Measuring markers of aging and knee osteoarthritis gait using inertial measurement units

Abstract

Differences in gait with age or knee osteoarthritis have been demonstrated in laboratory studies using optical motion capture (MoCap). While MoCap is accurate and reliable, it is impractical for assessment outside the laboratory. Inertial measurement units (IMUs) may be useful in these situations. Before IMUs are used as a surrogate for MoCap, methods that are reliable, repeatable, and that calculate metrics at similar accuracy to MoCap must be demonstrated. The purpose of this study was to compare spatiotemporal gait parameters and knee range of motion calculated via MoCap to IMU-derived variables and to compare the ability of these tools to discriminate between groups. MoCap and IMU data were collected from young, older, and adults with knee osteoarthritis during overground walking at three self-selected speeds. Walking velocity, stride length, cadence, percent of gait cycle in stance, and sagittal knee range of motion were calculated and compared between tools (MoCap and IMU), between participant groups, and across speed. There were no significant differences between MoCap and IMU outcomes, and root mean square error between tools was ≤0.05 m/s for walking velocity, ≤0.07 m for stride length, ≤0.5 strides/min for cadence, ≤5% for percent of gait cycle in stance, and ≤1.5° for knee range of motion. No interactions were present, suggesting that MoCap and IMU calculated metrics similarly across groups and speeds. These results demonstrate IMUs can accurately calculate spatiotemporal variables and knee range of motion during gait in young and older, asymptomatic and knee osteoarthritis cohorts.

Introduction

Differences in gait between young and older adults with or without knee osteoarthritis are well-documented. These differences include slower walking speed (Bohannon, 1997), shorter stride lengths, higher cadence, longer relative stance phases (Murray et al., 1969), and smaller sagittal knee range of motion (Boyer et al., 2017, Messier et al., 1992) for older compared with young adults, and in adults with vs without knee osteoarthritis. Often, these differences are identified using optical motion capture (MoCap). MoCap allows for precise measurement of positions of markers on the body, enabling calculation of kinematics in a controlled setting, but data collection is labor-intensive and requires staff with advanced technical skills and expensive equipment. This limits the use of MoCap for routine gait analyses to monitor longitudinal change or assess real-world function. For certain variables, this limitation may be overcome using wearable sensors, including inertial measurement units (IMUs). However, before IMUs are used as a surrogate for MoCap, the validity of IMU-derived metrics must be reported, and reliable and reproducible methods for IMU gait analysis must be demonstrated. This would allow researchers to determine if group differences or intervention outcomes previously demonstrated with MoCap extend to real-world settings and would also enable proper comparison of findings across IMU studies.

The use of IMUs in gait analysis has increased in prevalence, but methodology is not yet standardized. Many studies have used IMUs to measure spatiotemporal variables or joint angles or angular displacements during gait in healthy young (Findlow et al., 2008, Jaysrichai et al., 2015, Lebel et al., 2017, Monda et al., 2015, Nüesch et al., 2017, Rebula et al., 2013, Washabaugh et al., 2017) or older (Monda et al., 2015) adults or in individuals with knee osteoarthritis (Bolink et al., 2015, Bolink et al., 2012, Calliess et al., 2014, Chiang et al., 2017, Kobsar et al., 2017, Kobsar and Ferber, 2018). Several of these studies compared IMU and MoCap metrics (Findlow et al., 2008, Jaysrichai et al., 2015, Lebel et al., 2017, Monda et al., 2015, Nüesch et al., 2017, Rebula et al., 2013, Washabaugh et al., 2017). While most of these studies report agreement between technologies, IMU methods are often unclear as many studies used proprietary manufacturer-provided algorithms for segment orientations, drift correction, and gait event detection (Findlow et al., 2008, Monda et al., 2015, Nüesch et al., 2017, Washabaugh et al., 2017). The use of manufacturer-provided algorithms is convenient, but may result in a lack of standardized, reproducible methodology, making it difficult to determine whether different studies are comparable.

Standardization of reproducible methodology may be achieved in a similar manner for IMUs as has been done for MoCap. Investigators can collect IMU data in raw format (akin to raw marker trajectories in MoCap), report procedures for orienting IMU coordinate systems to global or segment coordinate systems (similar to laboratory and anatomical calibration in MoCap), and describe details of IMU data transformation and processing used to calculate outcome variables. Several groups of investigators have worked to develop IMU methods including sensor calibration to global and segment coordinate systems (Cain et al., 2016), calculation of spatiotemporal variables (Rebula et al., 2013), and correction for integration drift (Ojeda and Borenstein, 2007). These works provide reproducible IMU methodology that researchers can use to measure features of gait without proprietary software. However, to our knowledge, the ability of these IMU methods to both accurately capture variables of interest and detect differences between clinically-relevant groups similarly to MoCap has not been demonstrated.

Therefore, the purpose of this study was to compare spatiotemporal gait parameters and sagittal knee range of motion calculated via MoCap and IMU and to compare the ability of MoCap and IMU calculations to discriminate between groups as a first, pilot step towards creating standards for gait analysis using IMUs. We compared metrics between tools and across three self-selected walking speeds for young and older asymptomatic adults, and for older adults with knee osteoarthritis. Towards the goal of identifying parameters and methods that could be applied to monitor real-world gait patterns, we compared variables that are easily calculated from raw IMU data, and we used methods that could be reproduced and that do not require assumptions about sensor placement, participant walking speed, or data length. Where applicable, we used previously-published methods (e.g., (Cain et al., 2016, Ojeda and Borenstein, 2007, Rebula et al., 2013)) and reported details of techniques to enable reproducibility in future studies.