# US 2021/0343412 A1 · Intervention content estimation device, 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 2021/0343412 A1
- Document ID: US20210343412A1
- Status: Published application
- Kind: A1
- Application number: 17/271,295
- Filed: 2019-07-16
- Published: 2021-11-04
- Earliest claimed: 2018-08-31
- Assignees: Nippon Telegraph and Telephone Corporation
- Applicants: Nippon Telegraph and Telephone Corporation
- Inventors: Hisashi KURASAWA; Shozo AZUMA; Naoki ASANOMA; Akihiro CHIBA; Kana EGUCHI; Tsutomu YABUUCHI; Kazuhiro YOSHIDA; Tomohiro YAMADA
- Web page: https://patentgrep.com/patent/US20210343412A1

## CPC

- G16H 20/70: relating to mental therapies, e.g. psychological therapy or autogenous training
- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- G06N 3/0499: Feedforward networks
- G06N 3/08: Learning methods
- G06N 3/092: Reinforcement learning
- G06N 20/00: Machine learning
- G16H 10/60: for patient-specific data, e.g. for electronic patient records
- G16H 20/60: relating to nutrition control, e.g. diets
- G06N 3/048: Activation functions
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

An object of an aspect of the present invention is to enable estimation of more effective intervention content in order to make a person's health state approximate to an ideal health state, and in a learning phase, measurement values and target values of a health state for a plurality of days in the past are sequentially input to a learning machine configured by a multilayer neural network, and the learning machine is caused to perform learning such that a target achievement expectation value obtained by using a success rate that allows the user's health state to approximate to an ideal health state, continuity that allows the health state approximate to the ideal health state to be maintained, and a target value of the health state to be subsequently recommended and the target achievement expectation value thereof that reflect a temporal change in the health state and a history of interventions until a present time are output. In an estimation phase, measurement values and target values of the user's health state in the last three days are input to the estimation model after learning, and the target value of the health state to be recommended, which is output from the estimation model at this time, is presented to the user.

## Claims

### Claim 1

An intervention content estimation apparatus comprising:
a processor; and
a storage medium having computer program instructions stored thereon, when executed by the processor, perform to:
acquire, on a per user basis, record information that includes a target value of a health state, which is determined based on a current health state and a future ideal health state set in advance, and a measurement value of the health state of the user after the target value of the health state is presented; and
generate an intervention content estimation model by inputting the record information acquired by the first acquisition portion as training data to a learning machine and causing the learning machine to perform learning such that information representing a target value of the health state to be subsequently recommended is output as an evaluation result from the learning machine.

### Claim 2 (depends on 1)

The intervention content estimation apparatus according to claim 1, wherein the computer program instructions further causes the learning machine to perform learning such that information reflecting a target achievement expectation value is output as the evaluation result, the target achievement expectation value being obtained by using a success rate that allows the current health state to approximate to the ideal health state.

### Claim 3 (depends on 1)

The intervention content estimation apparatus according to claim 1, wherein the computer program instructions further
causes the learning machine to perform learning such that information reflecting a target achievement expectation value is output as the evaluation result, the target achievement expectation value being obtained by using a success rate that allows the current health state to approximate to the ideal health state, and a discount rate provided to the success rate as a coefficient.

### Claim 4 (depends on 2)

The intervention content estimation apparatus according to claim 2, wherein the computer program instructions further perform to input the acquired record information of a plurality of days set in advance as training data to the learning machine and thereby causes the learning machine to perform learning such that information reflecting the target achievement expectation value, and a temporal change in the measurement value of the health state and a history of a change in the target value of the health state of the user until a present time is output as the evaluation result.

### Claim 5 (depends on 1)

The intervention content estimation apparatus according to claim 1 wherein the computer program instructions further perform to acquire, on a per user basis, most recent record information including the target value of the health state presented and the measurement value of the health state of the user after the target value of the health state is presented; and
input the most recent record information as evaluation data to the intervention content estimation model and output, as estimation data, information representing a target value of the health state to be subsequently recommended next, which is output from the intervention content estimation model in response to the input.

### Claim 6

An intervention content estimation method that is executed by an information processing apparatus having a processor and a memory, the intervention content estimation method comprising:
acquiring, on a per user basis, record information that includes a target value of a health state determined based on a current health state and a future ideal health state set in advance, and a measurement value of the health state of the user after the target value of the health state is presented; and
generating, as a learning process, an intervention content estimation model by inputting the acquired record information as training data to a learning machine and causing the learning machine to perform learning such that information representing a target value of the health state to be subsequently recommended is output as an evaluation result from the learning machine.

### Claim 7 (depends on 6)

The intervention content estimation method according to claim 6, further comprising:
acquiring, on a per user basis, most recent record information including the target value of the health state presented and the measurement value of the health state of the user after the target value of the health state is presented; and
inputting, as an estimation process, the acquired most recent record information as evaluation data to the intervention content estimation model and outputting, as estimation data, information representing a target value of the health state to be subsequently recommended, which is output from the intervention content estimation model in response to the input.

### Claim 8 (depends on 1)

A non-transitory computer-readable medium having computer-executable instructions that, upon execution of the instructions by a processor of a computer, cause the computer to function as the intervention content estimation apparatus according to claim 1.
