Development and Validation of 2D-LiDAR-Based Gait Analysis Instrument and Algorithm

• Main Authors: Seongjun Yoon, Hee-Won Jung, Heeyoune Jung, Keewon Kim, Suk Koo Hong, Hyunchul Roh, Byung-Mo Oh
• Format: Article
• Language: English
• Published: MDPI AG 2021-01-01
• Series: Sensors
• Subjects: frailty; gait; physical performance; sarcopenia; LiDAR
• Online Access: https://www.mdpi.com/1424-8220/21/2/414

Acquiring gait parameters from usual walking is important to predict clinical outcomes including life expectancy, risk of fall, and neurocognitive performance in older people. We developed a novel gait analysis tool that is small, less-intrusive and is based on two-dimensional light detection and ranging (2D-LiDAR) technology. Using an object-tracking algorithm, we conducted a validation study of the spatiotemporal tracking of ankle locations of young, healthy participants (<i>n</i> = 4) by comparing our tool and a stereo camera with the motion capture system as a gold standard modality. We also assessed parameters including step length, step width, cadence, and gait speed. The 2D-LiDAR system showed a much better accuracy than that of a stereo camera system, where mean absolute errors were 46.2 ± 17.8 mm and 116.3 ± 69.6 mm, respectively. Gait parameters from the 2D-LiDAR system were in good agreement with those from the motion capture system (r = 0.955 for step length, r = 0.911 for cadence). Simultaneous tracking of multiple targets by the 2D-LiDAR system was also demonstrated. The novel system might be useful in space and resource constrained clinical practice for older adults.