PROJECT SUMMARY Physical disability in individuals with cerebral palsy (CP) is caused by a failure to learn and/or refine motor control during the first two years of life, during a critical period for neuroplasticity in motor control centers. Despite this, very little is known about how infants at high risk for CP learn (or fail to learn) to move and acquire early locomotor skills that are building blocks for later functional mobility. The long-term goal of our work is to reduce physical disability in individuals with CP by promoting the development of evidence-based motor training strategies to train and shape motor learning during infancy and toddlerhood. The objective of this project is to characterize the evolution of locomotor learning over the first 18 months of life in infants at high risk for CP. To characterize how locomotor skill is learned (or not learned) during this critical period, we will combine our established protocols using robust, unbiased robotic and sensor technology to longitudinally study infant movement across three consecutive stages during the development of impaired human motor control – early spontaneous movement, prone locomotion (crawling), and upright locomotion (walking). The spontaneous leg movements of sixty infants will be captured in real time using wearable sensors in their natural environment over the first 4 months of life. Infants who remain at high risk for CP at 4 months of age, which we expect to be approximately 50% of those enrolled, will progress to novel locomotor training protocols using robotic technology for crawling (5-9 months of age) and walking (9-18 months of age). The locomotor training protocols are introduced before the infants’ locomotor milestones emerge or fail to emerge, and are reward- and error-based, allowing ample opportunities for error learning while providing assistance to move when even small effort is made (reinforcement). Repeated measures of training characteristics and locomotor skill will quantify locomotor learning over time in infants at high risk for CP. Key training characteristics, combined with neurobehavioral moderators such as cognition and motivation to move, will be identified and used to develop a predictive model of locomotor learning. We hypothesize that experiencing and responding to error is a key mechanism for locomotor learning in CP, that learning is mediated by neurobehavioral factors outside of training, and that critical thresholds of error and other training characteristics predict the transfer of locomotor skill from prone to upright locomotion. Our team includes experts in motor control and development, computer science, robotics, biomechanics and neonatology. Characterizing and predicting locomotor learning in infants at high risk for CP will inform the design of future multi-site efficacy studies and the development of new treatments to shape motor learning during the critical years of the development of motor control, supporting the Eunice Kenne...