Evaluation of Lower Limb Exoskeleton for Improving Balance during Squatting Exercise using Center of Pressure Metrics

Abstract

The foot center of pressure (COP) variability is an important indicator of balance, particularly relevant for rehabilitation and training using wearable lower limb exoskeletons. This study aimed to evaluate the effectiveness of our exoskeleton in assisting squatting motion using the COP variability as a metric. Six human subjects performed alternate squatting and standing movements while their foot pressure and COP trajectories were recorded using insole pressure sensors. The exercises were performed under three conditions: i) no device, ii) unpowered device, and iii) device with optimal stiffness. Results showed that the variability of the COP trajectory in the anterior-posterior direction of the foot during squatting tended to be lower for the optimal stiffness condition than the no device and unpowered device conditions, indicating the potential usefulness of the device in improving balance during squatting. This study has implications for human-inthe-loop optimization and balance control of the exoskeleton based on COP.

Predicting the metabolic cost of exoskeleton-assisted squatting using foot pressure features and machine learning

“…In this paper, we analyzed the foot pressure data obtained from the study ( Kantharaju et al., 2022 ) on human-in-the-loop optimization of an ankle exoskeleton to minimize the physical effort (metabolic cost) of squatting using a machine learning method, random forest regression analysis. Based on our preliminary results ( Ramadurai et al., 2022 ), we hypothesized that metabolic cost can be estimated from CoP-derived metrics using machine learning methods like the random forest for improved time-efficiency and user comfort in HIL optimization. We also hypothesized that the random forest can reveal important CoP features that are correlated with the metabolic cost of squatting.…”

“…We used data from the 1st and 2nd days for our analysis. The 2nd day’s data was used for initial analysis, where we compared the CoP variability between unpowered and optimal assistance conditions of the exoskeleton ( Ramadurai et al., 2022 ). The 1st day’s data was used for the machine learning and foot pressure—metabolic cost correlation analysis, which are the main outcomes of this paper.…”

“…For the preliminary analysis ( Figure 1 ), six subjects’ foot pressure data from the validation study on the 2nd day of the experiment were selected based on data quality and analyzed ( Ramadurai et al., 2022 ). The CoP variability at the optimal assistance condition was compared with the unpowered device condition.…”

“…In our preliminary study ( Ramadurai et al., 2022 ), we conducted an initial analysis using previously collected data on personalized squat assistance using HIL optimization of a unilateral ankle exoskeleton ( Kantharaju et al., 2022 ), and we found that CoP variability tended to be lower (indicating higher postural stability) for the optimal assistance condition compared to the unpowered device condition ( Figure 1 ). Larger variability in the CoP trajectory has been associated with poorer ability to maintain balance ( Abrahamova and Hlavačka, 2008 ).…”

Personalized Wearable Ankle Robot Using Modular Additive Manufacturing Design

“…Ankle eversion taping can improve static and dynamic balance [21], and in/eversion assistance has reduced effort associated with balance in individuals with below-knee amputation [13]. The in/eversion assistance in a robotic AFO has been recently emphasized to improve muscle strength, promote a more comprehensive gait rehabilitation, including balance [9], and mitigate injury risk [22]. Implementing ankle in/eversion assistance in AFO could increase the capabilities of wearable robots during gait rehabilitation and training of elderly populations.…”