Examination of inertial measurement units to evaluate lower body segmental angles in persons with autism

Inertial measurement units (IMU's) have been used to collect human movement data in practical, real-world environments in both non-pathological and pathological gait. IMU sensors have produced strong results for their validity and repeatability compared to the gold-standard of motion capture, specifically in college-aged populations. Due to their small size and lightweight design, IMU's can be very convenient for the collection of movement data across various populations. Given the unique movement patterns reported in persons with autism (i.e., variability both intra- and inter-subject), the purpose of this proposed study was to compare gait events and segmental angles of the foot, shank, and thigh during gait via IMU's compared to the "gold standard" of motion capture. Gait events were calculated using the superior-inferior acceleration/anterior-posterior acceleration (SIacc/APacc) methodology and the 10 N vertical ground reaction force (vGRF) methodology. Gait event detection was matched between SIacc/APacc and vGRF methods for heel strike yet significantly different for toe off. Our results support the existing literature suggesting an agreement in sagittal plane lower extremity segmental angles between IMU's and motion capture, with less agreement between modalities in the frontal and transverse planes. However, these findings should be approached with caution and accompanied by recommendations that take into account the attributes of each participant (e.g. stimming, sensitivities). Future work investigating specific functional tasks to define the axis system and limit out-of-plane motion is needed in this population.