Concurrent validity of skin-based motion capture systems in measuring dynamic lumbar intervertebral angles

Spine kinematics are commonly measured by external sensors such as motion capture and accelerometers. However, these skin-based measures cannot directly capture intervertebral motion of the lumbar spine. To date, research in this area has focused on the estimation of intervertebral kinematics using static trials but no study has analyzed agreement throughout the dynamic range of motion. This study investigated the agreement between skin-based sensors (accelerometers and motion capture) and quantitative fluoroscopy (QF) in measuring lumbar spine kinematics for the duration of complete flexion and extension motion in a healthy female population. Twenty female participants (age 30-57, BMI < 30) were guided through a standing flexion and extension bending protocol while spine kinematics were concurrently measured by QF (L2, L3, L4, L5, and S1) and motion capture sensors and accelerometers positioned over the spinous processes of L2, L4, and S1. Intervertebral angles (L2-L4, L4-S1, L2-S1) and individual vertebrae levels were compared between measures. Non-parametric limits of agreement between QF and skin-based markers were greatest at the end-range of motion for both flexion and extension, but differences increased variably between participants, sometimes over-and sometimes underestimating angles, thus, disproving the common assumption that it increases linearly. The two skin-based marker systems showed good agreement with one another showing that they can be used interchangeably but they can only be used to estimate lumbar spine kinematics. Normalizing angles to a change in angle and considering the posture of instrumentation would be beneficial to reduce potential sources of errors.