation can be divided into three phases: the acute phase (1-14 days), the recovery phase (up to several months), and the maintenance phase (several months to several years). In the acute phase, range of motion exercises are performed with the aim of preventing and improving rheumatoid arthritis and muscle contractures. In the recovery phase, with the assistance of a physical therapist, parallel bars and active exercises are performed to restore the ability to walk independently. In the maintenance phase, gait training using a treadmill is performed to maintain and improve muscle strength, endurance, and walking ability. In particular, rehabilitation in the recovery phase is a crucial period for patients to acquire the ability to walk normally, but while care by a physical therapist is indispensable, the increasing number of stroke patients has led to a shortage of physical therapists, resulting in a problem where patients cannot receive adequate rehabilitation. Therefore, there is a need for technology that can safely provide normal gait training to paralyzed patients without requiring constant care from a physical therapist during the recovery phase.

[0004] As a technique for improving a user's walking motion in a simple manner, a control device is known that acquires sensor values ​​for the user's walking, uses an estimation model to calculate an estimated value of the walking phase from the sensor values, and determines the amount of assistance from the calculated estimated phase according to a set assistance pattern (see Patent Document 1). This determines the amount of assistance given to the user by guiding the trajectory of the center of pressure toward the inner side of the forefoot or by randomly changing the amount of assistance according to the progress of the walking round. However, the control device described in Patent Document 1 automates the process by which physical therapists manually adjust the amount of assistance from a basic pattern to suit each patient. It does not enable patients to perform rehabilitation on their own, so the burden on physical therapists remains significant.

[0005] While it has long been known that ergometers are effective in the rehabilitation of stroke patients and that pedaling exercises promote the recovery of lower limb function for walking, this has been based on the subjective assessments of physical therapists and has not involved a quantitative analysis of the assistive force provided by the ergometer or its contribution to the lower limb muscles. Therefore, in order to find the characteristics of muscle activity during exercise, a muscle activity analysis device is known that uses a technology to decompose the amount of muscle activity into muscle patterns (muscle synergies) and their activity amounts based on electromyography measurement data, thereby detecting electromyographic potentials when pedaling an exercise bike (registered trademark), and enabling appropriate advice in response to changes in muscle activity (see Patent Document 2). This method involves attaching electrodes to the surface of the subject's skin and, based on the detected electromyographic activity, calculating instantaneous muscle synergy, which represents muscle activity when multiple types of muscles work together during a first time period, and also calculating baseline muscle synergy, which represents muscle activity during a second time period longer than the first. Feature quantities are then calculated for both the baseline and instantaneous muscle synergy. This allows for understanding changes in the subject's muscle activity and providing advice by focusing on the time-series changes in muscle synergy during specific exercises, such as cycling on an exercise bike. However, the muscle activity analysis device described in Patent Document 2 has the problem that it does not facilitate the independent performance of other, safer exercises as an alternative to exercises that are difficult for hemiplegic patients to perform on their own, such as walking. [Prior art documents] [Patent Documents]

[0006] [Patent Document 1] Japanese Patent Publication No. 2024-022117 [Patent Document 2] Japanese Patent Publication No. 2024-119689 [Overview of the project] [Problems that the invention aims to solve]

[0007] The need for alternative exercises that provide equivalent benefits exists not only for stroke patients but also for anyone who finds it difficult to perform necessary exercises on their own. Furthermore, the ability to easily perform alternative exercises is beneficial for a wide range of users, not only in cases of physical limitations such as the after-effects of a stroke, but also in cases where it is difficult to perform target exercises due to factors such as time, location, and cost. In view of these circumstances, the present invention aims to provide an exercise assistance device, an exercise assistance method, and an exercise assistance program that can induce voluntary movement from a user by generating muscle activity equivalent to that of the target exercise when the user performs an alternative exercise instead of the target exercise. The present invention also aims to provide an alternative exercise determination device that determines whether or not an exercise to be judged is substitutable for the target exercise. [Means for solving the problem]

[0008] The inventors have found that walking motion can be represented by the coordinated action of muscles forming a distributed coordination system (muscle synergy), and that muscle synergy similar to that of walking motion appears in seated pedaling exercise, which is considered useful based on the empirical rules of physical therapists (PTs). Furthermore, they have found that by applying assistive force to suit each patient's individual physical condition from the perspective of muscle synergy when performing pedaling exercise, it is possible to elicit spontaneous voluntary movement from the patient, and a device can be realized that allows patients to effectively train the muscle functions necessary for walking on their own. Moreover, these are not limited to the combination of walking and pedaling, and due to the similarity of muscle synergy patterns, they can be applied to a wide range of exercise items that have equivalent effects.

[0009] In other words, the exercise assistance device according to the first aspect of the present invention is a device that provides positive or negative assistive force to a user performing an alternative exercise instead of a target exercise, thereby generating muscle activity equivalent to that of the target exercise, and comprises the following parts 1-1) to 1-5). 1-1) A simulation unit comprising: a body movement data measurement unit that measures the user's body movement data related to alternative exercises; a muscle strength data input unit that accepts input of the user's muscle strength data; a musculoskeletal model generation unit that generates a musculoskeletal model of the user from the body movement data and muscle strength data; and an assist force adjustment unit that applies assist force to the musculoskeletal model by changing the assist force for each phase of the alternative exercise, and inputting an assist pattern that combines the phase and assist force in a time series to the musculoskeletal model to generate a group of electromyographic data for each assist pattern. 1-2) An electromyographic data input unit that accepts input of a first group of electromyographic data for each phase in the target movement and a second group of electromyographic data for each phase in the auxiliary pattern. 1-3) Muscle synergy analysis unit that analyzes the muscle synergies from the input first electromyography data set and second electromyography data set. 1-4) Output auxiliary pattern generation unit that compares the first muscle synergy analyzed from the first electromyographic data group with the second muscle synergy analyzed from the second electromyographic data group, extracts auxiliary patterns assigned to the second electromyographic data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or in which the expression of the second muscle synergy with the highest correlation coefficient is observed, and generates them as output auxiliary patterns. 1-5) An output unit that outputs an auxiliary pattern for output to an exercise device that performs alternative movements.

[0010] This configuration allows for the induction of muscle activity in alternative exercises that is similar to the coordinated muscle movements performed during the target exercise, enabling effective training. Here, the target exercise is the exercise that the user aims to perform. For example, if a stroke patient with hemiplegia is undergoing rehabilitation with the goal of walking, then walking would be the target exercise. An alternative exercise is an exercise that can induce muscle activity similar to the coordinated muscle movements performed during the target exercise, and can be said to be an exercise that provides a training effect close to that of the target exercise. In the example above, if walking is the target exercise for the stroke patient with hemiplegia, then, for example, pedaling using an ergometer would be the alternative exercise. In the simulation unit described in 1-1) above, the body motion data measured by the body motion data measurement unit preferably includes body motion data such as posture measured by motion capture. However, it is also possible to use a wide range of data that can be obtained from the user's body, such as the force applied by the user during exercise, like the pedaling force during pedaling. Here, assistive force refers to the force applied directly or indirectly to the user's body performing alternative movement. A positive assistive force is considered supportive force (assistance), while a negative assistive force is considered a load. In this specification, the electromyography data set refers to time-series electromyography data from multiple muscles, specifically represented as a matrix of n (number of muscles) × m (number of time-series electromyography data).

[0011] The electromyography data input unit described in 1-2) above may accept electromyography data input from an electromyography sensor attached to the subject's body, or it may accept electromyography data acquired in advance. The muscle synergy analysis unit in 1-3) analyzes muscle synergy for each of the first muscle potential data and the second muscle potential data. Since the second muscle potential data is generated for each auxiliary pattern, muscle synergy is analyzed for all the generated second muscle potential data. Note that muscle synergy means a cooperative structure found in the activities of multiple muscles and refers to the simultaneous activities of multiple muscles. By providing the output auxiliary pattern generation unit in 1-4), it becomes possible to quantitatively design muscle performance training based on the similarity of movements from the perspective of the muscle coordination system. The output unit in 1-5) may output the output auxiliary pattern to the motion device in either a wired or wireless manner.

[0012] The motion assistance device according to the first aspect of the present invention preferably further includes 1-6) a storage unit that stores at least the output auxiliary pattern. By providing the storage unit in 1-6), the convenience during repeated use is improved. Also, by periodically generating output auxiliary patterns for each user and storing them in the storage unit, it becomes possible to utilize the stored data for measuring the exercise effect of the user and creating future exercise plans. Note that in the storage unit, data input or generated in the simulation unit such as the type of exercise, body movement data, muscle strength data, musculoskeletal model, auxiliary pattern, the first or second muscle potential data group, muscle synergy analysis results, and the devices used for measurement may be stored.

[0013] In the motion assistance device according to the first aspect of the present invention, the first muscle potential data group is preferably muscle potential data during the walking motion of a healthy person or a hemiplegic patient after stroke, and more preferably, muscle potential data during the walking motion of a healthy person. By using the first muscle potential data group as muscle potential data during the walking motion of a healthy person, data targeting a more normal walking motion can be utilized. Also, by using muscle potential data during the walking motion of a hemiplegic patient after stroke, data targeting a walking motion adapted to the recovery state of the user can be utilized.

[0014] In the exercise assistance device according to the first aspect of the present invention, the exercise device is an ergometer having left and right independent pedals, and it is preferable to further include the exercise device. By using an ergometer having left and right independent pedals, it is possible to effectively prevent the situation where the rehabilitation effect cannot be sufficiently obtained by covering the paralyzed side foot with the healthy side foot, and it is possible to ensure the rehabilitation effect for patients with different left and right leg strengths, such as stroke hemiplegia patients. Here, as an example of an ergometer having left and right independent pedals, it refers to an ergometer that can move the opposite pedal as if it were a normal pedaling device from the paralyzed limb side, while the opposite pedal cannot be moved from the healthy limb side, and the patient is required to exert the force of the paralyzed limb side in order to perform the pedaling movement. In addition, the ergometer used as the exercise device is preferably an ergometer of a type provided with a backrest on a chair, called a recumbent ergometer or a recumbent bike. The pedaling exercise using a recumbent ergometer has less risk of the patient falling compared to walking, etc., so it becomes possible for the patient to perform rehabilitation more easily and safely alone.

[0015] In the exercise assistance device according to the first aspect of the present invention, the output assistance pattern generation unit preferably generates an output assistance pattern in which the assistance force acts individually on the left and right pedals included in the exercise device. Thereby, an appropriate assistance force can be given to a master-slave type exercise device, and effective training becomes possible.

[0016] The alternative exercise determination device of the present invention is a device that determines the presence or absence of substitutability with a target exercise regarding the exercise to be determined, and includes the following parts 1) to 4). 1) A simulation unit comprising: a body movement data measurement unit that measures the user's body movement data related to the exercise to be judged; a muscle strength data input unit that accepts input of the user's muscle strength data; a musculoskeletal model generation unit that generates a musculoskeletal model of the user from the body movement data and muscle strength data; and an assist force adjustment unit that applies assist force to the musculoskeletal model by changing the assist force for each phase of the exercise to be judged, and inputting an assist pattern that combines the phase and assist force in a time series to the musculoskeletal model to generate a group of electromyographic data for each assist pattern. 2) An electromyographic data input unit that accepts input of a first group of electromyographic data for each phase in the target movement and a second group of electromyographic data for each phase in the auxiliary pattern. 3) Muscle synergy analysis unit that analyzes the muscle synergies from the input first electromyography data set and second electromyography data set. 4) A substitute exercise determination unit that compares the first muscle synergy analyzed from the first electromyographic data group with the second muscle synergy analyzed from the second electromyographic data group, and determines that the exercise under evaluation is a substitute for the target exercise if the proportion of the second electromyographic data group in which the expression of the second muscle synergy, whose correlation coefficient exceeds a predetermined threshold with respect to the first muscle synergy, exceeds a predetermined threshold.

[0017] With this configuration, it is possible to determine the general relationship between different exercises, and also to determine whether the exercise being evaluated can serve as a substitute for the target exercise for a particular user. Furthermore, a memory unit may be provided to store the evaluation results from the substitute exercise evaluation unit. By accumulating the evaluation results in the memory unit, it is possible to use them as data to verify the relationship between different exercises. In addition, if there is a change in the user's motor function due to rehabilitation or injury, it is also effective to perform the evaluation again and store the data. If a memory unit is provided, it may not only store the result of whether the pedaling exercise, which is the exercise being evaluated, is a substitute for the walking exercise, which is the target exercise, but also the type of exercise, data input or generated in the simulation unit such as body movement data, muscle strength data, musculoskeletal model, and auxiliary patterns, first or second electromyography data sets, muscle synergy analysis results, and equipment used for measurement.

[0018] A second aspect of the present invention is an exercise assistance device that determines whether a target exercise is substitutable for a target exercise, and if it is usable as an alternative exercise, provides a positive or negative assistive force to the user performing the alternative exercise to generate muscle activity equivalent to that of the target exercise, and comprises the following parts 2-1) to 2-6). 2-1) A simulation unit comprising: a body movement data measurement unit that measures the user's body movement data related to the exercise to be judged; a muscle strength data input unit that accepts input of the user's muscle strength data; a musculoskeletal model generation unit that generates a musculoskeletal model of the user from the body movement data and muscle strength data; and an assist force adjustment unit that applies assist force to the musculoskeletal model by changing the assist force for each phase of the exercise to be judged, and inputting an assist pattern that combines the phase and assist force in a time series to the musculoskeletal model to generate a group of electromyographic data for each assist pattern. 2-2) An electromyographic data input unit that accepts input of a first group of electromyographic data for each phase in the target movement and a second group of electromyographic data for each phase in the auxiliary pattern. 2-3) Muscle synergy analysis unit that analyzes the muscle synergies from the input first electromyography data set and second electromyography data set. 2-4) A substitute exercise determination unit that compares the first muscle synergy analyzed from the first electromyographic data group with the second muscle synergy analyzed from the second electromyographic data group, and determines that the exercise under evaluation is a substitute exercise that is substitutable for the target exercise if the proportion of the second electromyographic data group in which the expression of the second muscle synergy, whose correlation coefficient exceeds a predetermined threshold with respect to the first muscle synergy, exceeds a predetermined threshold. 2-5) Output auxiliary pattern generation unit: When the target exercise is determined to be an alternative exercise, the first muscle synergy analyzed from the first electromyography data group and the second muscle synergy analyzed from the second electromyography data group are compared, and the auxiliary pattern given to the second electromyography data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or in which the expression of the second muscle synergy with the highest correlation coefficient is observed, is extracted and generated as an output auxiliary pattern. 2-6) An output unit that outputs an auxiliary output pattern generated for a motor device that performs alternative movement.

[0019] With this configuration, the system can determine whether the exercise to be judged can substitute for the target exercise, generate an appropriate auxiliary pattern for the user, and perform an alternative exercise that replaces the target exercise, thereby improving functionality. Furthermore, a memory unit may be provided to store the judgment results from the alternative exercise judgment unit and the generated output auxiliary patterns. By storing the judgment results and output auxiliary patterns in the memory unit, the stored data can be used to verify the relationships between exercises, measure the exercise effect for the user, and help in creating future exercise plans. If a memory unit is provided, it may store not only the result of whether the pedaling exercise, which is the exercise to be judged, is a substitute for the walking exercise, which is the target exercise, and the output auxiliary patterns generated by the output auxiliary pattern generation unit, but also data such as the type of exercise, body movement data, muscle strength data, musculoskeletal model, auxiliary patterns, data input or generated in the simulation unit, first or second electromyography data sets, muscle synergy analysis results, and the equipment used for measurement.

[0020] The exercise assistance method of the present invention provides a positive or negative assistive force to a user performing an alternative exercise instead of a target exercise, thereby generating muscle activity equivalent to that of the target exercise, and comprises the following steps a1) to a6). a1) A simulation step that measures the user's body movement data related to alternative exercises, accepts input of the user's muscle strength data, generates a musculoskeletal model of the user from the body movement data and muscle strength data, applies assistive force to the musculoskeletal model by changing the assistive force for each phase of the alternative exercise, inputs an assistive pattern that combines the phase and assistive force in a time series into the musculoskeletal model, and generates a set of electromyographic data for each assistive pattern. a2) An electromyographic data input step that accepts input of a first group of electromyographic data for each phase in the target movement and a second group of electromyographic data for each phase in the auxiliary pattern. a3) A muscle synergy analysis step that analyzes the muscle synergies from the input first electromyography data set and second electromyography data set. a4) An output auxiliary pattern generation step, which involves comparing the first muscle synergy analyzed from the first electromyography data group with the second muscle synergy analyzed from the second electromyography data group, extracting auxiliary patterns assigned to the second electromyography data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or in which the expression of the second muscle synergy with the highest correlation coefficient is observed, and generating them as output auxiliary patterns. a5) Output step that outputs an auxiliary output pattern generated for a motor device that performs alternative movement.

[0021] The exercise assistance method of the present invention preferably further comprises a6) a storage step of storing at least an output assistance pattern.

[0022] The exercise assistance program of the present invention is a program that provides positive or negative assistance to a user performing an alternative exercise instead of a target exercise, thereby generating muscle activity equivalent to that of the target exercise, and causes a computer to execute the following steps b1) to b6). b1) A simulation step in which measurement data obtained from the user's body movement data related to alternative exercises and the user's muscle strength data are input, a musculoskeletal model of the user is generated from the body movement data and muscle strength data, the assisting force is changed for each phase of the alternative exercise and the assisting force is applied to the musculoskeletal model, and assisting patterns that combine the phase and assisting force in a time series are input to the musculoskeletal model to generate a set of electromyographic data for each assisting pattern. b2) An electromyographic data input step that accepts input of a first group of electromyographic data for each phase in the target movement and a second group of electromyographic data for each phase in the auxiliary pattern. b3) A muscle synergy analysis step that analyzes the muscle synergies from the input first electromyography data set and second electromyography data set. b4) An output auxiliary pattern generation step, which involves comparing the first muscle synergy analyzed from the first electromyography data group with the second muscle synergy analyzed from the second electromyography data group, extracting auxiliary patterns assigned to the second electromyography data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or in which the expression of the second muscle synergy with the highest correlation coefficient is observed, and generating them as output auxiliary patterns. b5) Output step: Outputs an auxiliary output pattern generated for a motor device that performs alternative movement.

[0023] The exercise assistance program of the present invention preferably further comprises b6) a storage step of storing at least an output assistance pattern. [Effects of the Invention]

[0024] The exercise assistance device, exercise assistance method, and exercise assistance program of the present invention have the effect of eliciting voluntary movements from users who perform alternative exercises instead of target exercises by generating muscle activity equivalent to that of the target exercise. Furthermore, the alternative movement determination device of the present invention has the effect of being able to determine whether or not the movement to be determined is substitutable with the target movement. [Brief explanation of the drawing]

[0025] [Figure 1] Functional block diagram of the alternative motion determination device of Example 1 [Figure 2] Schematic diagram of the alternative motion determination device of Example 1 [Figure 3] Functional block diagram of the exercise assistance device of Example 2 [Figure 4] Schematic diagram of the exercise assistance device in Example 2 [Figure 5] Schematic flowchart of the exercise assistance method in Example 2 [Figure 6] Diagram illustrating body movement data measurement and pedal angle. [Figure 7] Results of muscle synergy analysis during walking in healthy subjects [Figure 8] Muscle synergy analysis results during simulation [Figure 9] A graph showing the pedaling force and power assist patterns of the subject during pedaling motion. [Figure 10] Results of muscle synergy analysis during pedaling motion in user subjects. [Figure 11] Results of muscle synergy analysis during walking in subjects with mild symptoms [Figure 12] Results of muscle synergy analysis during pedaling exercise in subjects with mild symptoms (Example) [Figure 13] Muscle synergy analysis results during pedaling exercise in subjects with mild symptoms (comparative example) [Figure 14] Results of muscle synergy analysis during walking in severely ill subjects [Figure 15] Results of muscle synergy analysis during pedaling exercise in severely ill subjects (Example) [Figure 16] Functional block diagram of the exercise assistance device of Example 4 [Figure 17] Diagram illustrating the muscles of the lower limbs [Modes for carrying out the invention]

[0026] Hereinafter, an example of an embodiment of the present invention will be described in detail with reference to the drawings. First, in Example 1, an alternative exercise determination device will be described, which is a preliminary step to using the exercise assistance device according to the first aspect of the present invention, and which can determine the relationship between two seemingly unrelated exercises and find an alternative exercise to the target exercise. Then, in Examples 2 and 3, an exercise assistance device according to the first aspect of the present invention will be described, and in Example 4, an exercise assistance device according to the second aspect of the present invention will be described. It should be noted that the scope of the present invention is not limited to the following embodiments and illustrated examples, and numerous modifications and variations are possible. [Examples]

[0027] Human walking consists of several basic movements with different roles (lifting the leg, pushing off with the leg). Humans achieve these basic movements by coordinating multiple muscles. Healthy individuals' walking motion consists of three basic movements, and when the load is appropriately set during pedaling, three basic movements appear that consist of muscle combinations and muscle activation timings similar to walking motion. Therefore, in this embodiment, we will explain, using the walking motion of a healthy individual as the target movement, whether pedaling motion can serve as a substitute movement for the target walking motion, using a substitute movement determination device as an example. Figure 17 shows an explanatory diagram of the muscles of the lower limbs. In this example, eight muscles of the lower limbs, as shown in Figure 17, which are particularly known to generate walking motion, will be analyzed. Table 1 below shows the names, abbreviations, and functions of the target lower limb muscles.

[0028] [Table 1]

[0029] Figure 1 shows a functional block diagram of the alternative exercise determination device of Example 1. Figure 2 is an overview diagram of the alternative exercise determination device of Example 1, where (1) shows an image of walking motion and its muscle synergy, (2) shows an image of pedaling motion and its muscle synergy, and (3) shows a musculoskeletal model. Note that the walking motion and pedaling motion shown in Figure 2 are examples of those performed by healthy subjects. As shown in Figure 1, the alternative exercise determination device 1a of Embodiment 1 is a device that determines whether or not an exercise to be determined is substitutable with a target exercise, and comprises a simulation unit 11, an electromyography data input unit 12, a muscle synergy analysis unit 13, a storage unit 16, and an alternative exercise determination unit 17.

[0030] The simulation unit 11 inputs assistance patterns, which combine phase and assistance force in a time series, into a musculoskeletal model to generate a set of electromyographic data for each assistance pattern. It includes a body movement data measurement unit 11a, a muscle strength data input unit 11b, a musculoskeletal model generation unit 11c, and an assistance force adjustment unit 11d. The body movement data measurement unit 11a measures the user's body movement data related to the movement to be judged, and in this embodiment, motion capture is used. The muscle strength data input unit 11b accepts the user's muscle strength data. The musculoskeletal model generation unit 11c generates the user's musculoskeletal model 8 shown in Figure 2(3) from the body movement data and muscle strength data. The assistance force adjustment unit 11d applies assistance force to the musculoskeletal model by changing the assistance force for each phase of the movement to be judged.

[0031] The electromyographic data input unit 12 accepts input of a first group of electromyographic data for each phase in the target exercise and a second group of electromyographic data for each phase in the auxiliary pattern. For input of the first electromyography (EMG) data set, as shown in Figure 2(1), the EMG sensor 20 receives input of the first EMG data set for each phase during walking in a healthy person. Specifically, as shown in Figure 2(1), a healthy subject wearing the EMG sensor 20 performs walking motion (4a to 4d), and inputs the first EMG data set, consisting of EMG signals obtained from eight muscles during walking, into the EMG data input unit 12. Similarly, for the input of the second electromyography (EMG) data set, as shown in Figure 2(2), the EMG sensor 20 receives input of the second EMG data set for each phase during pedaling motion in a healthy person. Specifically, as shown in Figure 2(2), a subject wearing the EMG sensor 20 (a healthy subject in Figure 2) performs pedaling motion (5a to 5d) and inputs the second EMG data set, consisting of EMG signals obtained from eight muscles during walking motion, into the EMG data input unit 12.

[0032] The muscle synergy analysis unit 13 analyzes the muscle synergies from the input first and second electromyographic data sets and extracts the basic movements. In the walking movement shown in Figure 2(1), the expression of five muscle synergies (6a-6e) is observed, and it can be seen that the muscle synergies (6a-6e) act sequentially as the walking movement (4a-4d) progresses. Similarly, in the pedaling movement shown in Figure 2(2), the expression of five muscle synergies (7a-7e) is observed, and it can be seen that the muscle synergies (7a-7e) act sequentially as the pedaling movement (5a-5d) progresses. Figure 2 shows just one example, but the number of muscle synergies that are expressed will also change by changing the assisting force.

[0033] The alternative exercise determination unit 17 compares the first muscle synergy analyzed from the first electromyographic data group with the second muscle synergy analyzed from the second electromyographic data group. If the proportion of the second electromyographic data group in which the second muscle synergy, whose correlation coefficient exceeds a predetermined threshold, is observed exceeds a predetermined threshold, the unit determines that the exercise under evaluation is a substitute for the target exercise. Specifically, the unit calculates this by comparing the muscle synergies (6a-6e) shown in Figure 2 with multiple muscle synergies (7a-7e) obtained by changing the assisting force. The memory unit 16 stores the result of the alternative movement determination. The configuration of the alternative motion determination device 1a in Example 1 will be described in detail below.

[0034] (Regarding the simulation department) Regarding the input of the second electromyography (EMG) data group, in this embodiment, the simulation unit 11 receives input of the second EMG data group for each phase in the auxiliary pattern related to the pedaling motion of a stroke hemiplegia patient. Specifically, multiple markers are attached to the subject, who is a stroke hemiplegia patient, and the subject performs pedaling motion using an ergometer (Konami Sports "AEROBIKE 75XLII"), and body movements are measured using a motion capture device (not shown), which is the body movement data measurement unit 11a. In addition, the user's muscle strength data is input using the muscle strength data input unit 11b. The user's muscle strength data may be input from data acquired in advance, or it may be measured and input on the spot. Based on the input body movement data and muscle strength data, the musculoskeletal model generation unit 11c generates a musculoskeletal model 8 of the user performing pedaling motion, as shown in Figure 2(3). Furthermore, parameters such as the force that the lower limb muscles can exert, calculated based on the input data, are derived using known muscle strength estimation methods (Non-patent literature: D. Chugo, et al., “Human Models Simulating the Physical Conditions of the Elderly Individual and Standing Assistance Method Based on These Models,” In: 24th Int. Conf. on Climbing and Walking Robots (CLAWAR2022), pp.604-616, (2022).).

[0035] While typical ergometers allow for pedal load adjustment, the load remains constant. However, normal walking consists of several different basic movements, and in order to determine whether pedaling can induce movements similar to these basic movements, it is necessary to adjust the pedal load in various ways according to the corresponding basic movements. Therefore, the simulation unit 11 uses the assistive force adjustment unit 11d to change the assistive force for each phase of the pedaling movement and apply assistive force to the user's musculoskeletal model. This generates a large number of patterns (second electromyographic data group) for when various assistive forces are applied to the user's musculoskeletal model, which can be input to the electromyographic data input unit 12, allowing for precise comparison with the first electromyographic data group.

[0036] (About the Muscle Synergy Analysis Department) In this way, by measuring the electromyograms of each muscle in healthy individuals performing walking exercises or patients performing pedaling exercises, the force exerted by each muscle can be confirmed. However, since each basic movement is achieved through the coordination of multiple muscles, an index is needed to evaluate muscle coordination. Therefore, the muscle synergy analysis unit 13 evaluates the coordinated relationships of muscles using muscle synergy analysis. Human movements are controlled cooperatively by multiple muscles that constitute a redundant system of degrees of freedom, and the coordinated movement of multiple muscles that constitute one basic movement is called muscle synergy. The muscle synergy analysis unit 13 uses muscle synergy analysis to approximate and represent the vast amount of electromyographic data that constitutes pedaling movements as a linear sum of multiple synergies. This makes it possible to clarify what basic movements constitute a patient's pedaling movement from the perspective of muscle activity.

[0037] (Non-negative matrix factorization) The muscle synergy analysis method used in this embodiment is described below. Electromyogram data from patients performing pedaling exercises are converted to positive values ​​using RMS (Root Mean Square) processing to remove negative values, and are represented as a matrix M of size m (number of muscles) × n (number of time-series samples). Furthermore, matrix M can be approximated by two matrices W and H shown in Equation 1 below using Nonnegative Matrix Factorization (NMF). Here, matrix W is a matrix of size m (number of muscles) × k (number of muscle synergies), and matrix H is a matrix of size k (number of muscle synergies) × n (number of time-series samples).

[0038]

number

[0039] Each column in W represents a weighting coefficient indicating the extent to which each muscle contributes to muscle synergy, and each row in H represents an activation coefficient showing how muscle synergy is exerted over time. In other words, Equation 1 above shows which muscles are involved in a single basic human movement and at what timing that basic movement occurs. Equation 2 below shows the Euclidean distance between M and WH. However, M ij This represents the ij component of matrix M. The same applies to matrices W and H.

[0040]

number

[0041] To solve equation 1 above using NMF, the following ALS (alternating least squares) algorithm is applied. Step 1: The allowable value of J in equation 2 above is ε = 10 -4 This was the setting in this example. Furthermore, the number k of muscle synergies to be extracted is determined. Step 2: Initialize W and H by randomly selecting non-negative entries for the matrix. Step 3: W + HW=MHT to obtain W + Here, W ij + is given by the following Equation 3.

[0042] [Number]

[0043] Step 4: Set all negative entries in W + to 0. Step 5: H +T W T W = M T to obtain H + Here, H ij + is given by the following Equation 4.

[0044] [Number]

[0045] Set all negative entries in H to 0. The updated W + and H + are calculated for the Euclidean distance therebetween as in the above Equation 2. If J > ε, return to Step 3. Otherwise, output W and H and end.

[0046] (Identification of Muscle Synergy) In the aforementioned NMF, using the n muscle EMG data groups, up to n - 1 muscle synergies can be extracted. The original muscle EMG data group M is reproduced by multiplying the spatial component W representing the muscle synergy and the temporal component H representing the muscle synergy, and the reproducibility increases as the number of muscle synergies increases. On the other hand, the number of muscle synergies means the number of basic movements, and considering that humans perform the pedaling motion with a small number of simple basic movements, it is desirable that the number of muscle synergies to be extracted is small. Therefore, the reproducibility of M by W and H is evaluated using the VAF (Variance Accounted For) shown in Equation 5 below. In this example, the synergy number k is input to NMF from 1 to n-1, and the smallest k that takes a value of 90% or more in Equation 5 below is selected as the synergy number to be extracted. Here, ||Z|| F And tr(Z) are the norm and trace of matrix Z.

[0047]

number

[0048] (Regarding the determination of alternative movements) The alternative exercise determination unit 17 compares the first muscle synergy of walking exercise in healthy individuals, analyzed by the muscle synergy analysis unit 13, with the second muscle synergy of pedaling exercise using the patient's musculoskeletal model. If the proportion of second electromyographic data groups showing the expression of the second muscle synergy, whose correlation coefficient exceeds a predetermined threshold with respect to the first muscle synergy, exceeds a predetermined threshold, the unit determines that the pedaling exercise, which is the exercise under evaluation, is a substitute for walking exercise, which is the target exercise. Conversely, if the proportion of second electromyographic data groups showing the expression of the second muscle synergy, whose correlation coefficient exceeds a predetermined threshold with respect to the first muscle synergy, does not exceed a predetermined threshold, the unit determines that the pedaling exercise, which is the exercise under evaluation, is not a substitute for walking exercise, which is the target exercise. In this embodiment, the alternative movement determination unit 17 determined that the pedaling movement, which is the movement to be determined, is an alternative movement to the walking movement, which is the target movement.

[0049] The memory unit 16 stores the results of the alternative movement determination, but it does not simply store the result of whether the pedaling movement, which is the movement to be determined, is an alternative movement to the walking movement, which is the target movement. It also stores the type of movement, data input or generated in the simulation unit 11 such as body movement data, muscle strength data, musculoskeletal model 8, and auxiliary patterns, as well as the first or second electromyography data group, muscle synergy analysis results, and the equipment used for measurement. In this embodiment, not only is it determined that the target exercise is walking and the exercise to be judged is pedaling, and that pedaling is substitutable for walking, but also that user body movement data, muscle strength data, generated musculoskeletal model 8, auxiliary patterns, first or second electromyography data sets, muscle synergy analysis results, and measuring instruments are stored.

[0050] Thus, by using the alternative exercise determination device 1a of Example 1, it is possible to determine the general relationship between exercises, and also to determine whether the exercise being evaluated can serve as a substitute for the target exercise for a particular user. The determination results can be stored in the memory unit 16 and used as data to verify the relationships between exercises. Furthermore, if there is a change in the user's motor function due to rehabilitation or injury, it is also effective to perform the determination again. [Examples]

[0051] This embodiment describes an exercise assistance device that allows a user to perform an alternative exercise to replace the target exercise, assuming that the relationship between the target exercise and the alternative exercise is known. Figure 3 shows a functional block diagram of the exercise assistance device of Example 2. As shown in Figure 3, the exercise assistance device 1 of Example 2 is a device that provides positive or negative assistive force to a user performing an alternative exercise instead of a target exercise, thereby generating muscle activity equivalent to that of the target exercise. It comprises a simulation unit 11, an electromyographic data input unit 12, a muscle synergy analysis unit 13, an output assistive pattern generation unit 14, an output unit 15, a storage unit 16, and an exercise device 30.

[0052] The simulation unit 11 inputs assistance patterns, which combine phase and assistance force in a time series, into a musculoskeletal model to generate a set of electromyographic data for each assistance pattern. It includes a body movement data measurement unit 11a, a muscle strength data input unit 11b, a musculoskeletal model generation unit 11c, and an assistance force adjustment unit 11d. The body movement data measurement unit 11a measures the user's body movement data related to alternative movement, and in this embodiment, motion capture is used. The muscle strength data input unit 11b accepts the user's muscle strength data. The musculoskeletal model generation unit 11c generates the user's musculoskeletal model 8 from the body movement data and muscle strength data. The assistance force adjustment unit 11d applies assistance to the musculoskeletal model 8 by changing the assistance force for each phase of alternative movement. Thus, the simulation unit 11 in the exercise assistance device 1 of Example 2 differs from the simulation unit 11 in the alternative exercise determination device 1a of Example 1, which performs simulations regarding the exercise to be determined, in that it performs simulations regarding alternative exercises. However, other than this point, it is the same as in Example 1.

[0053] The electromyography (EMG) data input unit 12 receives input of a first EMG data set for each phase during the target exercise and a second EMG data set for each phase during the auxiliary pattern. In this embodiment, the EMG sensor 20 receives input of the first EMG data set for each phase during the target exercise, and the simulation unit 11 receives input of the second EMG data set for each phase during the auxiliary pattern. As the first EMG data set, an EMG data set from a healthy person's walking exercise was used. The muscle synergy analysis unit 13 analyzes the muscle synergies from the input first electromyographic data set and the second electromyographic data set. The electromyographic data input unit 12 or muscle synergy analysis unit 13 has the same configuration as the electromyographic data input unit 12 or muscle synergy analysis unit 13 in the alternative exercise determination device 1a of Example 1.

[0054] The output auxiliary pattern generation unit 14 compares the first muscle synergy analyzed from the first electromyography data group with the second muscle synergy analyzed from the second electromyography data group. It extracts the auxiliary pattern assigned to the second electromyography data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or in which the expression of the second muscle synergy with the highest correlation coefficient is observed, and generates it as an output auxiliary pattern. The output unit 15 outputs an auxiliary output pattern to the exercise device that performs the substitute movement. In this embodiment, the auxiliary output pattern is output to the exercise device 30, but it is not limited to a specific exercise device and can be output to a variety of devices. The memory unit 16 stores the generated output auxiliary patterns, but also data such as the type of movement, body movement data, muscle strength data, musculoskeletal model 8, auxiliary patterns, data input or generated in the simulation unit 11, the first or second electromyography data set, muscle synergy analysis results, and the equipment used for measurement.

[0055] Exercise assistance device 1 is a device that assists in the rehabilitation of patients using ergometer 3, with walking motion in healthy individuals as the target exercise and pedaling motion in stroke hemiplegia patients as the alternative exercise. Figure 4 shows a schematic diagram of the configuration of the exercise assistance device of Example 2. As shown in Figure 4, the exercise assistance device 1 of Example 2 is realized by a computer 2 and an ergometer 3. In Figure 4, the computer 2 and the ergometer 3 are connected by a wired cable 18, but communication may also be performed wirelessly. Ergometer 3 functions as an exercise device 30. Ergometer 3 is a recumbent ergometer with a backrest attached to the chair, making it a safe ergometer to use with a low risk of the patient falling during pedaling. Pedal 3a is equipped with a pedal strap to keep the foot in close contact with the pedal, so that the patient's foot is always on the pedal during pedaling.

[0056] (Independent left and right pedaling mechanism) Conventional ergometers have linked cranks on both sides, and the user pedals with both feet. However, when stroke patients with hemiplegia use an ergometer, they rely most of their movement on their normal limbs, making it unsuitable for training the paralyzed limbs. Therefore, there is a need for an exercise device that encourages patients to spontaneously move their paralyzed limbs.

[0057] Therefore, the ergometer 3 in this embodiment has an independent left and right pedaling mechanism. Although not shown in detail, each pedal is equipped with an actuator that generates load / assistance force and an encoder that measures the rotation angle. Both pedals are controlled as master-slave, with the pedal on the patient's paralyzed limb side as the master and the pedal on the healthy limb side as the slave. As a result, the ergometer 3 operates in the same way as a conventional ergometer when the patient tries to pedal from the paralyzed limb side. On the other hand, when the patient tries to pedal from the healthy limb side, the pedal becomes a slave and cannot be pedaled. Therefore, the patient will pedal from the paralyzed limb side, and it is expected that the effectiveness of normal gait training will be improved. The output assistance pattern generation unit 14 generates output assistance patterns in which assistance force is applied individually to the left and right pedals 3a of the ergometer 3.

[0058] To provide positive or negative assistive force in response to the patient's paralyzed limb's pedaling force, the ergometer 3 employs a control law on the master side that combines position control with damping control, as shown in equations 6 and 7 below. This controller operates at a constant speed in space and has the property of pushing with a constant force when it comes into contact with an external object. Therefore, it can be said that this is a safe control law for ergometer 3, where the patient's foot is always on the pedal. On the other hand, the slave side employs position control that maintains a phase of 180 degrees with the master side, as shown in equation 7 below. As a result, both pedals behave as if they were normal pedals, but only the master side can be used by the post-stroke hemiplegic patient to pedal at will, while the slave side cannot move the master pedal.

[0059]

number

[0060]

number

[0061] In the above equation 6, θ master θ is the angle of the master pedal. slave This is the angle of the slave pedal. user τ is the pedaling force exerted by the master pedal (i.e., the lower limb being trained). ref is the assist force command value of the master pedal (see the explanation of assist force in Figure 9 below). B and K are arbitrary constants that define the strength of the damping control and the strength of the position control. This controller operates according to the final rotational speed command value of the master pedal and the final rotational speed command value of the slave pedal derived in Equation 6 above. Equation 7 above shows that the position command value of the slave pedal maintains a phase of 180 degrees with the angle of the master pedal.

[0062] Next, the method of using the exercise assistance device 1 will be described in accordance with the flow of the exercise assistance method of this embodiment. Figure 5 shows a schematic flow diagram of the exercise assistance method of Embodiment 2. As shown in Figure 5, the exercise assistance method of Embodiment 2 is a method in which a computer provides positive or negative assistive force to a user who is performing an alternative exercise instead of a target exercise, thereby generating muscle activity equivalent to that of the target exercise, and comprises the following steps.

[0063] (Simulation step) The body movement data measurement unit 11a measures the user's body movement data related to the alternative movement (step S01). Next, the muscle strength data input unit 11b accepts the user's muscle strength data (step S02). The musculoskeletal model generation unit 11c generates the user's musculoskeletal model from the body movement data and muscle strength data (step S03). The assist force adjustment unit 11d applies assist force to the musculoskeletal model by changing the assist force for each phase of the alternative movement (step S04). The simulation unit 11 inputs assist patterns, which combine phase and assist force in a time series, into the musculoskeletal model and generates a second electromyographic data set for each assist pattern (step S05). At this point, the simulation unit 11 determines whether the input of all assist patterns has been completed (step S06), and if not, it again applies assist force to the musculoskeletal model by changing the assist force for each phase of the alternative movement (step S04).

[0064] Figure 6 shows an explanatory diagram of body movement data measurement and pedal angle. In measuring the user's body movement data, as shown in Figure 6, the subject's feet are placed on the pedals 3a of the ergometer 3, and the pedals 3a are rotated in the direction indicated by the arrows. Since motion capture is used as the body movement data measurement unit 11a, markers are attached to parts of the subject's body that are characteristic of human movement, such as joints, and the position and movement of the markers are measured in three dimensions, converted into data, and recorded. In addition, the assistive force is adjusted in correspondence with the rotation angle of the pedals 3a shown in Figure 6.

[0065] (Electromyography data input step) Once all auxiliary patterns have been entered, the electromyographic data input unit 12 accepts input of the first electromyographic data set for each phase in the target movement and the second electromyographic data set for each phase in the auxiliary patterns (step S07). Note that the input of electromyographic data sets to the electromyographic data input unit 12 may be performed each time an electromyographic data set is generated.

[0066] (Muscle synergy analysis step) The muscle synergy analysis unit 13 analyzes the muscle synergies from the input first electromyography data set and the second electromyography data set (step S08). In this embodiment, the first electromyography data set is a set of electromyography data from a healthy subject during walking. Figure 7 is a graph showing the results of muscle synergy analysis during walking in a healthy subject, where (1) is the weighting coefficient (W) of each muscle and (2) is the time-series change (H) of the muscle synergy. The period (%) on the horizontal axis of Figure 7(2) is defined as a series of walking movements, such as adjusting the position of the foot in preparation for landing, the landing foot supporting the body weight, kicking the foot backward relative to the direction of movement, lifting the kicked foot, and moving the foot forward. As shown in Figure 7(1) or (2), three muscle synergies are extracted in the order of synergy 3, synergy 1, and synergy 2 during walking in a healthy subject. Synergy 1 is observed in the movement of kicking the ground backward, synergy 2 in the movement of kicking the foot backward and bringing it forward, and synergy 3 in the movement of lowering the foot to the ground and bearing weight.

[0067] (Auxiliary pattern generation step for output) The output auxiliary pattern generation unit 14 compares the first muscle synergy analyzed from the first electromyography data group with the second muscle synergy analyzed from the second electromyography data group, extracts the auxiliary pattern given to the second electromyography data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or in which the expression of the second muscle synergy with the highest correlation coefficient is observed, and generates it as an output auxiliary pattern (step S09). Figure 8 shows the results of muscle synergy analysis during simulation, comparing the first and second muscle synergies and illustrating an example of a second electromyographic data set where the second muscle synergy, whose correlation coefficient exceeds a predetermined threshold, is observed relative to the first muscle synergy. Alternatively, a second electromyographic data set showing the second muscle synergy with the highest correlation coefficient may be used. Figure 8(1) shows the weight coefficient (W) for each muscle, and Figure 8(2) shows the time-series change (H) of muscle synergies. The period (%) on the horizontal axis of Figure 8(2) is defined as a series of pedaling actions: adjusting the pedal to the direction where force is easiest to apply, moving the foot forward by pedaling, applying force to the pedal by pedaling, returning the foot in time with the returning pedal, and moving the foot forward again. As shown in Figure 8(2), the expression of muscle synergies can be confirmed in the order of synergy 3, synergy 1, and synergy 2. This shows a similar trend to the muscle synergy analysis results during walking in healthy subjects shown in Figure 7, indicating that alternative exercises can serve as substitutes for target exercises.

[0068] Figure 9 is a graph showing the pedaling force and output assistance patterns of the subject during pedaling motion, derived from musculoskeletal simulation. Specifically, the assistance force shown by the dashed line in Figure 9 is generated as an output assistance pattern by extracting the assistance pattern given to the second electromyographic data set (see Figure 8) in which the expression of the second muscle synergy was observed, with the correlation coefficient exceeding a predetermined threshold for the first muscle synergy. A positive load value indicates that a load force is being applied to the pedal, meaning a negative assist force is being provided. Conversely, a negative load value indicates that a support force is being applied to the pedal, meaning a positive assist force is being provided. Since simply placing your foot on the pedal applies a load, a large pedaling force suggests that you are using your lower limb muscles to pedal. On the other hand, when there is no significant load on the pedal, it is thought that the system's positive assist force is driving the pedal. As shown in Figure 9, in the Synergy 3 section from 0° to 60°, the system is set to provide support to guide the rider into a more comfortable pedaling position; in the Synergy 1 section from 60° to 230°, the load is set to be stronger to induce stronger lower limb pedaling force; and in the Synergy 2 section from 230° to 360°, the system is set to provide support to guide the rider's feet forward.

[0069] (Output step) The output unit 15 outputs an output assistance pattern to the exercise device 30 that performs the alternative movement (step S10: output step). In this embodiment, the output assistance pattern shown in Figure 9 is output to the ergometer 3, and the user, who is a subject, performs a pedaling motion. Figure 10 shows the results of muscle synergy analysis during pedaling exercise in a user subject. Figure 10(1) shows the weight coefficient (W) of each muscle, and Figure 10(2) shows the time-series change (H) of muscle synergy. The period (%) on the horizontal axis of Figure 10(2) is the same as in Figure 8(2). As shown in Figure 10(2), the expression of muscle synergy can be confirmed in the order of synergy 3, synergy 1, and synergy 2. This shows a similar trend to the muscle synergy analysis results during walking exercise in a healthy subject shown in Figure 7, indicating that alternative exercise can serve as a substitute for the target exercise. Thus, the exercise assistance device of Example 2 can provide the user with appropriate assistance, enabling the user to independently train to acquire voluntary movements necessary for normal walking during the recovery period, without the support of a physical therapist. [Examples]

[0070] In this example, an experiment was conducted using the exercise assistance device 1 of Example 2 with subjects who were post-stroke hemiplegia patients. Unlike Example 2, in this experiment, the first electromyography data group used was the electromyography data group during walking exercise of post-stroke hemiplegia patients. Two subjects, both post-stroke hemiplegia patients, participated in the experiment. Subject A was 45 years old (age of onset 32), male, 165 cm tall, weighed 65 kg, and had paralysis on the right side. Subject B was 53 years old (age of onset 49), female, 158 cm tall, weighed 60 kg, and had paralysis on the left side. Furthermore, the Brunnstrom stages, which are used as a criterion for predicting the recovery process by evaluating the symptoms of hemiplegia in the upper limbs, hands, feet, and lower limbs on a 6-point scale, were as follows: Subject A: upper limbs 4, lower limbs 5; Subject B: upper limbs 2, lower limbs 3-4.

[0071] In the experiment, subjects were first asked to walk independently, and muscle synergies were extracted during the walking experiment. Furthermore, subjects practiced pedaling using ergometer 3 to acquire muscle synergies. Subject A also practiced pedaling using the comparative ergometer to compare the effectiveness of ergometer 3 (example) and the conventional ergometer (comparative example). Subject B, who was severely ill, did not participate in this for safety reasons. The experiment was conducted under the supervision of a physical therapist.

[0072] Figures 11-13 show the muscle synergy analysis results for subjects with mild symptoms, while Figure 14 or 15 shows the results for subjects with severe symptoms. Figures 11 and 14 show the muscle synergy analysis results during walking, Figures 12 and 15 show the muscle synergy analysis results for the example during pedaling, and Figure 13 shows the muscle synergy analysis results for the comparative example during pedaling. In the figures, (1) represents the weighting coefficient (W) for each muscle, and (2) represents the time-series change in muscle synergy (H). As shown in Figure 11, three muscle synergies were extracted when Subject A walked. The muscles constituting these muscle synergies and the timing of their generation were similar to the results of the muscle synergy analysis during walking in a healthy subject, as shown in Figure 7. This is based on the fact that Subject A had mild hemiplegia and was able to walk independently.

[0073] When subject A performed pedaling exercises using ergometer 3, three muscle synergies similar to those induced during walking were induced, as shown in Figure 12(1) or (2). On the other hand, as shown in Figure 13(1) or (2), when the comparative ergometer was used, muscle synergies similar to those induced during walking were not induced. From this, it can be concluded that the exercise assistance device 1 using ergometer 3 is an effective alternative to conventional walking training.

[0074] On the other hand, as shown in Figure 14(1) or (2), Subject B only exhibits synergy 2 muscle coordination during walking. Subject B is a patient who cannot walk independently without a cane, without making large movements of their paralyzed limbs. From this, it can be seen that Subject B walks using a principle completely different from normal walking, and has symptoms of severe hemiplegia.

[0075] When Subject B performed pedaling exercises using Ergometer 3, two muscle synergies (Synergy 1 and Synergy 2) were extracted, as shown in Figure 15(1) or (2). In particular, Synergy 1, which corresponds to the basic movement of forcefully pushing off with the foot, was significantly exhibited, indicating that the subject was spontaneously attempting to perform the pedaling motion and that the exercise assistance device 1 was actively guiding the muscle activity necessary for the pedaling motion using appropriate assistive force. Synergy 2, which represents basic movements such as pulling the foot back to adjust the foot position, was continuously and weakly exhibited, meaning that Subject B was unable to move the lower limbs skillfully. However, Subject B was able to continue pedaling with the paralyzed limb, which is thought to be due to the effective assistive force of the exercise assistance device 1. Based on the above, the exercise assistance device 1 can enable pedaling training using only the muscle strength of subject B's paralyzed limb, and can elicit lower limb muscle activity in a pattern close to normal walking. If subject B undergoes long-term rehabilitation training using the exercise assistance device 1, there is a high probability that subject B's normal walking function will improve. [Examples]

[0076] Figure 16 shows a functional block diagram of the exercise assistance device of Embodiment 4. As shown in Figure 16, the exercise assistance device 1b is a device that determines whether a target exercise is substitutable for a target exercise, and if it is usable as an alternative exercise, provides a positive or negative assistive force to the user performing the alternative exercise to generate muscle activity equivalent to that of the target exercise, and comprises a simulation unit 11, an electromyographic data input unit 12, a muscle synergy analysis unit 13, an output assist pattern generation unit 14, an output unit 15, a storage unit 16, an alternative exercise determination unit 17, and an exercise device 30. Unlike the alternative movement determination device 1a of Embodiment 1, the exercise assistance device 1b includes an output auxiliary pattern generation unit 14, an output unit 15, and an exercise device 30. Furthermore, unlike the exercise assistance device 1 of Embodiment 2, the exercise assistance device 1b includes an alternative movement determination unit 17. In other words, the exercise assistance device 1b is a device that has the functions of both the alternative movement determination device 1a of Embodiment 1 and the exercise assistance device 1 of Embodiment 2. With this configuration, it is possible to determine whether the movement to be determined can be used as a substitute for the target movement, and then generate an appropriate auxiliary pattern for the user, enabling the user to perform an alternative movement that can be used as a substitute for the target movement. [Industrial applicability]

[0077] This invention is useful for rehabilitation, the transmission of craftsmanship skills, and sports instruction. [Explanation of symbols]

[0078] 1,1b Exercise aids 1a Alternative motion determination device 2 Computers 3 Ergometer 3a Pedal 4a~4d Walking motion 5a~5d Pedaling motion 6a-6e, 7a-7e Muscle Synergies 8 Musculoskeletal Models 11. Simulation Department 11a Body motion data measurement unit 11b Muscle strength data input section 11c Musculoskeletal Model Generation Unit 11d Auxiliary force adjustment section 12. Electromyography Data Input Unit 13. Muscle Synergy Analysis Department 14. Output auxiliary pattern generation unit 15 Output section 16 Memory section 17 Alternative exercise determination section 18 Wired Cable 20 Electromyography Sensors 30 Exercise Devices

Claims

1. A device that provides positive or negative assistive force to a user performing an alternative exercise instead of a target exercise, thereby generating muscle activity equivalent to that of the target exercise. 1) A body movement data measurement unit that measures the user's body movement data related to alternative exercises; a muscle strength data input unit that receives input of the user's muscle strength data; a musculoskeletal model generation unit that generates a musculoskeletal model of the user from the body movement data and the muscle strength data; an assist force adjustment unit that changes the assist force for each phase of the alternative exercise and applies the assist force to the musculoskeletal model; and a simulation unit that inputs an assist pattern combining the phase and assist force in a time series to the musculoskeletal model and generates a group of electromyographic data for each of the assist patterns. 2) An electromyographic data input unit that receives input of a first electromyographic data set for each phase in the target movement and a second electromyographic data set for each phase in the auxiliary pattern, 3) A muscle synergy analysis unit that analyzes the muscle synergies from the input first electromyography data set and second electromyography data set, 4) An output auxiliary pattern generation unit that compares the first muscle synergy analyzed from the first electromyographic data group with the second muscle synergy analyzed from the second electromyographic data group, extracts the auxiliary pattern given to the second electromyographic data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or in which the expression of the second muscle synergy with the highest correlation coefficient is observed, and generates it as an output auxiliary pattern, 5) An output unit that outputs the output auxiliary pattern to an exercise device that performs alternative movement, An exercise assistance device characterized by being equipped with [a specific feature].

2. The exercise assist device according to claim 1, further comprising a storage unit for storing at least the output auxiliary patterns.

3. The exercise assistance device according to claim 1 or 2, characterized in that the first electromyographic data set is a set of electromyographic data taken during walking in a healthy person or a patient with hemiplegia after a stroke.

4. The exercise device is an ergometer having independent left and right pedals. The exercise assist device according to claim 1 or 2, further comprising the aforementioned exercise device.

5. The exercise assist device according to claim 4, characterized in that the output assist pattern generation unit generates an output assist pattern in which an assisting force acts individually on the left and right pedals of the exercise device.

6. A device for determining whether a target movement is substitutable for a movement to be judged, 1) A motion data measurement unit that measures the user's motion data related to the exercise to be judged; a muscle strength data input unit that receives input of the user's muscle strength data; a musculoskeletal model generation unit that generates a musculoskeletal model of the user from the motion data and the muscle strength data; an assist force adjustment unit that changes the assist force for each phase of the exercise to be judged and applies the assist force to the musculoskeletal model; a simulation unit that inputs an assist pattern combining the phase and assist force in a time series to the musculoskeletal model and generates a group of electromyographic data for each of the assist patterns; 2) An electromyographic data input unit that receives input of a first electromyographic data set for each phase in the target movement and a second electromyographic data set for each phase in the auxiliary pattern, 3) A muscle synergy analysis unit that analyzes the muscle synergies from the input first electromyography data set and second electromyography data set, 4) A substitute exercise determination unit compares the first muscle synergy analyzed from the first electromyographic data group with the second muscle synergy analyzed from the second electromyographic data group, and determines that the exercise under evaluation is a substitute exercise for the target exercise if the proportion of the second electromyographic data group in which the expression of the second muscle synergy, whose correlation coefficient exceeds a predetermined threshold with respect to the first muscle synergy, exceeds a predetermined threshold. An alternative motion determination device characterized by comprising the following:

7. A device that determines whether a target exercise is substitutable for a given exercise, and if it is usable as a substitute exercise, provides positive or negative assistive force to the user performing the substitute exercise to generate muscle activity equivalent to that of the target exercise. 1) A motion data measurement unit that measures the user's motion data related to the exercise to be judged; a muscle strength data input unit that receives input of the user's muscle strength data; a musculoskeletal model generation unit that generates a musculoskeletal model of the user from the motion data and the muscle strength data; an assist force adjustment unit that changes the assist force for each phase of the exercise to be judged and applies the assist force to the musculoskeletal model; a simulation unit that inputs an assist pattern combining the phase and assist force in a time series to the musculoskeletal model and generates a group of electromyographic data for each of the assist patterns; 2) An electromyographic data input unit that receives input of a first electromyographic data set for each phase in the target movement and a second electromyographic data set for each phase in the auxiliary pattern, 3) A muscle synergy analysis unit that analyzes the muscle synergies from the input first electromyography data set and second electromyography data set, 4) A substitute exercise determination unit compares the first muscle synergy analyzed from the first electromyographic data group with the second muscle synergy analyzed from the second electromyographic data group, and determines that the exercise under evaluation is a substitute exercise that is substitutable for the target exercise if the proportion of the second electromyographic data group in which the expression of the second muscle synergy, whose correlation coefficient exceeds a predetermined threshold with respect to the first muscle synergy, exceeds a predetermined threshold. 5) When the exercise to be judged is determined to be an alternative exercise, the first muscle synergy analyzed from the first electromyography data group and the second muscle synergy analyzed from the second electromyography data group are compared, and the auxiliary pattern given to the second electromyography data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or the second muscle synergy with the highest correlation coefficient is observed, is extracted and generated as an output auxiliary pattern, an output auxiliary pattern generation unit, 6) An output unit that outputs the output auxiliary pattern generated for an exercise device that performs alternative movement, An exercise assistance device characterized by being equipped with [a specific feature].

8. A method for providing positive or negative assistive force to a user performing an alternative exercise instead of a target exercise, thereby generating muscle activity equivalent to that of the target exercise, 1) A simulation step which involves measuring the user's body movement data related to alternative exercises, receiving input of the user's muscle strength data, generating a musculoskeletal model of the user from the body movement data and muscle strength data, applying assistive force to the musculoskeletal model by changing the assistive force for each phase of the alternative exercise, inputting an assistive pattern that combines the phase and assistive force in a time series into the musculoskeletal model, and generating a set of electromyographic data for each of the assistive patterns, 2) An electromyographic data input step that receives input of a first electromyographic data set for each phase in the target movement and a second electromyographic data set for each phase in the auxiliary pattern, 3) A muscle synergy analysis step in which the muscle synergies of the first and second electromyographic data sets input are analyzed, 4) An output auxiliary pattern generation step in which the first muscle synergy analyzed from the first electromyography data group and the second muscle synergy analyzed from the second electromyography data group are compared, and the auxiliary pattern given to the second electromyography data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or in which the expression of the second muscle synergy with the highest correlation coefficient is observed is extracted and generated as an output auxiliary pattern, 5) An output step that outputs the generated auxiliary output pattern to an exercise device that performs alternative movement, A method of assisting exercise, characterized by comprising:

9. A program that provides positive or negative assistive force to users performing alternative exercises instead of target exercises, thereby generating muscle activity equivalent to that of the target exercise. An exercise assistance program that instructs a computer to perform the following steps 1) to 5): 1) A simulation step in which measurement data obtained by measuring the user's body movement data related to alternative exercises and the user's muscle strength data are input, a musculoskeletal model of the user is generated from the body movement data and the muscle strength data, the assisting force is changed for each phase of the alternative exercise and the assisting force is applied to the musculoskeletal model, and an assisting pattern obtained by combining the phase and the assisting force in a time series is input to the musculoskeletal model to generate a group of electromyographic data for each of the assisting patterns, 2) An electromyographic data input step that receives input of a first electromyographic data set for each phase in the target movement and a second electromyographic data set for each phase in the auxiliary pattern, 3) A muscle synergy analysis step in which the muscle synergies of the first and second electromyographic data sets input are analyzed, 4) An output auxiliary pattern generation step in which the first muscle synergy analyzed from the first electromyography data group and the second muscle synergy analyzed from the second electromyography data group are compared, and the auxiliary pattern given to the second electromyography data group in which the correlation coefficient with respect to the first muscle synergy exceeds a predetermined threshold, or in which the expression of the second muscle synergy with the highest correlation coefficient is observed is extracted and generated as an output auxiliary pattern, 5) An output step that outputs the generated auxiliary output pattern to an exercise device that performs alternative movement.

Images