结构化的小脑连通性支持恢复性的模式分离

美国哈佛大学医学院Wei-Chung Allen Lee研究团队发现，结构化的小脑连通性支持恢复性的模式分离。2022年11月23日，国际知名学术期刊《自然》在线发表了这一成果。

研究人员表示，小脑被认为有助于检测和纠正预定命令和执行命令之间的错误，对社会行为、认知和情感至关重要。运动控制的计算必须快速进行，以实时纠正错误，并且应该对模式之间的微小差异敏感，从而进行精细的错误纠正，同时对噪音有抵抗力。有影响力的小脑信息处理理论在很大程度上假设了随机网络连接，这增加了网络第一层的编码能力。然而，最大限度地提高编码能力会降低对噪声的抵抗力。

为了了解神经元回路如何解决这一基本的权衡问题，研究人员通过使用自动大规模透射电子显微镜和基于卷积神经网络的图像分割，绘制了小鼠小脑皮层的前馈连接。研究人员发现，回路的输入层和输出层都表现出冗余和选择性的连通性模体，这与普遍的模型形成了对比。数值模拟表明，这些冗余的、非随机的连通性模体增加了对噪声的抵抗力，而对整体编码能力的代价可以忽略不计。

(资料图)

这项工作揭示了神经元网络结构如何支持编码能力和冗余之间的权衡，并揭开了生物网络结构的原则，这对人工神经网络的设计有意义。

附：英文原文

Title: Structured cerebellar connectivity supports resilient pattern separation

Author: Nguyen, Tri M., Thomas, Logan A., Rhoades, Jeff L., Ricchi, Ilaria, Yuan, Xintong Cindy, Sheridan, Arlo, Hildebrand, David G. C., Funke, Jan, Regehr, Wade G., Lee, Wei-Chung Allen

Issue&Volume: 2022-11-23

Abstract: The cerebellum is thought to help detect and correct errors between intended and executed commands1,2 and is critical for social behaviours, cognition and emotion3,4,5,6. Computations for motor control must be performed quickly to correct errors in real time and should be sensitive to small differences between patterns for fine error correction while being resilient to noise7. Influential theories of cerebellar information processing have largely assumed random network connectivity, which increases the encoding capacity of the network’s first layer8,9,10,11,12,13. However, maximizing encoding capacity reduces the resilience to noise7. To understand how neuronal circuits address this fundamental trade-off, we mapped the feedforward connectivity in the mouse cerebellar cortex using automated large-scale transmission electron microscopy and convolutional neural network-based image segmentation. We found that both the input and output layers of the circuit exhibit redundant and selective connectivity motifs, which contrast with prevailing models. Numerical simulations suggest that these redundant, non-random connectivity motifs increase the resilience to noise at a negligible cost to the overall encoding capacity. This work reveals how neuronal network structure can support a trade-off between encoding capacity and redundancy, unveiling principles of biological network architecture with implications for the design of artificial neural networks.

DOI: 10.1038/s41586-022-05471-w

Source: https://www.nature.com/articles/s41586-022-05471-w

来源：科学网 小柯机器人

版权声明：本文转载仅仅是出于传播信息的需要，并不意味着代表本网站观点或证实其内容的真实性；如其他媒体、网站或个人从本网站转载使用，须保留本网站注明的“来源”，并自负版权等法律责任；作者如果不希望被转载或者联系转载稿费等事宜，请与我们接洽。