Suppr超能文献

在视觉噪声存在的情况下,猕猴 MT 神经元感受野大小的减小。

Reduction in receptive field size of macaque MT neurons in the presence of visual noise.

机构信息

Department of Neurophysiology, Graduate School of Medicine, Juntendo University, Bunkyo, Tokyo, Japan.

出版信息

J Neurophysiol. 2012 Jul;108(1):215-26. doi: 10.1152/jn.00710.2011. Epub 2012 Apr 11.

Abstract

The visual system faces a trade-off between increased spatial integration of disparate local signals and improved spatial resolution to filter out irrelevant noise. Increased spatial integration is beneficial when signals are weak, whereas increased spatial resolution is particularly beneficial when focusing on a small object in a cluttered natural scene. The receptive field (RF) size of visual cortical neurons can be modulated depending on various factors such as sensory context, allowing adaptive integration of sensory signals. In this study, we explored the spatial integration properties of neurons in macaque middle temporal visual area (MT). We hypothesized that spatial resolution would increase when high-contrast noise was presented simultaneously with a visual stimulus, enabling focus on a small object in a cluttered scene. To test this hypothesis, we mapped the RFs of MT neurons of two fixating monkeys in a 5 × 5 grid manner using a small patch of random-dot motion. To examine the effects of noise on RF profile, a dynamic noise (0% coherence dots) of varying diameter was concurrently presented at the RF center. We found that RF size decreased when noise diameter increased. Analyses based on the response normalization model and area summation provided evidence for the potential contribution of spatial summation properties within the RF and surround suppression to the apparent contraction of RF size. Our results suggest that MT neurons integrate smaller regions of motion signals when signals are embedded in noise, an efficient strategy to filter out surrounding noise.

摘要

视觉系统在增加不相关局部信号的空间集成和提高空间分辨率以滤除无关噪声之间面临权衡。当信号较弱时,增加空间集成是有益的,而当聚焦于杂乱自然场景中的小物体时,增加空间分辨率特别有益。视觉皮层神经元的感受野 (RF) 大小可以根据各种因素(如感觉上下文)进行调节,从而允许对感觉信号进行自适应整合。在这项研究中,我们探索了猕猴中颞视觉区 (MT) 神经元的空间整合特性。我们假设,当高对比度噪声与视觉刺激同时呈现时,空间分辨率会提高,从而能够聚焦于杂乱场景中的小物体。为了验证这一假设,我们使用一小片随机点运动以 5×5 网格的方式绘制了两只注视猴子的 MT 神经元的 RF。为了检查噪声对 RF 轮廓的影响,在 RF 中心同时呈现具有不同直径的动态噪声(0%相干点)。我们发现,随着噪声直径的增加,RF 大小减小。基于响应归一化模型和面积总和的分析为 RF 内和周围抑制的空间总和特性对 RF 大小的明显收缩的潜在贡献提供了证据。我们的结果表明,当信号嵌入噪声中时,MT 神经元整合较小的运动信号区域,这是一种过滤周围噪声的有效策略。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验