Somatosensory Plasticity in Pediatric Cerebral Palsy following Constraint-Induced Movement Therapy

Figure 3

CIMT-induced effects on ERP topography across less and more affected hand. (a) Spatial correlation between ERPs from the pre-CIMT less affected hand and each of the other Side and Session conditions as a function of time. Note that while there is high positive spatial correlation over the initial ~200 ms poststimulus onset across all conditions, only the post-CIMT more affected hand ERP exhibited a high spatial correlation with the pre-CIMT less affected hand ERP over the ~250–550 ms poststimulus period. Both other conditions exhibited spatial correlations near 0. (b) Hierarchical topographic cluster analysis identified 3 template maps that characterized the 250–550 ms poststimulus period across all conditions (see insets). Single-subject fitting based on spatial correlation between these template maps and individual ERPs yielded the average percentage of time each of these three template maps characterized each condition. The bar graphs show that Map 1 was both clinically relevant and also that its presence changed in a manner consistent with beneficial plasticity; its mean duration decreased in the more affected hand post- vs. pre-CIMT. The presence of Map 1 also demonstrated a nonsignificant trend for maladaptive plasticity; its mean duration increased in the less affected hand post- vs. pre-CIMT. The graphs concerning Map 2 show that despite differing pre- vs. post-CIMT patterns for the less affected and more affected hands, there was no reliable evidence that the presence of this map was clinically relevant. The presence of Map 3 did not differ across conditions.

(a)

(b)