Rehabilitation Sensor Model Predicts Walking Ability

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current rehabilitation systems face challenges in accurately assessing the movement ability of patients, particularly due to the requirement for extensive space and the need for constant assistance from trained staff, leading to infrequent and potentially biased scoring of the walking FIM index.

Innovation Solution

A rehabilitation support system comprising an actuator for assisting movement, a sensor for data detection, and a learned model that predicts the walking ability of a patient based on input data, allowing for frequent and unbiased assessment of movement ability without the need for extensive space or constant staff assistance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the walking FIM is measured by assessing the amount of assistance when the trainee walks 50 m on flat ground, then the movement ability of the trainee can be assessed, but it is impossible to frequently measure the walking FIM due to space limitations and the need for constant staff assistance

Engineering Contradiction:
Improvemovement ability assessmentVSAvoidmeasurement frequency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces the manual assessment method (mechanical observation and scoring by staff) with an automated sensor-based system. Sensors detect movement data objectively, and a learned model automatically calculates the walking FIM score, eliminating the need for staff to manually assess and score patient movement ability during walking trials.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-assessment of movement ability by using the trainee's own movement data captured by sensors. The learned model processes the sensor data to generate the walking FIM score without requiring external assessment by staff members, allowing frequent automated measurements during rehabilitation sessions.

Inventive Principle:
Principle #25Self-service

2Measurement precision

If a trainee must move 50 m or more flat ground to measure the walking FIM, then the walking FIM can be assessed, but it may be difficult for the trainee to move to a place for measuring the FIM in a hospital where space is limited

Engineering Contradiction:
Improvewalking FIM assessmentVSAvoidmeasurement space
Core Design Contradiction:
Measurement precisionVSArea of stationary object

Solution Approach 1:

The patent changes the measurement parameters by using sensor-collected movement data (such as step count, distance, speed, or other kinematic parameters) to calculate the walking FIM score. This allows the assessment to be performed over shorter distances or in confined spaces while maintaining measurement validity through the learned model that correlates sensor parameters with the standard walking FIM scale.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If the amount of assistance by an assistant during walking on flat ground is assessed, then the walking FIM can be measured, but a deviation may occur in scoring of the walking FIM depending on who the assistant is

Engineering Contradiction:
Improvewalking FIM scoringVSAvoidscoring consistency
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent replaces human assistants with an automated sensor-based assessment system. Sensors objectively detect movement parameters, and a learned model automatically determines the walking FIM score based on these parameters. This eliminates inter-rater variability and ensures consistent scoring regardless of which staff member would have performed the assessment.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3756543B1Rehabilitation support system, prediction device, learning device, methods, and storage medium
Publication Date: 2024.05.01 TOYOTA JIDOSHA KK
  • EP3756543B1 patent drawingFigure 1
  • EP3756543B1 patent drawingFigure 2
  • EP3756543B1 patent drawingFigure 3

AI summary

A learning device includes a data acquisition unit (520) configured to acquire rehabilitation data from a rehabilitation support device having an actuator configured to assist a rehabilitation movement of a trainee and a sensor configured to detect data on the rehabilitation movement assisted by the actuator, a data generation unit (510a) configured to generate, as data for learning, the rehabilitation data that includes detected data corresponding to a detection result by the sensor, and a learning unit (510b) configured to generate, by performing machine learning using the data for learning, a learning model that outputs an index indicating a movement ability of the trainee, upon receiving an input of the detected data.