Pii: S0893-6080(98)00076-8
نویسندگان
چکیده
In lamprey, the supraspinal control of velocity is mainly accomplished by the reticulospinal (RS) system. During locomotion, RS neurones are rhythmically active with a cycle duration corresponding to the duration of the swim cycle. While the velocity of the muscular contraction wave changes as swimming velocity changes, the conduction velocity of RS axons remains constant. Thus, an action potential generated during a specific phase of the swim cycle will, depending on swimming velocity, provide input to a particular downstream segment during different phases of its rhythmic activity. In order to investigate the importance of this effect for the control of locomotion, the temporal and spatial characteristics of the propagation of the population of action potentials along RS axons in the spinal cord were investigated. The results suggest that if RS neurones are recruited independently of their sizes and conduction velocities, a phasic wave of action potentials in these fibers will reach some segments during the inhibited phase of their rhythmic activity. Such an input could hinder a smooth propagation of the contraction wave and disrupt swimming. In contrast, by recruiting successively larger and hence more rapidly conducting neurones for successively more rapid swimming, the phasic wave of action potentials may propagate with the same velocity as that of the muscular contraction wave. Under such conditions, reticulospinal activity would support and stabilise locomotion. q 1998 Elsevier Science Ltd. All rights reserved.
منابع مشابه
Explanation of the "virtual input" phenomenon
We write this letter to comment on the "virtual input" phenomenon reported by Thaler (Neural Networks, 8(1) (1995) 55-65). The author attributed the phenomenon to the network's ability to perform pattern classification and completion, and reported that pruning probability affects the number of virtual inputs observed. Our independent study of Thaler's results, however, reveals a simpler explana...
متن کاملModel selection in neural networks
In this article, we examine how model selection in neural networks can be guided by statistical procedures such as hypothesis tests, information criteria and cross validation. The application of these methods in neural network models is discussed, paying attention especially to the identification problems encountered. We then propose five specification strategies based on different statistical ...
متن کاملNeural classifier construction using regularization, pruning and test error estimation
In this paper we propose a method for construction of feed-forward neural classifiers based on regularization and adaptive architectures. Using a penalized maximum likelihood scheme, we derive a modified form of the entropic error measure and an algebraic estimate of the test error. In conjunction with optimal brain damage pruning, a test error estimate is used to select the network architectur...
متن کاملParallel and robust skeletonization built on self-organizing elements
A massively parallel neural architecture is suggested for the approximate computation of the skeleton of a planar shape. Numerical examples demonstrate the robustness of the method. The architecture is constructed from self-organizing elements that allow the extension of the concept of skeletonization to areas remote to image processing.
متن کاملExploiting inherent relationships in RNN architectures
We provide the relationship between the learning rate and the slope of a nonlinear activation function of a neuron within the framework of nonlinear modular cascaded systems realised through Recurrent Neural Network (RNN) architectures. This leads to reduction in the computational complexity of learning algorithms which continuously adapt the weights of such architectures, because there is a sm...
متن کامل