Fall Risk Assessment Dataset: Older-Adult Participants undergoing the Time Up and Go Test

Published: 26 September 2023| Version 2 | DOI: 10.17632/78mw556vjc.2
Boonserm Kaewkamnerdpong,


The dataset comprises signal data collected from IMU sensors during the administration of the Time Up and Go (TUG) test for assessing fall risk in older adults. The dataset is divided into two main sections. The first section contains personal, behavioral, and health-related data from 34 participants. The second section contains signal data from tri-axial acceleration and tri-axial gyroscope sensors embedded in an IMU sensor, which was affixed to the participants' waist area to capture signal data while they walked. The chosen assessment method for fall risk analysis is the TUG test, requiring participants to walk a 3-meter distance back and forth. To prepare the dataset for subsequent analysis, the raw signal data underwent processing to extract only the walking periods during the TUG test. Additionally, a low-pass filter technique was employed to reduce noise interference. This dataset holds the potential for the development of effective models for fall risk detection based on insights garnered from questionnaires administered to specialists who observed the experiments. The dataset also contains deanonymized participant information that can be explored to investigate fall risk, along with other health-related conditions or behaviors that could influence the risk of falling. This information is invaluable for devising tailored treatment or rehabilitation plans for individual older adults.


Steps to reproduce

Participant data and IMU sensor signals were collected. Participant information, obtained through interviews, questionnaires, and assessments, focused on personal and health details. Signal data from an IMU sensor linked to an ESP8266 controller within the measuring device captured raw signals of acceleration and gyroscope sensors during the Timed Up and Go (TUG) test. Signal data covered three TUG rounds. The data was stored on a memory card connected to the measuring device's controller. Raw data underwent pre-processing, eliminating extraneous signals during idle periods before and after the task.


King Mongkut's University of Technology Thonburi


Fall in the Elderly, Accelerometer, Sensor, Gyroscope