Reliability and Sources of Variability of 3D Kinematics and Muscle Activation Analysis of Newly-acquired Gait in Toddlers With a Typical Development and a Unilateral Cerebral Palsy

This study aim to assess lower limb kinematics and muscle EMG reliability during gait in typically developing toddlers and in toddlers with unilateral cerebral palsy with less than 6 months of independent walking experience and secondary to determine three sources of variability (between subjects, between sessions and between trials) using an optoelectronic system and a surface EMG system. The third aim was to identify the most reliable gait parameters.

This was a cross-sectional study conducted between January 2016 and May 2018 in the Motion Analysis Laboratory of the National Rehabilitation Center for Children (NRCC) "Dr. N. Robanescu", Bucharest (Romania). The toddlers with UCP were recruited among the inpatients and outpatients of the Pediatric Rehabilitation department. TD toddlers were recruited among the siblings of the toddlers with UCP and the children of people known to the investigators. Given the exploratory nature of this study, the target size of each group was set to a minimum of 10 participants per group.Toddlers from both groups were younger than 3 years old and independent walking experience of maximum 6 months. Every child was proposed to be evaluated in two gait analysis sessions within a period of 30 days. The second gait analysis session was planned approximately at 14 days from the first one. The two sessions were performed by the same experienced assessor.

The toddlers walked barefoot at self-selected pace. Each toddler was equipped with 19 reflective markers and 8 surface EMG electrodes. Data was collected with a BTS motion capture system equipped with 8 infrared cameras and 8 channels EMG system. The biomechanical model used was "Helen Hayes with medial markers".The EMG electrodes were positioned according to the SENIAM recommendations over the rectus femoris, the medial hamstrings, the tibialis anterior and the triceps surae.Visual 3D software (C-motion, Rockville, MD, USA) was used to build the kinematic model and identify gait cycles. A qualitative analysis was conducted for each gait cycle and each child in order to exclude the cycles during running or non-straight walking. Data was filtered and normalized. Three sources of variability (between subjects, between sessions and between trials) for kinematics and EMG data were quantified by calculating the standard deviations for each source.Standard Error of Measurement (SEM) for intertrials and intersessions were also calculated. For TD toddlers, a single value was reported as the quadratic mean for the right and left sides.