Decomposition of spontaneous movements of infants as combinations of limb synergies.

We hypothesized that a variety of limb movements in infants, including spontaneous movements and movements during interactions with the environment, can be represented as combinations of limb synergies, which are building blocks for generating coordinated movements of multiple limbs. A decomposition algorithm based on a nonnegative matrix factorization was applied to the discrete data segments taken from continuous data of limb movements in 298 infants (age, 3-4 months). The data were linearly decomposed into bases, which were referred to as synergies. The results showed that approximately 70% of the variance in the velocity profiles of the data segments of the four limbs can be explained by a set of five simple synergies that represent single-limb movements and the synchronous movement of all limbs. The present method showed that the complex properties of limb movements can be represented as combinations of synergies. Furthermore, comparisons of movement patterns across different age groups showed that in older infants, the contribution ratios of each synergy were different between spontaneous movements and movements during playing with a toy, whereas in younger infants, there were no differences in the contribution ratios between the different movement conditions. These results demonstrate that decomposition into limb synergies is useful for determining the spatiotemporal properties of interlimb coordination during spontaneous movements and task-constrained movements in infants.