Granted Patent B2

US 12,642,452 · App. 18/156,565 · Granted Jun 2, 2026

Inventors: Tatsuhiko Matsumoto (Nagaokakyo, JP); Atsushi Naito (Nagaokakyo, JP); Naoki Kawara (Nagaokakyo, JP); Yutaka Takamaru (Nagaokakyo, JP)

Assignee: MURATA MANUFACTURING CO., LTD.

A61B5/1107A61B5/1101A61B5/7264A61B5/7267

Abstract

Claims (26)

1 . An action state estimation apparatus comprising:

a displacement detection sensor configured to detect a displacement measurement signal by converting displacement on a skin surface of a test subject due to tremor and deformation into a voltage that corresponds to the displacement measurement signal;

a storage medium; and

a processing apparatus that includes a first sampling portion configured to sample the displacement measurement signal of the test subject that is detected by the displacement detection sensor and is within a predetermined time and to generate displacement measurement data based on the sampled displacement measurement signal;

an action state model storage configured to store in the storage medium an action state model modeled by associating the displacement measurement data with a loaded state of a muscle of the test subject; and

an estimation calculation portion configured to estimate the loaded state by setting the displacement measurement data as an input vector and using the action state model.

2 . The action state estimation apparatus according to claim 1 , wherein the action state model sets an importance level according to a time range of the displacement measurement data for each muscle of the test subject of which the loaded state is to be estimated.

3 . The action state estimation apparatus according to claim 1 , further comprising a second sampling portion configured to sample a motion measurement signal of the test subject within a predetermined time and to generate motion measurement data based on the sampled motion measurement signal.

4 . The action state estimation apparatus according to claim 3 , wherein:

the action state model storage stores the action state model modeled by associating the displacement measurement data, the motion measurement data, and the loaded state of the muscle of the test subject; and

the estimation calculation portion is configured to estimate the loaded state of the muscle by setting the displacement measurement data and the motion measurement data as the input vector and using the action state model.

5 . The action state estimation apparatus according to claim 4 , wherein the action state model sets the importance level according to the time range of the displacement measurement data and an importance level based on a time range of the motion measurement data for each muscle of the test subject of which the loaded state is to be estimated.

6 . The action state estimation apparatus according to claim 5 , wherein the importance level according to the time range of the displacement measurement data and the importance level according to the time range of the motion measurement data are set by a common level of importance.

7 . The action state estimation apparatus according to claim 5 , wherein the importance level according to the time range of the displacement measurement data and the importance level according to the time range of the motion measurement data are set individually.

8 . The action state estimation apparatus according to claim 4 , wherein the motion measurement signal is a measurement signal of at least one of acceleration and angular velocity.

9 . The action state estimation apparatus according to claim 8 , wherein the motion measurement signal includes at least one of an orthogonal triaxial component and a composite component provided by combining the orthogonal triaxial component.

10 . The action state estimation apparatus according to claim 1 , wherein the input vector includes biological information.

11 . The action state estimation apparatus according to claim 1 , wherein the estimation calculation portion is configured to set a plurality of statistics as the input vector, with a level of importance of the plurality of statistics set as the action state model and based on a type of the muscle of the test subject to be estimated.

12 . An action state estimation method comprising:

detecting, by a displacement detection sensor, a displacement measurement signal by converting displacement on a skin surface of a test subject due to tremor and deformation into a voltage that corresponds to the displacement measurement signal;

sampling the displacement measurement signal of the test subject within a predetermined time and generating displacement measurement data based on the sampled displacement measurement signal; and

estimating, by a processing apparatus that executes instructions stored on a storage medium, a loaded state of a muscle of the test subject by using an action state model modeled by associating the displacement measurement data with the loaded state of the muscle and setting the displacement measurement data as an input vector.

13 . The action state estimation method according to claim 12 , further comprising sampling in time series a motion measurement signal of the test subject within a predetermined time and generating motion measurement databased on the sampled motion measurement signal.

14 . The action state estimation method according to claim 13 , further comprising:

estimating the loaded state by using the action state model modeled by associating the displacement measurement data, the motion measurement data, and the loaded state of the muscle of the test subject; and

setting the displacement measurement data and the motion measurement data as the input vector.

Assignments (1)

Priority Claims (2)

Continuity (2)

References Cited (18)