# US 2015/0106023 A1 · Training apparatus, training method, and program

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2015/0106023 A1
- Document ID: US20150106023A1
- Status: Published application
- Kind: A1
- Application number: 14/511,542
- Filed: 2014-10-10
- Published: 2015-04-16
- Earliest claimed: 2013-10-11
- Assignees: Toyota Jidosha Kabushiki Kaisha
- Applicants: Toyota Jidosha Kabushiki Kaisha
- Inventors: Shingo SHIMODA; Fady SK SHIBATA ALNAJJAR; Tytus WOJTARA; Hitoshi YAMADA
- Web page: https://patentgrep.com/patent/US20150106023A1

## CPC

- A61B 5/4519: {Muscles (measuring muscular strength A61B5/224; using electromyography [EMG] A61B5/389)}
- A61B 5/389: Electromyography [EMG]
- A61B 5/7246: {using correlation, e.g. template matching or determination of similarity}
- G16H 10/60: for patient-specific data, e.g. for electronic patient records
- G16H 20/00: ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H 20/30: relating to physical therapies or activities, e.g. physiotherapy, acupressure or exercising
- A61B 5/7275: {Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor}
- A61B 2505/09: Rehabilitation or training
- A61H 1/02: Stretching or bending {or torsioning} apparatus for exercising
- A61H 2230/60: Muscle strain, i.e. measured on the user, e.g. Electromyography [EMG]
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- A61H: PHYSICAL THERAPY APPARATUS, e.g. DEVICES FOR LOCATING OR STIMULATING REFLEX POINTS IN THE BODY; ARTIFICIAL RESPIRATION; MASSAGE; BATHING DEVICES FOR SPECIAL THERAPEUTIC OR HYGIENIC PURPOSES OR SPECIFIC PARTS OF THE BODY (electrotherapy, magnetotherapy, radiation therapy or ultrasound therapy A61N)

## Abstract

A training apparatus includes: a controller. The controller is configured to: (a) while a subject is performing a behavior, acquire time-series myoelectric potentials of each of muscles of the subject; (b) calculate a muscle synergy matrix W, a control matrix C and an error matrix E from a myoelectric potential matrix M so that the relationship M=WC+E is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of a corresponding one of the muscles, m being arranged in rows in the myoelectric potential matrix, n unit column vectors being arranged in columns in the muscle synergy matrix W, n row vectors being arranged in rows in the control matrix; (c) calculate a feature quantity indicating consistency among the unit column vectors included in the calculated muscle synergy matrix W; and (d) output the calculated feature quantity.

## Claims

### Claim 1

A training apparatus comprising:
a controller configured to:
(a) while a subject is performing a behavior, acquire time-series myoelectric potentials of each of muscles 1, 2, . . . , m of the subject;
(b) calculate a muscle synergy matrix W, a control matrix C and an error matrix E from a myoelectric potential matrix M so that the relationship M=WC+E is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of one of the muscles 1, 2, . . . , m being arranged in rows in the myoelectric potential matrix M, n unit column vectors W(1), W(2), . . . , W(n) being arranged in columns in the muscle synergy matrix W, n row vectors C(1), C(2), . . . , C(n) being arranged in rows in the control matrix C;
(c) calculate a feature quantity indicating consistency among the unit column vectors W(1), W(2), . . . , W(n) included in the calculated muscle synergy matrix W; and
(d) output the calculated feature quantity.

### Claim 2 (depends on 1)

The training apparatus according to claim 1, wherein
the controller is configured to apply non-negative matrix factorization, and
the feature quantity is an SCI that is calculated through computation of inner product p(x,y) of row vectors x, y as SCI=2/[n(n+2)]×Σi=1nΣj=1n,j≠ip(W(i),W(j)).

### Claim 3 (depends on 2)

The training apparatus according to claim 2, wherein
the controller is configured to output, to the subject, the fact that the SCI is larger than or equal to a predetermined threshold.

### Claim 4 (depends on 1)

The training apparatus according to claim 1, wherein
the controller is configured to calculate a right feature quantity for a muscle of a right side of a body of the subject and a left feature quantity for a muscle of a left side of the body of the subject, and
the controller is configured to output, to the subject, one of information about both the right feature quantity and the left feature quantity and information about a similarity between the right feature quantity and the left feature quantity.

### Claim 5

A training apparatus comprising:
a controller configured to:
(e) while a subject is performing a task consisting of behaviors k=1, 2, . . . , p, acquire time-series myoelectric potentials of each of muscles 1, 2, . . . , m of the subject;
(f) calculate a muscle synergy matrix Wk, a control matrix Ck and an error matrix Ek from a myoelectric potential matrix Mk for each of the behaviors k=1, 2, . . . , p so that the relationship Mk=WkCk+Ek is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of a corresponding one of the muscles 1, 2, . . . , m being arranged in rows in the myoelectric potential matrix Mk, n unit column vectors Wk(1), Wk(2), . . . , Wk(n) being arranged in columns in the muscle synergy matrix Wk, n row vectors Ck(1), Ck(2), . . . , Ck(n) being arranged in rows in the control matrix Ck;
(g) calculate a feature quantity indicating consistency among the unit column vectors W1(1), W2(1), . . . , Wp(1), consistency among the unit column vectors W1(2), W2(2), . . . , Wp(2), . . . , consistency among the unit column vectors W1(n), W2(n), . . . , Wp(n), these unit column vectors W1(1), W2(1), . . . , Wp(1), unit column vectors W1(2), W2(2), . . . , Wp(2), . . . , and unit column vectors W1(n), W2(n), . . . , Wp(n) being included in the calculated muscle synergy matrices W1, W2, . . . , Wp; and
(h) output the calculated feature quantity.

### Claim 6 (depends on 5)

The training apparatus according to claim 5, wherein
the controller is configured to apply non-negative matrix factorization, and
the feature quantity is an SSI that is calculated through computation of correlation coefficient r(x,y) of vectors x, y as SSI=2/[n×p(p−1)]×Σi=1nΣk=1pΣh=1p,h≠kr(Wk(i),Wh(i)).

### Claim 7 (depends on 6)

The training apparatus according to claim 6, wherein
the controller is configured to output, to the subject, one of the fact that the SSI is larger than or equal to a predetermined threshold and the fact that the SSI tends to increase with an increase in the p indicating the number of the behaviors.

### Claim 8 (depends on 5)

The training apparatus according to claim 5, wherein
the controller is configured to calculate a right feature quantity for a muscle of a right side of a body of the subject and a left feature quantity for a muscle of a left side of the body of the subject, and
the controller is configured to output, to the subject, one of information about both the right feature quantity and the left feature quantity and information about a similarity between the right feature quantity and the left feature quantity.

### Claim 9

A training apparatus comprising:
a controller configured to:
(i) while a subject is performing a task consisting of behaviors k=1, 2, . . . , p, acquire time-series myoelectric potentials of each of muscles 1, 2, . . . , m of the subject;
(j) calculate a muscle synergy matrix Wk, a control matrix Ck and an error matrix Ek from a myoelectric potential matrix Mk for each of the behaviors k=1, 2, . . . , p so that the relationship Mk=WkCk+Ek is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of a corresponding one of the muscles 1, 2, . . . , m being arranged in rows in the myoelectric potential matrix Mk, n unit column vectors Wk(1), Wk(2), . . . , Wk(n) being arranged in columns in the muscle synergy matrix Wk, n row vectors Ck(1), Ck(2), . . . , Ck(n) being arranged in rows in the control matrix Ck;
(k) calculate a feature quantity indicating consistency among the row vectors C1(1), C2(1), . . . , Cp(1), consistency among the row vectors C1(2), C2(2), . . . , Cp(2), . . . , consistency among the row vectors C1(n), C2(n), . . . , Cp(n), these row vectors C1(1), C2(1), . . . , Cp(1), row vectors C1(2), C2(2), . . . , Cp(2), . . . , and row vectors C1(n), C2(n), . . . , Cp(n) being included in the calculated control matrices C1, C2, . . . , Cp; and
(l) output the calculated feature quantity.

### Claim 10 (depends on 9)

The training apparatus according to claim 9, wherein
the controller is configured to apply non-negative matrix factorization, and
the feature quantity is an SSIC that is calculated through computation of correlation coefficient r(x,y) of vectors x, y as SSIC=2/[n×p(p−1)]×Σi=1nΣk=1pΣh=1p,h≠kr(Ck(i),Ch(i)).

### Claim 11 (depends on 10)

The training apparatus according to claim 10, wherein
the controller is configured to output, to the subject, one of the fact that the SSIC is smaller than or equal to a predetermined threshold and the fact that the SSIC tends to decrease with an increase in the p indicating the number of the behaviors.

### Claim 12 (depends on 9)

The training apparatus according to claim 9, wherein
the controller is configured to calculate a right feature quantity for a muscle of a right side of a body of the subject and a left feature quantity for a muscle of a left side of the body of the subject, and
the controller is configured to output, to the subject, one of information about both the right feature quantity and the left feature quantity and information about a similarity between the right feature quantity and the left feature quantity.

### Claim 13

A training method for an apparatus including a controller, the training method comprising:
while a subject is performing a behavior, acquiring, by the controller, time-series myoelectric potentials of each of muscles 1, 2, . . . , m of the subject;
calculating, by the controller, a muscle synergy matrix W, a control matrix C and an error matrix E from a myoelectric potential matrix M so that the relationship M=WC+E is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of a corresponding one of the muscles 1, 2, . . . , m being arranged in rows in the myoelectric potential matrix M, n unit column vectors W(1), W(2), . . . , W(n) being arranged in columns in the muscle synergy matrix W, n row vectors C(1), C(2), . . . , C(n) being arranged in rows in the control matrix C;
calculating, by the controller, a feature quantity indicating consistency among the unit column vectors W(1), W(2), . . . , W(n) included in the calculated muscle synergy matrix W; and
outputting, by the controller, the calculated feature quantity.

### Claim 14

A training method for an apparatus including a controller, the training method comprising:
while a subject is performing a task consisting of behaviors k=1, 2, . . . , p, acquiring, by the controller, time-series myoelectric potentials of each of muscles 1, 2, . . . , m of the subject;
calculating, by the controller, a muscle synergy matrix Wk, a control matrix Ck and an error matrix Ek from a myoelectric potential matrix Mk for each of the behaviors k=1, 2, . . . , p so that the relationship Mk=WkCk+Ek is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of a corresponding one of the muscles 1, 2, . . . , m being arranged in rows in the myoelectric potential matrix Mk, n unit column vectors Wk(1), Wk(2), . . . , Wk(n) being arranged in columns in the muscle synergy matrix Wk, n row vectors Ck(1), Ck(2), . . . , Ck(n) being arranged in rows in the control matrix Ck;
calculating, by the controller, a feature quantity indicating consistency among the unit column vectors W1(1), W2(1), . . . , Wp(1), consistency among the unit column vectors W1(2), W2(2), . . . , Wp(2), . . . , consistency among the unit column vectors W1(n), W2(n), . . . , Wp(n), these unit column vectors W1(1), W2(1), . . . , Wp(1), unit column vectors W1(2), W2(2), . . . , Wp(2), . . . , and unit column vectors W1(n), W2(n), . . . , Wp(n) being included in the calculated muscle synergy matrices W1, W2, . . . , Wp; and
outputting, by the controller, the calculated feature quantity.

### Claim 15

A training method for an apparatus including a controller, the training method comprising:
while a subject is performing a task consisting of behaviors k=1, 2, . . . , p, acquiring, by the controller, time-series myoelectric potentials of each of muscles 1, 2, . . . , m of the subject;
calculating, by the controller, a muscle synergy matrix Wk, a control matrix Ck and an error matrix Ek from a myoelectric potential matrix Mk for each of the behaviors k=1, 2, . . . , p so that the relationship Mk=WkCk+Ek is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of a corresponding one of the muscles 1, 2, . . . , m being arranged in rows in the myoelectric potential matrix Mk, n unit column vectors Wk(1), Wk(2), . . . , Wk(n) being arranged in columns in the muscle synergy matrix Wk, n row vectors Ck(1), Ck(2), . . . , Ck(n) being arranged in rows in the control matrix Ck;
calculating, by the controller, a feature quantity indicating consistency among the row vectors C1(1), C2(1), . . . , Cp(1), consistency among the row vectors C1(2), C2(2), . . . , Cp(2), . . . , consistency among the row vectors C1(n), C2(n), . . . , Cp(n), these row vectors C1(1), C2(1), . . . , Cp(1), row vectors C1(2), C2(2), . . . , Cp(2), . . . , and row vectors C1(n), C2(n), . . . , Cp(n) being included in the calculated control matrices C1, C2, . . . , Cp; and
outputting, by the controller, the calculated feature quantity.

### Claim 16

A non-transitory computer-readable information storage medium storing a program for causing a computer to execute a process, the process comprising:
(a) while a subject is performing a behavior, acquiring time-series myoelectric potentials of each of muscles 1, 2, . . . , m of the subject;
(b) calculating a muscle synergy matrix W, a control matrix C and an error matrix E from a myoelectric potential matrix M so that the relationship M=WC+E is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of a corresponding one of the muscles 1, 2, . . . , m being arranged in rows in the myoelectric potential matrix M, n unit column vectors W(1), W(2), . . . , W(n) being arranged in columns in the muscle synergy matrix W, n row vectors C(1), C(2), . . . , C(n) being arranged in rows in the control matrix C;
(c) calculating a feature quantity indicating consistency among the unit column vectors W(1), W(2), . . . , W(n) included in the calculated muscle synergy matrix W; and
(d) outputting the calculated feature quantity.

### Claim 17

A non-transitory computer-readable information storage medium storing a program for causing a computer to execute a process, the process comprising:
(e) while a subject is performing a task consisting of behaviors k=1, 2, . . . , p, acquiring time-series myoelectric potentials of each of muscles 1, 2, . . . , m of the subject;
(f) calculating a muscle synergy matrix Wk, a control matrix Ck and an error matrix Ek from a myoelectric potential matrix Mk for each of the behaviors k=1, 2, . . . , p so that the relationship Mk=WkCk+Ek is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of a corresponding one of the muscles 1, 2, . . . , m being arranged in rows in the myoelectric potential matrix Mk, n unit column vectors Wk(1), Wk(2), . . . , Wk(n) being arranged in columns in the muscle synergy matrix Wk, n row vectors Ck(1), Ck(2), . . . , Ck(n) being arranged in rows in the control matrix Ck;
(g) calculating a feature quantity indicating consistency among the unit column vectors W1(1), W2(1), . . . , Wp(1), consistency among the unit column vectors W1(2), W2(2), . . . , Wp(2), . . . , consistency among the unit column vectors W1(n), W2(n), . . . , Wp(n), these unit column vectors W1(1), W2(1), . . . , Wp(1), unit column vectors W1(2), W2(2), . . . , Wp(2), . . . , and unit column vectors W1(n), W2(n), . . . , Wp(n) being included in the calculated muscle synergy matrices W1, W2, . . . , Wp; and
(h) outputting the calculated feature quantity.

### Claim 18

A non-transitory computer-readable information storage medium storing a program for causing a computer to execute a process, the process comprising:
(i) while a subject is performing a task consisting of behaviors k=1, 2, . . . , p, acquiring time-series myoelectric potentials of each of muscles 1, 2, . . . , m of the subject;
(j) calculating a muscle synergy matrix Wk, a control matrix Ck and an error matrix Ek from a myoelectric potential matrix Mk for each of the behaviors k=1, 2, . . . , p so that the relationship Mk=WkCk+Ek is satisfied, row vectors each expressing the acquired time-series myoelectric potentials of a corresponding one of the muscles 1, 2, . . . , m being arranged in rows in the myoelectric potential matrix Mk, n unit column vectors Wk(1), Wk(2), . . . , Wk(n) being arranged in columns in the muscle synergy matrix Wk, n row vectors Ck(1), Ck(2), . . . , Ck(n) being arranged in rows in the control matrix Ck;
(k) calculating a feature quantity indicating consistency among the row vectors C1(1), C2(1), . . . , Cp(1), consistency among the row vectors C1(2), C2(2), . . . , Cp(2), . . . , consistency among the row vectors C1(n), C2(n), . . . , Cp(n), these row vectors C1(1), C2(1), . . . , Cp(1), row vectors C1(2), C2(2), . . . , Cp(2), . . . , and row vectors C1(n), C2(n), . . . , Cp(n) being included in the calculated control matrices C1, C2, . . . , Cp; and
(l) outputting the calculated feature quantity.
