Skip to main content
Fig. 12 | Journal of NeuroEngineering and Rehabilitation

Fig. 12

From: Interaction of network and rehabilitation therapy parameters in defining recovery after stroke in a Bilateral Neural Network

Fig. 12

Comparison of different therapies for chronic stroke model with lesion size of A 5, B 10, C 15 and D 20, under different values of structural integrity of CC are given. Lesion is induced in the penultimate layer having a total of 30 nodes. [Key for x-axis: 10L (CIMT Exploratory under local plasticity), 11L (BMT Exploratory under local plasticity), 10G (CIMT Exploratory under global plasticity), 11G (BMT Exploratory under global plasticity)]. For lesion size of 5 and 10 nodes, the network does not show preference of one plasticity condition over the other (p > 0.1 under all CC integrity values when comparing 10L with 10G and 11L with 11G). However, for a lesion size of 15, 70% and 90% show preference of global plasticity over local plasticity (p < 0.05 when comparing 10L and 10 G, p > 0.1 when comparing 11L and 11G). However, for a lesion size of 20, regardless of the value of CC integrity, the network prefers global over local plasticity (p < 0.005 when comparing 10L and 10G for all CC integrity values, and p < 0.009 when comparing 10L and 10G). Thus, as the integrity of the connections goes down, the network tries to compensate with retraining of only the local connections, and preferring retraining of the entire network (global plasticity) only when the damage is too high (20 nodes)

Back to article page