Fig. 2: Time-varying kinematics, electromyography (EMG), and neural population activity during walking. | Nature Communications

Fig. 2: Time-varying kinematics, electromyography (EMG), and neural population activity during walking.

From: Regional specialization of movement encoding across the primate sensorimotor cortex

Fig. 2

a Illustrative chronophotography of 7 successive corridor steps. b Right hindlimb kinematics, EMG, and neural signals recorded from Mk-Ek during 5.26 seconds of walking across the corridor. Top panel: changes in hip, knee and ankle joint angles from the right hindlimb contralateral to neural recordings. Middle panel: EMG activity from six hindlimb muscles. Bottom panel: raster plots reconstructed from the multiunit activity of 160 microelectrodes distributed over three Utah arrays implanted in the hindlimb region of left PMd, M1, and S1. The gray and white spaces at the bottom indicate the duration of the stance and swing gait phases of the right hindlimb, respectively. c Muscle activity varies substantially across tasks. The graphs show the mean peri-gait EMG of six hindlimb muscles recorded in Mk-Ek for all five tasks. d Hindlimb kinematics of both monkeys differs substantially across tasks. Top: We computed 58 variables from right hindlimb kinematic recordings of each gait cycle. These variables quantified different kinematic features of monkeys’ locomotor patterns (Supplementary Table 1). This dataset was arranged in a matrix with variables as the matrix columns and each row representing one gait cycle. Data from all five locomotor tasks were pooled together in a single matrix and z-scored across columns. We then applied principal component analysis on this dataset and visualized the outcome by plotting the dataset in a space spanned by the two leading principal components (PCs). The data for each task is represented by an ellipsoid with the center and principal semi-axis as the mean and standard deviation calculated across all the gait cycles for that task. Middle: The bar plots show the mean of two variables used for the principal component analysis: the step height (number of steps: Mk-Ek: corridor: 11, ladder: 10, stairs: 12, obstacles: 11, and treadmill: 23) and minimum knee angle across steps (number of steps: Mk-Nt: corridor: 31, ladder: 37, stairs: 14, obstacles: 10, and treadmill: 25). Bottom: The graphs show the mean peri-gait hip, knee and ankle angles for all five tasks. e Confusion matrix reporting decoding accuracy of classifying a task from EMG envelopes in Mk-Ek. f Confusion matrix reporting decoding accuracy of classifying a task from kinematic trajectories in Mk-Nt. Source data are provided as a Source Data file.

Back to article page