ITERATOR: A 3D Gait Identification from IR-UWB Technology.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Human Motion Analysis

Background:

• Increasing need for non-intrusive health monitoring in aging populations.
• Limitations of wearable sensors (privacy, security, comfort) impacting user adoption.
• Gait disorders require accurate, unobtrusive detection methods.

Purpose of the Study:

• To propose and validate a novel non-contact 3D human motion model for gait disorder identification.
• To utilize impulse radio ultra-wideband (IR-UWB) technology for gait analysis.
• To compare the proposed IR-UWB model with Kinect Xbox One for gait assessment.

Main Methods:

• Development of a non-contact 3D human motion model using IR-UWB, spherical trigonometry, and vector fields.
• Data collection from 24 participants (20 normal gait, 4 spasticity).
• Transformation of radar backscattered responses and Kinect skeletal data using vector algebra.
• Determination and validation of thigh angles using root mean square error (RMSE).

Main Results:

• The proposed IR-UWB model successfully distinguished between normal and spastic gait patterns.
• Validation against Kinect skeletal data showed a root mean square error (RMSE) of less than 0.5 for thigh angle measurements.
• The system effectively differentiates lower and upper body segments based on recorded height data.

Conclusions:

• The non-contact IR-UWB 3D human motion model is a viable and accurate method for gait disorder identification.
• This approach offers a privacy-preserving and comfortable alternative to existing wearable monitoring systems.
• The findings support the potential of IR-UWB for remote, unobtrusive health monitoring, particularly for the aging population.