Developing a Balance Rehabilitation System for Older Adults, Based on IMU and AI: Personalized Training and Preventive Strategies

Developing a Balance Rehabilitation System for Older Adults, Based on Inertial Measurement Unit Sensing and Artificial Intelligence: Personalized Training and Preventive Strategies

The aging physiological state of the elderly may lead to problems such as unstable gait, balance disorders, and falls. Previous research has confirmed that exercise training can help improve the physical function, quality of life, and reduce the risk of falls in the elderly. In order to achieve effective and continuous intervention training, somatosensory games have become a trend in recent years. Among them, the use of non-immersive virtual reality training methods not only provides training for the elderly, but also reduces the discomfort caused by the virtual environment; however, there are some limitations in clinical rehabilitation training methods, such as the lack of data-based evaluation and personalization. In order to solve the above problems, this research plan will use the inertial measurement unit as a tool for clinical monitoring and human movement assessment, and use artificial intelligence technology to evaluate and adjust the training plan according to its gait characteristics to achieve personalization Training and prevention strategies.

Study Overview

• Status: Recruiting
• Conditions: Community-dwelling Older Adults
• Intervention / Treatment: Other: IMU-based balance training; Other: general health education or exercise training

Detailed Description

The development of a balance rehabilitation system for older adults, integrating Inertial Measurement Unit (IMU) sensing and Artificial Intelligence (AI). The key technical components and methodology are as follows:

Technological Foundation:

IMU sensors will be used to monitor and assess human movement and posture. These sensors detect motion through accelerometers, gyroscopes, and magnetometers, allowing for precise gait analysis.

AI and Generative Adversarial Networks (GAN) will process the data to customize training regimens based on the individual's physiological and movement characteristics.

A Vicon 3D motion capture system will be used in conjunction with IMUs for validating and collecting data during the development phase.

Research Phases:

Year 1: Developing an AI-based gait training system using IMUs. This involves creating a gait database and balance training protocols using bilateral and unilateral movements.

Year 2: Optimizing the training system using AI and GAN to diversify the data and improve training efficacy.

Year 3: Clinical validation of the system by comparing results between participants undergoing IMU-based training versus standard physical exercises.

Training Protocols:

Exergame Environment: Participants engage in exercises within a virtual environment, which mimics real-world conditions but includes artificial elements to challenge balance and coordination.

Balance Training: Skateboard-based training focuses on unilateral leg movements, monitored by IMUs to provide feedback and adjust difficulty based on performance.

Data Analysis:

Gait Data: AI and GAN are used to generate personalized gait profiles, which will feed into the training system.

Statistical Analysis: Various statistical tests (e.g., ANOVA) will assess the effectiveness of the system compared to conventional rehabilitation methods.

This system aims to provide older adults with personalized rehabilitation, reducing fall risk and enhancing their quality of life.

• Study Type: Interventional
• Enrollment (Estimated): 120
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Hsu Wei-Li, Ph. D; Phone Number: 886-2-33668127; Email: wlhsu@ntu.edu.tw

Study Locations

• Taiwan
  • Taipei, Taiwan, 100
    • Recruiting
    • National Taiwan University, College of Medicine, School and Graduate Institute of Physical Therapy
    • Contact: Hsu Wei-Li, Ph. D; Phone Number: 886-2-33668127; Email: wlhsu@ntu.edu.tw

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

Aged between 18 and 80 years capable of independent walking-

Exclusion Criteria:

1. history of lower limb orthopedic surgery, ankylosing spondylitis, rheumatoid arthritis, osteoarthritis, and other medical joint diseases
2. Those who cannot communicate or follow instructions, and those with severe visual or hearing impairments
3. the neurological impairment or vestibular disorders, such as stroke, spinal cord injury, Meniere's syndrome.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Prevention
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: experimental group; IMU-based balance training	Other: IMU-based balance training; Leveraging AI technology to identify motion deficiencies, the experimental group will undergo IMU-based balance training
Other: control group; General health education or exercise training	Other: general health education or exercise training; general health education or exercise training

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Static Standing Balance Test	Balance Assessments	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)
Single Leg Standing Test	Balance Assessments	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)
Five Times Sit to Stand Test	Functional Tests	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)
Timed Up and Go Test	Functional Tests	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)
Six-Minute Walk Test	Functional Tests	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)
Over-ground walking	Walking test	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)
Walking on a treadmill	Walking test	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)
Delsys Trigno EMG analysis system	Three-Dimensional Motion Analysis	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)
Vicon Bonita	Three-Dimensional Motion Analysis	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)
Force plates	Three-Dimensional Motion Analysis	pre-training, post-training(after 6 weeks), follow-up(after 2 weeks)

Collaborators and Investigators

• Sponsor: National Taiwan University Hospital

Study record dates

Study Major Dates

• Study Start (Actual): November 3, 2023
• Primary Completion (Estimated): December 1, 2026
• Study Completion (Estimated): December 1, 2026

Study Registration Dates

• First Submitted: September 11, 2024
• First Submitted That Met QC Criteria: September 11, 2024
• First Posted (Actual): September 19, 2024

Study Record Updates

• Last Update Posted (Actual): November 19, 2025
• Last Update Submitted That Met QC Criteria: November 16, 2025
• Last Verified: October 1, 2025

More Information

Terms related to this study

• Keywords: Older Adults; Artificial Intelligence; Balance Rehabilitation; Inertial Measurement Unit Sensing
• Additional Relevant MeSH Terms: Motor Activity; Movement; Musculoskeletal Physiological Phenomena; Musculoskeletal and Neural Physiological Phenomena; Exercise
• Other Study ID Numbers: 202309084RIND

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No