JP6620329B2 - Electrical stimulation device, method for calibrating electrode probability matrix of electrical stimulation device, and muscle displacement sensor calibration method for sensor system - Google Patents

Description

The present invention relates to an electrical stimulation device that applies an electrical stimulation signal to a human arm muscle and detects the displacement of the arm muscle with a sensor, a method for calibrating an electrode probability matrix of the electrical stimulation device, and a muscle displacement sensor of the sensor system. The present invention relates to a calibration method .

  Conventionally, an attempt has been made to move a user's finger or hand in response to an external command by attaching a plurality of electrodes to the forearm of a person (user) and applying an electrical stimulation signal from the electrodes to the muscles of the forearm. It has been broken. For example, it is considered that finger rehabilitation, training, movement assistance, and the like are performed by an external command. In addition, using a head-mounted display, virtual reality processing (Virtual Reality) that presents virtual space images to the user, and augmented reality processing (Argumented Reality) that superimposes virtual object images on real space images It has also been proposed to improve reality by moving a finger in accordance with an image in a virtual space or the like in response to an instruction from the outside.

  The inventors of the present application have previously proposed an electrical stimulation apparatus as described in Patent Document 1. The electrical stimulation apparatus proposed in Patent Document 1 is an apparatus that applies electrical stimulation to muscles of the forearm by attaching a plurality of electrodes to a band worn on the user's forearm.

JP 2014-104241 A

  A user wearing the electrical stimulation device proposed in Patent Document 1 moves his / her finger or hand by applying electrical stimulation to the muscles of the forearm according to an external command. For example, it is known in advance which muscle of the forearm to move when five fingers are stimulated, and the finger corresponding to the muscle moves by stimulating a specific muscle with an electrode attached to the electrical stimulator. .

Here, it is assumed that the user wears a band-shaped electrical stimulation device on the forearm. At this time, the band-shaped electrical stimulation device includes a plurality of electrodes, but it is necessary to detect which muscle each electrode corresponds to. In order to perform this detection, when a user wears a band-shaped electrical stimulation device, a calibration operation is first performed to output an electrical stimulation signal from each electrode and confirm in advance which finger moves. I need it.
That is, when the user wears the band-shaped electrical stimulation device, it is necessary for the electrical stimulation device to accurately know in advance which muscle each electrode disposed on the electrical stimulation device faces. For this reason, calibration work is extremely important.

  On the other hand, the thickness of a person's arm varies greatly depending on individual differences such as age, sex, and weight. For this reason, in order to absorb individual differences in the thickness of the arm with only one electrical stimulation device, it is necessary to provide electrodes and sensors that exceed the number of five fingers. Then, an electrode that does not correspond to the movement of the finger is generated in the electrode. That is, in order for the electrical stimulation apparatus to operate correctly, it is necessary to clarify in advance the calibration relationship between the muscles that actually move the finger and the electrodes.

The present invention has been made in view of such problems, is not affected by the state worn on the user's arm and individual differences, clarify the correspondence between the movement of the fingers and the electrodes in a short time, and there are very few malfunctions, It is an object of the present invention to provide an electrical stimulation device capable of driving a target finger with high accuracy, a method for calibrating an electrode probability matrix of the electrical stimulation device, and a muscle displacement sensor calibration method for a sensor system .

In order to solve the above problems, an electrical stimulation device of the present invention detects a band wound around a user's forearm and a plurality of muscles disposed on one surface of the band and present on the user's arm, A plurality of muscle displacement sensors, a sensor multiplexer for selecting one muscle displacement sensor from the plurality of muscle displacement sensors, and a surface of the band on which the plurality of muscle displacement sensors are arranged are adjacent to the plurality of muscle displacement sensors. A plurality of electrodes. Furthermore, a short-distance wireless transmission unit that converts a signal related to the displacement of the muscle of the user's arm obtained from the muscle displacement sensor into digital data and transmits it to the host, and a multiplexer for selecting one electrode from a plurality of electrodes And an electrode probability matrix in which the correlation between the finger movement and the plurality of electrodes is described with probability. Further, a short-range wireless receiving unit that receives a command to move a finger from the host in the normal mode and receives an updated electrode probability matrix from the host in the calibration mode, and a maximum probability from the electrode probability matrix based on the command to move the finger And a finger electrode corresponding conversion unit that controls the electrode multiplexer and selects the specified electrode.
In the electrode probability matrix, the host converts the data of the muscle displacement sensor into the movement of the user's finger, detects the relative positional relationship between the current user's arm and the electrical stimulator, and detects the electrode probability matrix. The elements are updated after rearranging the elements.

The present invention, without being affected by the conditions and individual differences mounted on the user's arm, to clarify the relationship between the motion and the electrode finger in a short time, malfunctions is extremely small, electricity can drive the objective fingers with high precision A stimulation device, a method of calibrating an electrode probability matrix of an electrical stimulation device, and a muscle displacement sensor calibration method of a sensor system can be provided.
Problems, configurations, and effects other than those described above will be clarified by the following description of embodiments.

It is an external appearance perspective view of the electrical stimulation apparatus which is an example of embodiment of this invention. It is a top view of an electrode arrangement | positioning surface. It is a figure which shows the state immediately before mounting | wearing a forearm with an electric stimulator, and the state immediately after. It is a schematic diagram which shows an electrical stimulation system which has an electrical stimulation apparatus which is an example of the usage pattern of an electrical stimulation apparatus. It is a block diagram which shows the hardware constitutions of a host. It is a block diagram which shows the hardware constitutions of an electrical stimulation apparatus. It is a block diagram which shows the software function of the electrical stimulation apparatus and host in normal mode. It is a block diagram which shows the software function of the electrical stimulator and host in a calibration mode. It is a time chart which shows the flow of the calibration operation | movement which an electric stimulator and a host perform in a calibration mode. It is a flowchart explaining the flow of operation | movement of the first learning mode performed by a host when a user wears an electric stimulator for the arm for the first time. It is a time chart explaining the operation | movement of an electrical stimulation, the contraction state of a muscle, and a muscle displacement sensor. An explanatory diagram of the finger behavior matrix showing the movement of the finger with respect to the electrical stimulus, a diagram showing the procedure for selecting an effective electrical stimulus from the finger behavior matrix, and a flag matrix showing the result of selecting an effective electrical stimulus from the finger behavior matrix It is explanatory drawing and the figure which shows the electrode probability matrix produced | generated from the flag matrix. It is a flowchart explaining the flow of operation | movement of the learning mode after the 2nd time performed by a host, when a user wears an electrical stimulation apparatus after the 2nd time. It is a flowchart explaining the flow of operation | movement of the learning mode after the 2nd time performed by a host when a user wears an electrical stimulator after the 2nd time. A schematic diagram for explaining the arrangement relationship between arm muscles and electrodes when the user first wears the electrostimulator on the arm, and the arm of the arm when the user puts the electrostimulator on the arm again. It is a typical schematic diagram for explaining the arrangement relationship between muscles and electrodes. An example of a finger behavior matrix created in the second and subsequent learning modes, a virtual electrode probability matrix created based on this finger behavior matrix, and an electrode probability matrix before applying the learning mode stored in the host FIG. 4 is a diagram showing a matrix in which this electrode probability matrix is rearranged. It is a figure which shows an example of the generalized electrode probability matrix.

FIG. 1 is an external perspective view of an electrical stimulation apparatus 100 that is an example of an embodiment of the present invention.
The electrical stimulation device 100 includes a V-shaped band 101. The band 101 is made of a flexible resin sheet such as silicone rubber. Both wing parts of the band 101 are inclined by equal inclination angles θ1 and θ2 from the horizontal line L105. The inclination angles θ1 and θ2 are, for example, 32 °. A rectangular circuit storage box 103 is provided at the center of the surface side of the band 101. The circuit storage box 103 incorporates an arithmetic processing unit 150 (see FIG. 4) described later, a secondary battery, and the like.

  A first serial interface terminal 104 is provided on the side surface of one short side of the circuit storage box 103. The first serial interface terminal 104 is, for example, a microUSB terminal. The electrical stimulation device 100 charges a secondary battery (not shown) incorporated therein through the first serial interface terminal 104. Further, it is possible to expand functions such as connecting the first serial interface terminal 104 to a personal computer or the like and updating firmware as a component of the arithmetic processing unit.

The back surface (the lower surface in FIG. 1) opposite to the surface on which the circuit storage box 103 of the band 101 is attached is an electrode arrangement surface 100a which will be described later with reference to FIG.
The electrical stimulation device 100 is worn by the user by winding an electrode placement surface 100a, which is the back surface of the band 101, around the user's forearm, as will be described in the wearing example of FIGS. 3A and 3B described later.

FIG. 2 is a plan view of the electrode arrangement surface 100a.
The electrode placement surface 100a includes electrodes 201 to 208 for applying an electrical stimulation signal to the muscles of the user's forearm, and electrodes 211 to 213, 216, and 217 that are ground electrodes used in pairs with the electrodes 201 to 208. Is provided. However, since the ground electrode is commonly used for a plurality of opposing electrodes, the numbers of the electrodes 201 to 208 and the electrodes 211 to 213, 216, and 217 do not match.
In addition, the electrode placement surface 100a includes muscle displacement sensors 221 to 228 that detect the movement of muscles of the user's forearm.

  A right electrode arrangement location 231 is provided on the right side (left side in the drawing) of the electrode arrangement surface 100a, and four electrodes 201, 202, 211, and 212 are arranged in the right electrode arrangement location 231. Of the four electrodes 201, 202, 211, 212, the electrode 201 is a first electrode, and the electrode 202 is a second electrode. The electrode 211 is a ground electrode facing the electrode 201, and the electrode 212 is a ground electrode facing the electrode 202.

The electrodes 201 and 211 are electrodes for stimulating the forearm muscles, and are arranged adjacent to the longitudinal direction L of the arm when worn.
The electrodes 202 and 212 are also electrodes for stimulating the muscles of the forearm, and are substantially rectangular electrodes arranged in an inclined state at an inclination angle θ2 with respect to the circumferential direction H. The electrode 202 and the electrode 212 are disposed adjacent to the longitudinal direction L of the arm.

  A central electrode arrangement location 232 is provided at the center of the electrode arrangement surface 100 a, and five electrodes 203, 204, 205, 208, and 213 are arranged at the central electrode arrangement location 232. Of the five electrodes 203, 204, 205, 208, and 213, the electrode 203 is a third electrode, the electrode 204 is a fourth electrode, and the electrode 205 is a fifth electrode. These three electrodes 203, 204, and 205 extend in the longitudinal direction of the arm and are arranged side by side in substantially parallel to the circumferential direction of the arm. The electrode 208 is an eighth electrode. The electrode 208 is an electrode that extends long in the circumferential direction of the arm. The electrode 213 is a ground electrode commonly used facing the electrodes 203, 204, 205, and 208.

The electrode 203, the electrode 204, and the electrode 205 are electrodes for stimulating different muscles of the forearm for each channel, and the electrode 213 is commonly used as a ground electrode. The three electrodes 203, 204, 205 are arranged side by side in the circumferential direction H of the arm, and the electrode 213, which is a common ground electrode arranged adjacent to the three electrodes 203, 204, 205 in the longitudinal direction L, A rectangular electrode extending in the circumferential direction H of the arm.
The electrode 208 is a rectangular electrode that is adjacent to the electrode 213 and extends long in the circumferential direction H of the arm. The electrode 213 is also used as the ground potential of the electrode 208. Note that the electrode 208 is a spare electrode, and the electrode 208 extends long in the circumferential direction H of the arm, so that a plurality of muscles of the arm can be simultaneously stimulated.

  A left electrode arrangement portion 233 is provided on the left side portion 102 (right side in FIG. 2) of the electrode arrangement surface 100a, and four electrodes 206, 207, 216, and 217 are arranged in the left electrode arrangement portion 233. The Of the four electrodes 206, 207, 216, and 217, the electrode 206 is a sixth electrode, and the electrode 207 is a seventh electrode. The electrode 216 is a ground electrode facing the electrode 206, and the electrode 217 is a ground electrode facing the electrode 207.

The electrodes 206 and 216 are electrodes for stimulating the muscles of the forearm, and are substantially rectangular electrodes arranged in an inclined state at the same angle θ1 as the inclination angle θ1 of the left side portion 102 with respect to the circumferential direction H.
The electrodes 207 and 217 are electrodes that stimulate the forearm muscles and are disposed adjacent to the longitudinal direction L of the arm when worn.

  Muscle displacement sensors 221 and 222 are arranged at two locations in the vicinity of the right electrode arrangement location 231 on the electrode arrangement surface 100a. Muscle displacement sensors 223, 224, 225, and 226 are arranged at four locations in the vicinity of the central electrode arrangement location 232 on the electrode arrangement surface 100a. Muscle displacement sensors 227 and 228 are arranged at two positions in the vicinity of the left electrode arrangement place 233 on the electrode arrangement surface 100a.

The eight muscle displacement sensors 221 to 228 are well-known photo reflectors. Each of these muscle displacement sensors includes infrared light emitting elements 221a to 228a and infrared light receiving elements 221b to 228b (see FIG. 6), and detects changes in the distance from the muscle displacement sensor placement surface to the muscle surface of the arm. To do. The infrared light emitting elements 221a to 228a are, for example, near infrared LEDs, and the infrared light receiving elements 221b to 228b are, for example, phototransistors.
When the muscle contracts, the distance between the photo reflector and the surface portion of the muscle fluctuates due to the bulge that occurs in the portion of the skin where the muscle exists. The photoreflector detects the intensity of near-infrared reflected light caused by the change in distance with a phototransistor. Near-infrared rays have the property of transmitting through the skin surface and are therefore suitable for detecting the state of muscle bulge.

  In addition, a resin material (not shown) having adhesiveness is disposed at portions other than the right electrode arrangement portion 231, the central electrode arrangement portion 232, and the left electrode arrangement portion 233 on the electrode arrangement surface 100 a, and the adhesive property of the resin material Thus, the electrode placement surface 100a can be worn around the forearm.

[Installation example of electrical stimulation device 100]
FIG. 3A is a diagram showing a state immediately before the electrical stimulation device 100 is mounted on the forearm.
FIG. 3B is a diagram illustrating a state immediately after the electrical stimulation device 100 is mounted on the forearm.
As shown in FIG. 3A, the center portion of the electrode placement surface 100a (FIG. 2) of the band 101 is in contact with a portion near the wrist of the forearm RA of the user's right arm. At this time, as shown in FIG. Further, the circuit storage box 103 at the center of the substantially V-shaped band 101 is set to face the palm side.

Then, the user performs an operation of winding both wings of the band 101 around the wrist as indicated by arrows F1 and F2.
In this way, as shown in FIG. 3B, the electrical stimulation device 100 is worn while being wound around the forearm RA. At this time, the state of being wound around the forearm RA is maintained by the adhesiveness of the adhesive resin material arranged on the electrode arrangement surface 100a.
Note that the state of being wrapped around the forearm RA only by the adhesiveness of the resin material is merely an example. For example, some kind of clip mechanism may be provided at both ends of the band 101 so that the two remain overlapped.

Thus, the electrical stimulation apparatus 100 can be easily mounted because the band 101 is wound around the forearm RA. Since the band 101 is substantially V-shaped, the user can easily understand the mounting direction, and can be securely mounted in the determined direction as shown in FIG. 3B.
Although FIG. 3 shows an example in which the electrical stimulation device 100 is worn on the user's right arm, the electrical stimulation device 100 may be worn on the left arm.

Incidentally, as illustrated in FIGS. 3A and 3B, the electrical stimulation device 100 according to the embodiment of the present invention is wound around a portion of the user's forearm RA near the wrist. However, at this time, the forearm RA is not provided with an index that allows the electrical stimulation device 100 to be positioned at the home position of the user's forearm RA. That is, every time the user wears the electrical stimulation device 100 on the forearm RA, the wearing position often deviates slightly. Then, each time the user wears the electrical stimulation device 100 on the forearm, the relative positional relationship between the electrode provided on the electrode placement surface 100a of the electrical stimulation device 100, the muscle displacement sensor, and the muscle of the user's arm is changed. It will shift.
The present invention has been conceived for the purpose of correctly understanding the correspondence between the electrode and the movement of the finger in response to this “deviation”.

[Usage pattern of electrical stimulation device 100]
FIG. 4 is a schematic diagram illustrating an electrical stimulation system 400 having the electrical stimulation device 100, which is an example of a usage pattern of the electrical stimulation device 100.
The electrical stimulation device 100 has a short-range wireless communication function such as Bluetooth (registered trademark) described later. On the other hand, the electrical stimulation device 100 is connected to a personal computer with a short-range wireless communication function equivalent to that of the electrical stimulation device 100 or a peripheral device that provides a short-range wireless communication function. Establish two-way communication with. Hereinafter, a personal computer that establishes two-way communication with the electrical stimulation apparatus 100 through short-range wireless communication is referred to as a host 401. In FIG. 4, a short-range wireless communication unit 402 is connected to the host 401 and establishes bidirectional communication with the electrical stimulation apparatus 100.

In the host 401, an application program such as a shooting game is running. Then, in response to the user's operation on the application program, the host 401 transmits a command for applying electrical stimulation to the user's predetermined muscle to the electrical stimulation apparatus 100 through short-range wireless communication. The electrical stimulation device 100 applies electrical stimulation to a desired muscle of the user based on the electrical stimulation command received from the host 401.
Further, the electrical stimulation device 100 transmits the displacement information of the muscles of the user's arm to the host 401 as digital data by the above-described muscle displacement sensor.

  By the way, the driving of the muscle displacement sensor is accompanied by relatively large power consumption, including the light emission driving of the infrared LED. Therefore, the application program of the host 401 is configured to drive the muscle displacement sensor in the electrical stimulation device 100 according to the state of the application program in order to acquire the displacement information of the arm muscles from the electrical stimulation device 100 with the minimum power consumption. Send a command to stop. That is, while executing the application program, the host 401 does not cause the electrical stimulation device 100 to drive the muscle displacement sensor until the application program is in a situation that requires information on the displacement of the user's arm muscles. When the application program reaches a situation where the user's arm muscle displacement information is required, a command for driving the muscle displacement sensor is transmitted from the host 401 to the electrical stimulation device 100 at that time. In response to the command from the host 401, the electrical stimulation device 100 drives the muscle displacement sensor to acquire the displacement information of the arm muscles.

The host 401 transmits a command to the electrical stimulation apparatus 100 to stop driving the muscle displacement sensor when the acquisition of the displacement information of the arm muscles required by the application program is completed. In response to the command from the host 401, the electrical stimulation device 100 stops driving the muscle displacement sensor.
In other words, the electrical stimulation device 100 functions as an input device that collects displacement information of the user's arm muscles and an output device that applies displacement to the arm muscles to the host 401. It can also be said that the electrical stimulation device 100 is a terminal for the host 401 and / or an application program.

  As described with reference to FIG. 2, the electrode arrangement surface 100 a of the electrical stimulation device 100 has eight electrodes excluding the ground electrode. On the other hand, a human hand has five fingers. This is because the number of electrodes exceeding the number of five fingers is provided in order to absorb individual differences in the thickness of a person's arm with only one electrical stimulation device 100. That is, the eight electrodes include electrodes that do not correspond to finger movements. In addition, depending on the state in which the electrical stimulation device 100 is worn on the user's arm, the position of the muscle relative to the electrode may often shift due to the displacement of the wearing position. In order to allow the electrodes to correspond to muscles even if such a shift occurs, it is preferable to provide more electrodes than the number of five fingers.

In addition, in order for the electrical stimulation apparatus 100 to operate correctly, it is necessary to clarify the correspondence relationship between the muscle that moves the actual finger, the electrode, and the muscle displacement sensor by a calibration operation.
Therefore, as the operation mode of the electrical stimulation apparatus 100 according to the embodiment of the present invention, the normal mode that operates as a terminal of the application program, and the correspondence between the muscle that moves the actual finger, the electrode, and the muscle displacement sensor are clarified. There are two types of operation modes: a calibration mode for performing a calibration operation for the purpose.
In addition, in the block diagram explaining the software function mentioned later, it demonstrates by dividing into normal mode and calibration mode. Of these, the present invention relates to a calibration mode.

[Hardware configuration of host 401]
FIG. 5 is a block diagram illustrating a hardware configuration of the host 401.
As described above, the host 401 formed of a general personal computer includes the CPU 501, ROM 502, RAM 503, nonvolatile storage 504, display unit 505, operation unit 506, and short-range wireless communication unit 402 connected to the bus 507. The short-range wireless communication unit 402 is hardware for performing short-range wireless communication with the electrical stimulation device 100. The nonvolatile storage 504 stores an OS and an application program for operating the personal computer as the host 401 of the electrical stimulation apparatus 100.

[Hardware Configuration of Electrical Stimulator 100]
FIG. 6 is a block diagram illustrating a hardware configuration of the electrical stimulation apparatus 100.
The CPU 602, ROM 603, RAM 604, A / D converter 605, and second serial interface 606 (abbreviated as “second serial I / F” in FIG. 6) connected to the bus 601 are connected to a well-known one-chip microcomputer 607. Constitute.
The anodes of the infrared light emitting elements 221a, 222a,... 228a, which are infrared LEDs constituting the muscle displacement sensors 221, 222,... 228, are connected to the power supply voltage node + Vcc. The cathodes of the infrared light emitting elements 221a, 222a... 228a are connected to one end of a current limiting resistor R609 through a first multiplexer 608. The other end of the current limiting resistor R609 is grounded.

  The collectors of infrared light receiving elements 221b, 222b,... 228b, which are phototransistors constituting the muscle displacement sensors 221, 222,... 228, are connected to a power supply voltage node + Vcc. The emitters of the infrared light receiving elements 221b, 222b,... 228b are connected to the A / D converter 605 through the second multiplexer 610 and grounded through the resistors R611a, R611b,.

  The first multiplexer 608 and the second multiplexer 610 receive control signals from the second serial interface 606 and are periodically switched and controlled, so that the A / D converter 605 has eight muscle displacement sensors in time division. Voltage signals 221, 222,... 228 are input. The first multiplexer 608 and the second multiplexer 610 can be collectively referred to as a sensor multiplexer that selects one of the plurality of muscle displacement sensors 221, 222,.

A well-known 6-axis sensor 612 and a short-range wireless communication unit 613 are also connected to the bus 601 of the one-chip microcomputer 607, and posture information and acceleration information output from the 6-axis sensor 612 are obtained through the A / D converter 605. Are transmitted to the host 401 through the short-range wireless communication unit 613 together with the information of the eight muscle displacement sensors 221, 222.
A first serial interface 614 (abbreviated as “first serial I / F” in FIG. 6) is further connected to the bus 601 of the one-chip microcomputer 607. The first serial interface 614 is used to supply power to a storage battery (not shown) and to update firmware stored in the ROM 603.

Further, a booster circuit 615 including a known choke coil, a capacitor, and a transistor switch is connected to the second serial interface 606. Then, a rectangular wave pulse signal having a voltage substantially equal to the power supply voltage + Vcc at 100 kHz, for example, is supplied from the second serial interface 606 to the booster circuit 615. This rectangular wave pulse signal controls on / off of a transistor switch (not shown) in the booster circuit 615.
The voltage of the rectangular wave pulse signal is boosted twice by the booster circuit 615. The electrical stimulation voltage output from the booster circuit 615 is supplied to the electrodes 201, 202... 208 through the PWM switch 616 and the third multiplexer 617.

The PWM switch 616 is controlled by the second serial interface 606, and applies PWM modulation to the electrical stimulation voltage boosted by the booster circuit 615. Since the duty ratio of the electrical stimulation voltage is changed by PWM modulation, the electrical stimulation voltage applied to the muscle is changed to a desired voltage. The third multiplexer 617 is also controlled through the second serial interface 606, selects an electrode specified in the command received from the host 401 through the short-range wireless communication unit 613, and a PWM-modulated electrical stimulation voltage is applied to the electrode. The
The third multiplexer 617 can also be referred to as an electrode multiplexer that selects one of the plurality of electrodes 201, 202.

[Software Functions of Electrical Stimulation Device 100 and Host 401 in Normal Mode]
FIG. 7 is a block diagram illustrating software functions of the electrical stimulation apparatus 100 and the host 401 in the normal mode.
Electrostimulation device 100 to the host 401, a function of transmitting user's muscle change and electrostimulation device 100 itself arm posture and acceleration to the host 401, in accordance with instructions received from the host 401, the user's arm muscles This is an input / output terminal device having a function of applying electrical stimulation.
That is, analog signals output from the eight muscle displacement sensors 221 to 228 are converted into muscle displacement information by the A / D converter 605, and together with the posture information and acceleration information output from the six-axis sensor 612, the input / output control unit 701. And transmitted to the host 401 through the short-range wireless transmission unit 702.

When the host 401 receives muscle displacement information, posture information, and acceleration information from the electrical stimulation device 100 through the short-range wireless reception unit 711, the host 401 supplies these to the input / output control unit 712. The input / output control unit 712 supplies the muscle displacement information, posture information, and acceleration information received from the electrical stimulation apparatus 100 to the information processing unit 713 that is a predetermined application program such as a game, and the information processing unit 713 outputs the information. Based on the drawing information, predetermined screen drawing information is output to the display unit 505. Further, the input / output control unit 712 transmits the electrical stimulation information output from the information processing unit 713 to the electrical stimulation device 100 through the short-range wireless transmission unit 714.

When the finger electrode correspondence conversion unit 703 of the electrical stimulation apparatus 100 receives the electrical stimulation execution command output from the host 401 through the short-range wireless reception unit 704, the finger electrode correspondence conversion unit 703 refers to the electrode probability matrix 705 stored in the RAM 604. Then, the finger number specified in the command is converted into the electrode number, the PWM switch 616 and the third multiplexer 617 are controlled, and the electrical stimulation voltage is applied to the desired electrodes 201 to 208.
Details of the electrode probability matrix 705 will be described later with reference to FIG.

  The input / output control unit 712 that controls the operation timing of the first multiplexer 608 and the second multiplexer 610 that performs the switching operation of the muscle displacement sensors 221 to 228 and the operation timing of the third multiplexer 617 that performs the electrode switching operation are controlled. The finger electrode corresponding conversion unit 703 is completely asynchronous. For this reason, in FIG. 7, it is illustrated as separate functional blocks.

[Software Functions of Electrical Stimulator 100 and Host 401 in Calibration Mode]
FIG. 8 is a block diagram illustrating software functions of the electrical stimulation apparatus 100 and the host 401 in the calibration mode.
The difference between the software functions of the electrical stimulation apparatus 100 and the host 401 shown in FIG.
<1> The function of the 6-axis sensor 612 that is unnecessary in the calibration mode is stopped.
<2> The input / output control unit 712 of the host 401 stores sensor values in the sensor value storage unit 801 provided in the RAM 503 or the non-volatile storage 504, creates an electrode probability matrix 705, and updates it. Transmitting an electrode probability matrix 705 to the electrical stimulator 100;
<3> The electrical stimulation device 100 applies electrical stimulation to the electrode based on a command transmitted from the host 401, and scans the muscle displacement sensor in synchronization therewith.

In particular, because of the above <3>, the input / output control unit 712 of the electrical stimulation device 100 activates a built-in timer 803 after applying electrical stimulation to the electrode based on a command transmitted from the host 401. Then, after a predetermined time has elapsed, the muscle displacement sensor is scanned.
On the other hand, the input / output control unit 712 of the host 401 acquires the calculation result of the probability calculation unit 802 based on the information of the muscle displacement sensors 221 to 228 received from the electrical stimulation device 100, creates the electrode probability matrix 705, Or update. Then, the electrode probability matrix 705 that has been created or updated is transmitted to the electrical stimulation device 100 through the short-range wireless transmission unit 714. Further, the input / output control unit 712 displays the progress of the calibration mode and the like on the display unit 505 with a predetermined message or the like during the operation of the calibration mode.

[Software Operation of Host 401 in Calibration Mode]
The electrode probability matrix 705 is matrix data indicating the correspondence between the movement of the electrode and the finger corresponding to the muscle that moves the user's finger. FIG. 17 shows a generalized example of the electrode probability matrix 705. The vertical (row) of the electrode probability matrix 705 represents the electrode, and the horizontal (column) represents the movement of the finger. Each element stores a Bayesian posterior probability based on Bayesian estimation.

  In the normal mode, when receiving a command to move a desired finger from the host 401, the finger electrode correspondence conversion unit 703 of the electrical stimulation apparatus 100 refers to the electrode probability matrix 705 based on the finger information designated by the host 401. That is, the electrode probability matrix 705 designated by the host 401 corresponding to the finger information is viewed. Then, the probabilities for each electrode are stored in the elements constituting the row. Among these elements, the electrode corresponding to the element showing the highest probability is the electrode most likely to move the finger. As described above, the finger electrode correspondence conversion unit 703 converts the finger specified by the host 401 into an electrode number, and controls the third multiplexer 617 using this information.

When an electrical stimulation voltage is applied to an electrode, a predetermined muscle is stimulated and a finger corresponding to the muscle moves. Then, a muscle displacement sensor detects the displacement of the muscle, and refers to correspondence data stored in the sensor value storage unit 801 to determine which finger is bent. That is, the relationship between the finger movement and the electrode is set to 1: 1.
In FIG. 2, it has been described that the electrode arrangement surface 100a of the electrical stimulation device 100 is provided with eight electrodes excluding the ground electrode and eight muscle displacement sensors. As described above, the reason for providing eight electrodes and eight muscle displacement sensors for five fingers is to absorb individual differences in the thickness of a person's arm with only one electrical stimulation device 100. Because. Then, an electrode that does not correspond to the movement of the finger is generated in the electrode. That is, in order for the electrical stimulation apparatus 100 to operate correctly, it is necessary to clarify the correspondence between the actual finger movement and the electrode by calibration work.

FIG. 9 is a time chart showing the flow of the calibration operation executed by the electrical stimulation device 100 and the host 401 in the calibration mode.
When the electrical stimulation device 100 is worn on the user's arm, the muscle displacement sensor detects that the user's skin is close to the detection region of the muscle displacement sensor. When the input / output control unit 712 of the electrical stimulation device 100 detects that the electrical stimulation device 100 is worn on the user's arm (S901), the host 401 is requested to establish communication through the short-range wireless communication unit 613. (S902). The host 401 receives a communication request from the electrical stimulation device 100 and returns a status message indicating the establishment of communication (S903). The electrical stimulation device 100 receives the status message from the host 401, and returns a status message indicating that the establishment of communication is recognized to the host 401 (S904).

  When the input / output control unit 712 of the host 401 receives a status message indicating that the establishment of communication is recognized from the electrical stimulation apparatus 100 through the short-range wireless reception unit 704, the input / output control unit 712 starts the calibration process of the muscle displacement sensor. First, the moving image data “stationary state guide video” stored in the nonvolatile storage 504 of the host 401 is reproduced and displayed on the display unit 505. And the command which collects the data of a muscle displacement sensor is transmitted to the electrical stimulation apparatus 100 (S905). “Stand-by state guidance video” is video data that displays an illustration of a hand with no power applied with the electrostimulator 100 attached to the upper arm and a message “Please wait with no force applied to the hand” It is. When receiving the command from the host 401, the electrical stimulation device 100 returns data of all muscle displacement sensors to the host 401 (S906). The value of the muscle displacement sensor at this time is the value of the muscle displacement sensor in a state where the user is not putting force on the finger, and is a value serving as a basis for detecting whether or not force is applied to the muscle.

  When the host 401 receives data of the eight muscle displacement sensors from the electrical stimulation device 100, the host 401 stores the data as “stationary state data” (S907). Next, the moving image data “first operation guide video” stored in the nonvolatile storage 504 of the host 401 is reproduced and displayed on the display unit 505. Then, a command for collecting the data of the muscle displacement sensor is transmitted to the electrical stimulation device 100 (S908). “First motion guide video” refers to an illustration of a hand with the electric stimulator 100 attached to the upper arm with the thumb bent and a message saying “Please wait with the thumb bent” It is data. When receiving the command from the host 401, the electrical stimulation apparatus 100 returns data of all muscle displacement sensors to the host 401 (S909). The value of the muscle displacement sensor at this time is the value of the muscle displacement sensor when the user puts force on the thumb.

  Upon receiving the data of the eight muscle displacement sensors from the electrical stimulation device 100, the host 401 stores this as “first motion state data” (S910). Next, the moving image data “second operation guidance video” stored in the nonvolatile storage 504 of the host 401 is reproduced and displayed on the display unit 505. And the command which collects the data of a muscle displacement sensor is transmitted to the electrical stimulation apparatus 100 (S911). The “second motion guide video” is a video that displays an illustration of a hand with the electrical stimulation device 100 attached to the upper arm, with the index finger bent, and a message “Please wait with the index finger bent” It is data.

After step S911, operations corresponding to steps S908, S909, and S910 are performed a total of eight times from step S908. At that time, moving image data reproduced by the host 401 and data received by the host 401 from the electrical stimulation device 100 and stored in the sensor value storage unit 801 are as follows.
“First motion guidance video”: motion guidance with the thumb bent, and first motion status data are stored.
“Second motion guidance video”: Stores motion guidance with the index finger bent and second motion status data.
“Third motion guide video”: Stores motion guide with the middle finger bent and third motion state data.
“Fourth operation guide video”: Stores the operation guide and fourth operation state data when the ring finger or little finger is bent.
“Fifth motion guidance video”: Stores motion guidance and fifth motion state data in a state where the wrist is bent straight toward the palm with the hand straightened (palm flexion).
“Sixth motion guidance video”: Stores motion guidance and sixth motion state data in a state where the wrist is bent straight toward the back of the hand (dorsiflexion) with the hand straightened.
“Seventh motion guide video”: Stores motion guide and seventh motion state data in a state where the wrist is bent in the direction of the thumb with the hand extended straight (flexion).
“Eighth motion guidance video”: Stores motion guidance and eighth motion state data in a state where the wrist is bent in the direction of the little finger with the hand stretched straight (scaled).

Upon receiving the data of the eight muscle displacement sensors from the electrical stimulation device 100, the host 401 stores this as “seventh motion state data” (S912). Next, the moving image data “eighth operation guide video” stored in the nonvolatile storage 504 of the host 401 is reproduced and displayed on the display unit 505. And the command which collects the data of a muscle displacement sensor is transmitted to the electrical stimulator 100 (S913). The “eighth motion guide video” is an illustration of a hand with the electrostimulator 100 attached to the upper arm, with the hand stretched straight and the wrist bent in the direction of the little finger (scaled), Is a moving image data that displays a message “Please wait with your wrist straightened and your wrist bent in the direction of the little finger”. When receiving the command from the host 401, the electrical stimulation device 100 returns data of all muscle displacement sensors to the host 401 (S914). The value of the muscle displacement sensor at this time is the value of the muscle displacement sensor in a crooked state in which the user bends the wrist in the direction of the little finger.
When the host 401 receives the data of the muscle displacement sensor from the electrical stimulation device 100, the host 401 stores it as “eighth motion state data” (S915).
As described above, from step S905 to step S915, the host 401 has grasped the correspondence between the change in the value of the muscle displacement sensor and the movement of the finger.

When step S915 ends, the input / output control unit 712 of the host 401 calculates the relative value of each muscle displacement sensor. Specifically, the stationary state data is subtracted from the first operating state data to obtain a first reference value relating to a state where the thumb is bent. The stationary state data is subtracted from the second operation state data to obtain a second reference value relating to the state where the index finger is bent. In the same manner, the third reference value relating to the bent middle finger, the fourth reference value relating to the bent or little finger bent, the fifth reference value relating to the palm flexion state, and the sixth reference value relating to the dorsiflexion state. , The seventh reference value relating to the bent state, and the eighth reference value relating to the bent state. These first to eighth reference values are a set of relative values of the eight muscle displacement sensors. The input / output control unit 712 of the host 401 stores these first to eighth reference values in the sensor value storage unit 801.
Next, the input / output control unit 712 of the host 401 multiplies the first to eighth reference values by a predetermined ratio to obtain a threshold value. Here, the ratio of multiplying the first to eighth reference values is, for example, 50%. The input / output control unit 712 of the host 401 also stores these threshold values in the sensor value storage unit 801.

  The intensity of reflected light detected by the phototransistor of the muscle displacement sensor depends on factors such as the condition of the user's skin and muscle relative to the muscle displacement sensor and the relative positional relationship between the muscle displacement sensor and the user's skin. It varies greatly depending on the element. Therefore, the input / output control unit 712 of the host 401 stores the values of the muscle displacement sensors in the state in which the user hardly applies force to the finger and the state in which the user has bent a specific finger, as the sensor value storage unit 801. And the difference is calculated. Since the obtained reference value is a relative variation value of the muscle displacement sensor, the influence of uncertain factors can be eliminated.

An A / D converter 605 that digitizes an analog signal output from the muscle displacement sensor is, for example, a 10-bit unsigned integer (0 to 1023). When the inventors created the electrical stimulation apparatus 100 experimentally, it was found that the relative fluctuation value obtained from the muscle displacement sensor that detected the displacement of the muscle was approximately 300 to 900.
After step S916, which will be described later, all the data of the muscle displacement sensor received by the host 401 from the electrical stimulation device 100 is converted into a relative value of the muscle displacement sensor by subtracting the stationary state data. Then, the input / output device compares the relative value of the muscle displacement sensor with a threshold value and determines whether or not a predetermined finger has moved.

From step S905 to step S915, the host 401, which has grasped the correspondence between the change in the value of the muscle displacement sensor and the finger movement, next enters an operation for grasping the correspondence between the electrode and the finger movement. .
The input / output control unit 712 of the host 401 transmits a command to collect data of the muscle displacement sensor after applying the electrical stimulation voltage to the first electrode to the electrical stimulation device 100 (S916).
When the electrical stimulation device 100 receives a command from the host 401, it applies an electrical stimulation voltage to the first electrode (S917), collects data of the muscle displacement sensor after a predetermined time has elapsed, and sends this back to the host 401 ( S918).
When receiving the muscle displacement sensor data from the electrical stimulation device 100, the input / output control unit 712 of the host 401 stores the data in the sensor value storage unit 801 as sensor value data for the first electrode (S919). Next, the input / output control unit 712 of the host 401 transmits a command to collect data of the muscle displacement sensor after applying the electrical stimulation voltage to the second electrode to the electrical stimulation device 100 (S920).

After step S920, operations corresponding to steps S916, S917, S918, and S919 are performed a total of eight times from step S916. At that time, the command that the host 401 transmits to the electrical stimulation device 100 and the data that the host 401 receives from the electrical stimulation device 100 and stores in the sensor value storage unit 801 are as follows.
Command for collecting muscle displacement sensor data after applying electrical stimulation voltage to the first electrode: Store sensor value data at the first electrode.
Command to collect muscle displacement sensor data after applying electrical stimulation voltage to the second electrode: store sensor value data at the second electrode.
Command to collect muscle displacement sensor data after applying electrical stimulation voltage to the third electrode: Store sensor value data at the third electrode.
Command for collecting muscle displacement sensor data after applying electrical stimulation voltage to the fourth electrode: Store sensor value data at the fourth electrode.
Command for collecting muscle displacement sensor data after applying electrical stimulation voltage to the fifth electrode: Store sensor value data at the fifth electrode.
Command for collecting muscle displacement sensor data after applying electrical stimulation voltage to the sixth electrode: Store sensor value data at the sixth electrode.
Command for collecting muscle displacement sensor data after applying electrical stimulation voltage to the seventh electrode: Store sensor value data at the seventh electrode.
Command for collecting muscle displacement sensor data after applying electrical stimulation voltage to the eighth electrode: Store sensor value data at the eighth electrode.

When receiving the muscle displacement sensor data from the electrical stimulation device 100, the input / output control unit 712 of the host 401 stores the data in the sensor value storage unit 801 as sensor value data for the seventh electrode (S921). Next, the input / output control unit 712 of the host 401 transmits a command for collecting data of the muscle displacement sensor after applying the electrical stimulation voltage to the eighth electrode to the electrical stimulation device 100 (S922).
When the electrical stimulation device 100 receives a command from the host 401, it applies an electrical stimulation voltage to the eighth electrode (S923), collects the data of the muscle displacement sensor after a predetermined time has elapsed, and sends this back to the host 401 ( S924).

When receiving the data of the muscle displacement sensor from the electrical stimulation device 100, the input / output control unit 712 of the host 401 stores this in the sensor value storage unit 801 as sensor value data for the eighth electrode (S925). Next, the input / output control unit 712 of the host 401 stores the first to eighth operation state data corresponding to the movement of each finger and the steps S916 to S925 stored in the series of operations from step S905 to S915. The sensor value data at each electrode stored or generated in a series of operations is generated or updated.
If the electrode probability matrix 705 exists in the host 401 as a result of the execution of the first learning mode (described later in FIG. 10 and subsequent figures), the electrode probability matrix 705 is also updated through the probability calculation unit 802. Do. The input / output control unit 712 of the host 401 transmits the generated or updated electrode probability matrix 705 to the electrical stimulation device 100 (S926). Then, the electrical stimulation apparatus 100 stores the electrode probability matrix 705 received from the host 401 in the RAM 503 (S927), and ends a series of processing.

Steps S905 to S915 in FIG. 9 are muscle displacement sensor calibration modes for clarifying the correlation between the muscle displacement sensor and the movement of the finger. When the input / output control unit 712 of the host 401 executes the muscle displacement sensor calibration mode, it becomes clear from the data obtained from the muscle displacement sensors 221 to 228 which finger is currently moving.
Then, steps S916 to S927 surrounded by a broken line in FIG. 9 are learning modes in which the correlation between the electrode and the finger movement is clarified by creating and updating the electrode probability matrix 705. . When the input / output control unit 712 of the host 401 executes the learning mode, it becomes clear to which electrode of the electrodes 201 to 208 the electrical stimulation voltage should be applied in order to move a desired finger.
That is, the calibration mode includes a muscle displacement sensor calibration mode and a learning mode. Hereinafter, the learning mode will be described with reference to FIG.

FIG. 10 is a flowchart showing the flow of the operation in the first learning mode executed by the host 401 when the user first wears the electrical stimulation device 100.
When the processing is started (S1001), the input / output control unit 712 of the host 401 first initializes the counter variable i to 1 and initializes all elements of the electrode probability matrix 705 to “0” (S1002).
The rest is a loop. The input / output control unit 712 transmits a command for collecting data of the muscle displacement sensor after applying the electrical stimulation voltage to the i-th electrode to the electrical stimulation device 100 (S1003). Then, a difference value is calculated from the data of the muscle displacement sensor received from the electrical stimulation device 100, and compared with a threshold value, it is checked whether or not there has been a finger movement (S1004). If it is determined that any finger has moved due to electrical stimulation by the i-th electrode (YES in S1005), the input / output control unit 712 corresponds to the x-th finger whose movement is detected in the electrode probability matrix 705. “1” is stored in the i-th column element corresponding to the i-th electrode in the x row (S1006).

After executing step S1006, or when it is determined in step S1005 that no finger is moved by electrical stimulation by the i-th electrode (NO in S1005), the input / output control unit 712 sets the counter variable i to the maximum of i. Check the value, ie whether the total number of electrodes has been reached. If the counter variable i has not reached the total number of electrodes (NO in S1007), the input / output control unit 712 increments the counter variable i by 1 (S1008) and repeats the processing from step S1003 again.
In step S1007, when the counter variable i has reached the total number of electrodes (YES in S1007), the input / output control unit 712 stores the created electrode probability matrix 705 in the nonvolatile storage 504 and the electrical stimulation apparatus 100. (S1009), and a series of processing ends (S1010).

Subsequently, the operation of step S1003 will be described.
FIG. 11 is a time chart for explaining the electrical stimulation, the contraction state of the muscle, and the operation of the muscle displacement sensor.
In FIG. 11, (A) an electrical stimulus applied to the electrode from above, (B) a muscle contraction state, (C) a gate signal in the input / output control unit 712 indicating the operation period of the muscle displacement sensor, (D) (E) operation period of the second muscle displacement sensor, (F) operation period of the third muscle displacement sensor, (G) operation period of the seventh muscle displacement sensor, H) The operation period of the eighth muscle displacement sensor. (B) The vertical axis represents the amount of displacement of muscle contraction only in the contraction state of muscle, and all other values are logical values.

At time T1101, when an electrical stimulation voltage is applied to the electrodes, the muscle begins to contract. From the time T1102 when the muscle contraction is stabilized, the gate signal indicates logic true, and in response, the scan of the muscle displacement sensor starts. The time required for data collection of one muscle displacement sensor is about several msec to several tens msec. At the time T1103 when scanning of all the muscle displacement sensors is completed, the logic of the gate signal is inverted to false, and at the same time, the application of the electrical stimulation voltage to the electrodes is completed.
The operation | movement which applies an electrical stimulation voltage to an electrode and scans a muscle displacement sensor shown in FIG. 11 is performed with respect to all eight electrodes.

  Human muscle contracts when an electrical stimulation voltage is applied from the outside through electrodes. At that time, it takes about 0.1 second until the muscle contracts to a predetermined contraction amount and stabilizes after the electrical stimulation voltage is applied. In the electrical stimulation apparatus 100 according to the embodiment of the present invention, a margin time of 0.2 seconds is provided in anticipation of the margin. The margin time is from time T1101 when the electrical stimulation voltage shown in FIG. 11 is applied to time T1102 when the muscle contraction is stabilized.

Next, the electrode probability matrix 705 created by the processing from steps S1003 to S1008 will be described.
FIG. 12A is an explanatory diagram of a matrix showing the movement of fingers with respect to electrical stimulation. Hereinafter, this matrix is referred to as a finger behavior matrix.
FIG. 12B is a diagram illustrating a procedure for selecting an effective electrical stimulus from the finger behavior matrix.
FIG. 12C is an explanatory diagram of a matrix showing a result of selecting an effective electrical stimulus from the finger behavior matrix. This matrix is called a flag matrix.
FIG. 12D is a diagram showing an electrode probability matrix 705 generated from the flag matrix.

  In step S1004, the input / output control unit 712 calculates a difference value of the muscle displacement sensor. Then, a difference value of the muscle displacement sensor corresponding to the movement of the finger is derived. When there is only one muscle displacement sensor corresponding to a certain finger movement, the difference value is adopted as it is. In the case of a combination of two or more muscle displacement sensors corresponding to a certain finger movement, the average value of the difference values of each muscle displacement sensor is adopted. In this way, numerical values as shown in FIG. 12A are stored as elements of the finger behavior matrix.

The row of the finger behavior matrix shown in FIGS. 12A and 12B, the row of the flag matrix shown in FIG. 12C, and the row of the electrode probability matrix 705 shown in FIG. 12D are as follows from the top.
First line: The thumb is bent, that is, the thumb is bent.
Second line: the index finger is bent, that is, the index finger is bent.
Third line: the middle finger is bent, that is, the middle finger is bent.
Fourth line: A state in which the ring finger or the little finger is bent, that is, the ring finger or the little finger is bent.
5th line: A state where the wrist is bent in the direction of the palm with the hand stretched straight (palm flexion), that is, a wrist palm flexion state.
6th line: A state in which the wrist is bent in the direction of the back of the hand (dorsiflexion) in a state where the hand is straightened out, that is, a wrist dorsiflexion state.
Seventh line: a state where the wrist is bent in the direction of the thumb with the hand straightened (flexion), that is, the wrist is flexed.
8th line: A state where the wrist is bent in the direction of the little finger with the hand straightened out (flexion), that is, the wrist is bent.
The column of the finger behavior matrix shown in FIGS. 12A and 12B, the column of the flag matrix shown in FIG. 12C, and the column of the electrode probability matrix 705 shown in FIG. 12D are from the first electrical stimulus to the eighth electrical stimulus. It is.

Next, the procedure for determining whether or not the finger has moved in step S1004 in FIG. 10 will be described.
Focusing on the elements from the 1st row and the 1st column to the 8th row and 1st column of the finger behavior matrix shown in FIG. 12B, it is composed of elements “595 115 92 0 0 0 0 0” in order from the top. Among the elements in this row, the element indicating the maximum value is “595” in one row and one column. This value is stored in the maximum value array 1201. The maximum value array 1201 is an array that stores the maximum value of each column.
The position of the element “595” is (1, 1) of the finger behavior matrix, and the row (one row) corresponds to the thumb flexion state. Therefore, it is determined whether or not this value “595” exceeds the thumb flexion state threshold. As a result of the determination, since it is determined that the threshold value is exceeded, it is assumed that the element in the 1st row and 1st column of the finger behavior matrix is valid, and the logical truth is stored in the flag array 1202. In FIG. 12B, it is marked with “◯”.
In this way, the maximum value is selected for each row of the finger behavior matrix and stored in the maximum value array 1201. And it compares with the threshold value in the operation | movement of the finger to which the element belongs from the position of these elements. As a result of the comparison, if the maximum value is a value equal to or greater than the threshold value, the logical true is stored in the flag array 1202. Then, with respect to the position of the element of the maximum value array 1201 in which the logical truth is added to the flag array 1202, the element existing at the same position in the flag matrix is regarded as logical true. This is the flag matrix of FIG. 12C.
For an element whose flag matrix is true in logic, the probability is 100% (= 1). This is the electrode probability matrix 705 of FIG. 12D.

The electrode probability matrix 705 is a matrix indicating which hand moves with a certain probability when an electrical stimulation voltage is applied to a certain electrode. The probability stored as an element of the electrode probability matrix 705 is a posterior probability such as Bayesian estimation. That is, the creation of the electrode probability matrix 705 in the flowchart of FIG. 10 is the first learning of Bayesian estimation.
At the time when the flowchart of FIG. 10 is executed, since it is still the first learning, only 0 or 1 exists. These values fluctuate in the second and subsequent learning modes described below.

FIG. 13 and FIG. 14 are flowcharts showing the flow of operations in the second and subsequent learning modes that are executed by the host 401 when the user wears the electrical stimulation device 100 for the second and subsequent times.
When the processing is started (S1301), the input / output control unit 712 of the host 401 first initializes the counter variable j to 1 and initializes the flag variable MFflag to logic false (S1302). The flag variable MFflag is a flag for recording that a state in which a finger has moved due to electrical stimulation has occurred.

  The rest is a loop. The input / output control unit 712 transmits a command to collect data of the muscle displacement sensor after applying the electrical stimulation voltage to the jth electrode to the electrical stimulation device 100 (S1303). Then, a difference value is calculated from the data of the muscle displacement sensor received from the electrical stimulation device 100, and compared with a threshold value, it is checked whether or not there has been a finger movement (S1304). Steps S1303 and S1304 have the same processing contents as steps S1003 and S1004 of FIG.

  If it is determined that any finger has moved due to electrical stimulation by the j-th electrode (YES in S1305), the input / output control unit 712 then determines whether or not the flag variable MFflag is logic false, that is, Confirm whether or not the finger has moved for the first time at this point. If it is determined that the finger has moved for the first time at this time (YES in S1306), the electrode probability matrix 705 stored in the nonvolatile storage 504 of the host 401 is read into the RAM 503, and the electrode probability matrix 705 on the RAM 503 is read. Are moved for each row and / or for each column (S1307). In this way, by moving the elements of the electrode probability matrix 705, the relative positional relationship between the electrodes and the muscles of the electrical stimulation device 100 currently worn on the user's arm is reflected in the electrode probability matrix 705. .

  Next, the input / output control unit 712 operates the probability calculation unit 802, and the Bayesian posterior probability that the finger did not move with respect to the elements of the electrode probability matrix 705 corresponding to the first to j-1th electrodes. “P (x | j)” is calculated, and the corresponding element of the electrode probability matrix 705 is updated (S1308). Then, the flag variable MFflag is changed to logic true (S1309), and the process proceeds to step S1410 in FIG. In other words, after it is determined in step S1005 that any finger has moved (YES in S1005), the MFflag has changed to logic true in step S1006, so the processing in steps S1307, S1308, and S1309 is performed. First, the process proceeds to step S1410 in FIG.

Next, moving elements of the electrode probability matrix 705 in step S1307 will be described in detail with reference to FIGS. 15A and 15B.
FIG. 15A is a schematic diagram for explaining an arrangement relationship between arm muscles and electrodes when the user first wears the electrical stimulation apparatus 100 on the arm.
FIG. 15B is a schematic diagram for explaining an arrangement relationship between arm muscles and electrodes when the user puts the electrical stimulation apparatus 100 on the arm again.
In FIG. 15A, the electrode 1501 is away from the muscle 1502. An electrode 1503 is close to the muscle 1502. An electrode 1505 is close to the muscle 1504. Note that the electrode 1506 is a ground electrode commonly used by the electrode 1501, the electrode 1503, and the electrode 1505.

Comparing FIG. 15A and FIG. 15B, in FIG. 15B, the relative positional relationship between the user's arm and the electrode arrangement surface 100a of the electrical stimulation device 100 is shifted. For this reason, the electrode 1501 is close to the muscle 1502, and the electrode 1503 is close to the muscle 1504. The electrode 1505 is away from the muscle 1504.
As described above, when the relative positional relationship between the user's arm and the electrode arrangement surface 100a of the electrical stimulation device 100 is shifted, the arrangement of the electrodes relative to the muscle also changes. This phenomenon appears as a shift of the elements of the finger behavior matrix and the flag matrix created based on this. For this reason, it is necessary to match the elements of the electrode probability matrix 705 previously stored in the nonvolatile storage 504 with the detected finger behavior matrix.

FIG. 16A is an example of a finger behavior matrix created by the second and subsequent learning modes.
FIG. 16B is a virtual electrode probability matrix 705 created based on the finger behavior matrix of FIG. 16A.
FIG. 16C is the electrode probability matrix 705 of FIG. 12D.
FIG. 16D is a matrix in which the electrode probability matrix 705 of FIG. 16C is rearranged.
First, the electrode probability matrix 705 in FIG. 16C is matrix data created in the first learning mode as described in FIG. 12D. On the other hand, the virtual electrode probability matrix 705 shown in FIG. 16B is matrix data created by the second learning mode.

  At first glance, just comparing the matrix data, the relative positional relationship between the user's arm and the electrical stimulation device 100 in the second learning mode (FIG. 16B) is It seems that it is not clear how much the relative positional relationship with the stimulating device 100 (FIG. 16C) is deviated. However, when attention is paid to the element in which the finger moves (the element “1”) in the upper left of the matrix data, the position of the element (P1603) at the time when the finger moves for the first time in the first learning mode, and the second In the second learning mode, it can be clearly seen that the position is shifted from the position of the element (P1601) when the finger moves for the first time. That is, the deviation between the relative positional relationship between the user's arm and the electrical stimulation device 100 at the time of previous use and the current relative positional relationship between the user's arm and the electrical stimulation device 100 is the displacement of the element position. As clearly understood.

As described above, the electrode probability matrix 705 is matrix data representing the correlation between the electrode and the movement of the finger as a probability. The probability applies the Bayesian posterior probability by Bayesian estimation. However, before Bayesian estimation, the difference between the relative positional relationship between the user's arm and the electrical stimulation device 100 at the time of previous use and the current relative positional relationship between the user's arm and the electrical stimulation device 100 is If it appears in the electrode probability matrix 705, the accuracy of learning will be significantly reduced. Therefore, the input / output control unit 712 of the host 401 matches the electrode probability matrix 705 read from the non-volatile storage 504 and held in the RAM 503 with a matrix according to the relative positional relationship between the current user's arm and the electrical stimulation device 100. Swap elements of. In the case of FIG. 16C, the column of the electrode probability matrix 705 in FIG. 16C is set to 1 in the right direction so that the position (1, 1) of the element P1603 is the same position as the position (1, 2) of the element P1601 in FIG. Shift columns. Then, the element group A 1604 in FIG. 16C is arranged at the same position as the element group A 1602 in FIG. 16B, and the element group A 1605 in FIG. 16C is ejected with the movement of the element group A 1604 and left. Arranged in one row. This is the electrode probability matrix 705 of FIG. 16D.
Note that the process of step S1308 is a process of updating the element group A 1605 in FIG. 16D.

  In step S1307 in FIG. 13, for the sake of simplicity, the process of searching for an element indicating an effective electrical stimulation that exists in the upper left of the electrode probability matrix 705 is performed by applying the electrical stimulation voltage sequentially to the electrodes. The electrode probability matrix 705 is constructed on the premise that the movement of the finger with the smallest number occurs. However, since this is convenient, an electrical stimulation voltage is sequentially applied to the electrodes, and elements of the electrode probability matrix 705 corresponding to the first detected finger movement may be referred to.

  In step S1306 in FIG. 13, when the finger that has moved for the first time is not in the bent state of the thumb, any electrode causes a bent state of the thumb in the relative positional relationship between the current user's arm and the electrical stimulation device 100. I'm not letting you. That is, the arrangement of the electrodes fails to flex the thumb. In such a case, the element in the same row of the electrode probability matrix 705 corresponding to the finger that has moved for the first time is referred to. That is, the first line indicating the thumb bending state of the electrode probability matrix 705 is ignored. The electrode probability matrix 705 is rearranged only in column units, and is not rearranged in row units.

Returning to FIG. 13 and FIG. 14 again, the description of the flowchart will be continued.
After step S1309 or when the flag variable MFflag is logically true in step S1306 (NO in S1306), the process proceeds to FIG.
As illustrated in FIG. 14, the input / output control unit 712 operates the probability calculation unit 802 to indicate that the finger has moved with respect to the element in which the finger has moved with respect to the element of the electrode probability matrix 705 corresponding to the j-th electrode. The Bayesian posterior probability “P (x | j)” is calculated. For the element element for which the finger did not move, the Bayes posterior probability “P (x | j)” that the finger did not move is calculated, and the corresponding element of the electrode probability matrix 705 is updated (S1410).

Then, the input / output control unit 712 confirms whether or not the counter variable j has reached the maximum value of j, that is, the total number of electrodes. If the counter variable j has not reached the total number of electrodes (NO in S1411), the input / output control unit 712 increments the counter variable j by 1 (S1412) and repeats the processing from step S1303 again.
If the counter variable j reaches the total number of electrodes in step S1411 (YES in S1411), the input / output control unit 712 stores the created electrode probability matrix 705 in the non-volatile storage 504 and sends it to the electrical stimulation apparatus 100. Transmit (S1413), and the series of processing ends (S1414).

Returning to FIG. 13, if no finger is moving in step S1305 (NO in S1305), then the input / output control unit 712 determines whether or not the flag variable MFflag is logically false, that is, at this time. It is confirmed whether or not the finger has not moved yet (S1415). If the finger has not moved yet (YES in S1415), the counter variable j is checked in step S1411 without doing anything.
In step S1415, if the flag variable MFflag is logically true, that is, if the finger has already moved at this time (NO in S1415), the input / output control unit 712 sets the probability calculation unit 802. The Bayes posterior probability “P (x | j)” indicating that the finger has not moved is calculated for all elements of the electrode probability matrix 705 corresponding to the j-th electrode, and the corresponding of the electrode probability matrix 705 The element to be updated is updated (S1416). In step S1411, the counter variable j is checked.

  The determination in step S1415 is a process for determining whether or not the electrode probability matrix 705 has been rearranged in step S1307. If the electrode probability matrix 705 indicating the relative positional relationship between the user's arm and the electrical stimulation device 100 at the time of previous use is not matched with the relative positional relationship between the current user's arm and the electrical stimulation device 100, the electrode Learning based on Bayesian estimation cannot be correctly performed on the probability matrix 705. Accordingly, the Bayesian estimation calculation is not performed until the process of Step S1307 is completed (YES in S1415), but after the process of Step S1307 is completed, the Bayesian estimation calculation is performed (NO from S1308, S1410, and S1415 to S1416) ).

The electrical stimulation device 100 according to the embodiment of the present invention is configured by eight muscle displacement sensors and eight electrodes, but the number of muscle displacement sensors and electrodes is not necessarily eight. Rather, the greater the number of muscle displacement sensors and electrodes, the more precisely the detection of muscle contraction and the control of muscle contraction can be performed.
FIG. 17 is a diagram illustrating an example of a generalized electrode probability matrix 705. When the user repeatedly uses the electrical stimulation device 100, the electrical stimulation device 100 is repeatedly attached to and detached from the user's arm each time. That is, every time the electrical stimulation apparatus 100 is attached to or detached from the user's arm, the learning mode is executed. When the learning mode is repeatedly executed, as a result, a correlation between a certain electrode and a finger movement is determined. Each element of the electrode probability matrix 705 stores a Bayesian posterior probability “P (x | j)” for x rows and j columns.

The embodiment of the present invention described above can be applied as follows.
(1) It is assumed that one electrical stimulator 100 is introduced into the family. Dad, mom, and child have different arm thicknesses. Therefore, when one electrical stimulation apparatus 100 is shared by a plurality of users, it is preferable to provide a user authentication function in the host 401 and associate the electrode probability matrix 705 with a user ID.
The user authentication may be any means having a function for uniquely identifying a user. For example, in addition to general password authentication using a keyboard, biometric authentication such as fingerprints, veins, and irises can be used. A biometric unit using fingerprints or veins is convenient to use when stored in the circuit storage box 103 of the electrical stimulation device 100.
Further, when this user authentication function is integrated with the user authentication function of the information processing unit 713, which is an application program, the user authentication of the electrical stimulation device 100 and the user authentication of the application program can be unified, so that the usability is further improved. You can expect to improve. In this case, the user ID is associated with the user data of the information processing unit 713 and the electrode probability matrix 705 of the electrical stimulation device 100.
That is, a user authentication function is included in the device driver program of the electrical stimulation apparatus 100 installed in the host 401, and the application program uses the user authentication function of the device driver program. Centralized user authentication can be realized.

  (2) In the flowcharts shown in FIG. 10, FIG. 13 and FIG. 14, after applying electrical stimulation to the electrodes, the data of the muscle displacement sensor is taken in, and after the movement of the finger is examined, the determination and learning processing proceeds sequentially. However, it is also possible to first apply electrical stimulation to the electrodes and execute the operation of taking in the data of the muscle displacement sensor for all the electrodes, create finger behavior matrices in advance, and then perform the determination and learning processing. Of course, step S1307 in FIG. 13 is indispensable also at that time.

  (3) In the electrical stimulation apparatus 100 according to the embodiment of the present invention, Bayesian estimation is adopted as the learning algorithm, but the learning algorithm is not limited to this. For example, other learning algorithms such as a support vector machine may be used.

In the present embodiment, the electrical stimulation device 100 and the electrical stimulation system 400 have been disclosed.
In order to clarify the correlation between the electrical stimulation by the electrode and the movement of the finger, the host 401 creates an electrode probability matrix 705 in which a Bayes posterior probability indicating which finger movement corresponds to the electrode is described as an element. To the electrical stimulator 100. In the second and subsequent learning modes, in order to match the previous electrode probability matrix 705 with the wearing state of the electrical stimulation device 100 in the current user's arm, the element in which the finger movement existing at the upper left of the electrode probability matrix 705 has occurred Are rearranged, and the columns of the electrode probability matrix 705 are rearranged as necessary.
By configuring the electrical stimulation device 100 and the host 401 in this way, the correspondence between the finger movement and the electrode is clarified in a short time regardless of the state worn on the user's arm, and there is very little malfunction and high accuracy. Thus, the electrical stimulation device 100 and the electrical stimulation system 400 that can drive the target finger can be realized.

The embodiments of the present invention have been described above. However, the present invention is not limited to the above-described embodiments, and other modifications and application examples are provided without departing from the gist of the present invention described in the claims. including.
For example, the above-described embodiment is a detailed and specific description of the configuration of the apparatus and the system in order to explain the present invention in an easy-to-understand manner, and is not necessarily limited to one having all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Moreover, it is also possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

Each of the above-described configurations, functions, processing units, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. Further, each of the above-described configurations, functions, and the like may be realized by software for interpreting and executing a program that realizes each function by the processor. Information such as programs, tables, and files that realize each function must be held in a volatile or non-volatile storage such as a memory, hard disk, or SSD (Solid State Drive), or a recording medium such as an IC card or an optical disk. Can do.
In addition, the control lines and information lines are those that are considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. Actually, it may be considered that almost all the components are connected to each other.

DESCRIPTION OF SYMBOLS 100 ... Electrical stimulation apparatus, 101 ... Band, 102 ... Left side part, 103 ... Circuit storage box, 104 ... First serial interface terminal, 150 ... Arithmetic processing part, 201, 202, 203, 204, 205, 206, 207, 208 211, 212, 213, 216 ... electrodes, 217 ... electrodes, 221, 223, 227 ... muscle displacement sensors, 231 ... right electrode placement locations, 232 ... central electrode placement locations, 233 ... left electrode placement locations, 400 ... electrical stimulation System 401 ... Host 402 ... Near Field Communication Unit 501 ... CPU 502 ... ROM 503 ... RAM 504 ... Non-volatile Storage 505 ... Display Unit 506 ... Operation Unit 507 ... Bus 601 ... Bus 602 ... CPU, 603 ... ROM, 604 ... RAM, 605 ... A / D converter, 606 ... second serial Interface, 607 ... one-chip microcomputer, 608 ... first multiplexer, 610 ... second multiplexer, 612 ... 6-axis sensor, 613 ... short-range wireless communication unit, 614 ... first serial interface, 615 ... step-up circuit, 616 ... PWM Switch 617... Third multiplexer 701 Input / output control unit 702 Short range wireless transmission unit 703 Finger electrode corresponding conversion unit 704 Short range wireless reception unit 705 Electrode probability matrix 711 Short range wireless Reception unit, 712 ... Input / output control unit, 713 ... Information processing unit, 714 ... Short-range wireless transmission unit, 801 ... Sensor value storage unit, 802 ... Probability calculation unit, 803 ... Timer, 1201 ... Maximum value array, 1202 ... Flag Array

Claims (6)

A band wrapped around the user's forearm,
A plurality of muscle displacement sensors arranged on one side of the band to detect displacement of a plurality of muscles existing in the user's arm;
A sensor multiplexer for selecting one muscle displacement sensor from the plurality of muscle displacement sensors;
A plurality of electrodes disposed adjacent to the plurality of muscle displacement sensors on the surface of the band on which the plurality of muscle displacement sensors are disposed;
An electrode multiplexer for selecting one electrode from the plurality of electrodes;
A short-range wireless transmission unit that converts a signal relating to the displacement of the muscle of the user's arm obtained from the muscle displacement sensor into digital data and transmits the digital data to a predetermined host;
An electrode probability matrix in which the correlation between the movement of the finger and the plurality of electrodes is described by probability;
A short-range wireless receiving unit that receives a command to move a finger from the host in a normal mode and receives the updated electrode probability matrix from the host in a calibration mode;
A finger electrode corresponding conversion unit that specifies an electrode showing the maximum probability from the electrode probability matrix based on an instruction to move the finger, and controls the electrode multiplexer to select the specified electrode;
Comprising
In the electrode probability matrix, the host converts the displacement data of the muscle displacement sensor into the movement of the user's finger, and the relative positional relationship between the current arm of the user and the plurality of electrodes is calculated. After detecting and reordering the elements of the electrode probability matrix, the elements are updated.
Electrical stimulator. The electrical stimulation device according to claim 1, wherein elements of the electrode probability matrix are updated by calculating a posteriori probability based on a learning algorithm. A band wrapped around the user's forearm,
A plurality of muscle displacement sensors arranged on one side of the band to detect displacement of a plurality of muscles existing in the user's arm;
A sensor multiplexer for selecting one muscle displacement sensor from the plurality of muscle displacement sensors;
A plurality of electrodes disposed adjacent to the plurality of muscle displacement sensors on the surface of the band on which the plurality of muscle displacement sensors are disposed;
An electrode multiplexer for selecting one electrode from the plurality of electrodes;
An electrode probability matrix in which the correlation between the movement of the finger and the plurality of electrodes is described by probability;
A finger electrode corresponding conversion unit that specifies an electrode showing the maximum probability from the electrode probability matrix based on an instruction to move the finger, and controls the electrode multiplexer to select the specified electrode;
A method for calibrating the electrode probability matrix of an electrical stimulator comprising:
Applying electrical stimulation to the electrode;
Examining movements of all fingers of the forearm to which the user's electrical stimulation device is worn through the muscle displacement sensor;
When it is determined that any finger of the forearm to which the electrical stimulation device of the user is attached has been moved by electrical stimulation by any of the plurality of electrodes, and when it is determined that the finger has moved for the first time at this point, Elements of the electrode probability matrix before the calibration of the electrode probability matrix are performed, so that elements of the electrode probability matrix at the time when the finger is moved for the first time in the calibration operation match the current elements. Sorting by row and / or column
Calibrating the electrode probability matrix of the electrical stimulator. Reordering the elements of the electrode probability matrix by row and / or column and then updating the elements of the electrode probability matrix by calculating a posteriori probability based on a learning algorithm;
The method of calibrating an electrode probability matrix of an electrical stimulator according to claim 3. A band wrapped around the user's upper arm,
A plurality of muscle displacement sensors arranged on one side of the band to detect displacement of a plurality of muscles existing in the user's arm;
A sensor multiplexer for selecting one muscle displacement sensor from the plurality of muscle displacement sensors;
A sensor device comprising:
A host that communicates with the sensor device;
A muscle displacement sensor calibration method for a sensor system comprising:
A stationary state guide including a message for prompting a state in which the user puts the sensor device on the upper arm and the user is not putting power on the finger on the display unit of the host that has established communication with the sensor device A first step of displaying a video and the host sending a stationary muscle displacement sensor data collection command for collecting data of the muscle displacement sensor to the sensor device;
A second step in which the sensor device returns data of all the muscle displacement sensors to the host in response to the sensor device receiving the stationary muscle displacement sensor data collection command;
A third step of storing all the muscle displacement sensor data received by the host from the sensor device as stationary state data;
In the state where the user wears the sensor device on the upper arm, the display unit displays a first motion guide video including a message prompting the user to bend the first finger with force. A fourth step in which the host sends a first muscle displacement sensor data collection command to collect data of the muscle displacement sensor to the sensor device;
In response to the sensor device receiving the first muscle displacement sensor data collection command, a fifth step in which the sensor device returns data of all the muscle displacement sensors to the host;
A sixth step of storing all the muscle displacement sensor data received by the host from the sensor device as first operation state data;
In the state where the user wears the sensor device on the upper arm, the display unit displays a second motion guidance video including a message prompting the user to bend the second finger with force. A seventh step in which the host sends a second muscle displacement sensor data collection command to collect data of the muscle displacement sensor to the sensor device;
In response to the sensor device receiving the command, an eighth step in which the sensor device returns data of all the muscle displacement sensors to the host;
A ninth step of storing data of all the muscle displacement sensors received by the host from the sensor device as second operation state data;
A method for calibrating a muscle displacement sensor of a sensor system. Furthermore,
  After the ninth step, the host subtracts the stationary state data from the first operational state data to obtain a first reference value relating to a state in which the first finger is bent, and a second motion A tenth step of calculating a relative value of each of the muscle displacement sensors by subtracting the stationary state data from the state data to obtain a second reference value relating to the state of bending the second finger;
The muscle displacement sensor calibration method for a sensor system according to claim 5.

Images