Decoding Methods for Predicting Postsynaptic Responses to Spike Trains

[421388]劳伦斯·阿伯特神经系统使用电尖峰序列来传输信息，这些信息在运动神经元和肌肉之间的突触上转化为行动。各种不同模式的神经元尖峰序列到达神经肌肉连接处，并由肌肉转化为同样广泛的肌肉收缩。这些研究人员希望了解这种转导过程，并确定运动神经元输出的变化对肌肉反应的影响。为了做到这一点，他们将构建一个数学形式，使他们能够预测肌肉对任何尖峰训练的反应。预测将通过线性积分核和数学神经网络进行，使用学习规则，随着时间的推移提高系统的性能，同时各种尖峰序列被传递到肌肉中。由此产生的网络将能够预测新型尖峰列车的影响。此外，模型的形式将提供关于突触转导背后的生物物理过程的重要信息。这些技术可以应用于神经系统内神经元之间的突触，以及神经元和肌肉之间的突触。这项工作将对我们理解大脑如何处理来自环境的刺激并将其转化为最终行为产生重要影响。

9421388 Laurence Abbott The nervous system uses electrical spike trains to transport information that is converted into actions at the synapses between motor neurons and muscles. A wide variety of different patterns of the neuronal spike trains arrive at the neuromuscular junction and are converted by the muscle into an equally wide range of muscle contractions. These investigators would like to understand this transduction process and determine what effect changes in the output of the motor neuron have on the muscle response. To do this they will construct a mathematical formalism that allows them to predict the response of the muscle to any spike train. The prediction will be made by linear integration kernels and mathematical neural networks using learning rules that improve the performance of the system over time while a variety of spike trains are delivered to the muscle. The resulting network will then be able to predict the effect of novel spike trains. In addition, the form of the model will provide important information about the biophysical processes underlying the synaptic transduction. The techniques that will be developed can then be applied to synapses between neurons within the nervous system as well as those between neurons and muscles. This work will have an important impact on our understanding of how the brain processes stimuli from the environment and converts them into a resultant behavior.