Recurrence Quantification Analysis of Postural Stability with Pendant-Based IMU under Varying Balance Conditions

Abstract

Recurrence quantification analysis (RQA) has been widely implemented as a nonlinear data analysis method in multiple research areas and has shown good potential for the study of human postural stability. Dynamic features identified by RQA may be useful for characterizing balance disorders using inertial measurement units (IMUs) placed on the lower back near the body’s center of mass. However, wearing IMU sensors may be challenging for some specific populations at risk for falling such as pregnant women due to discomfort and lack of fit. A novel pendant-based IMU sensor was proposed and shown to have good sensitivity to postural instability in different standing tasks, when compared with the sensor worn on the lower back with a waist belt. This study’s objective is to evaluate the sensitivity of the pendant sensor to balance perturbations using RQA-the nonstationary method. Six RQA measurements (%recurrence, %deterministic, linemax, entropy, %laminarity, trapping time) were extracted for data analysis. The objective of this research is to validate whether the pendant-based sensor can distinguish different balance conditions as well as or better than a sensor on a waist belt using RQA measurements.