‘Holistic’ representation of learned complex items in bursting neurons

报告题目： ‘Holistic’ representation of learned complex items in bursting neurons

报  告  人：廖祥 (重庆大学 神经智能研究中心 )

时    间：  2019年6月5日（周三）下午15:40  

地    点：  理科楼 LE201  

邀请人：  柴一晟 （物理学院，低温物理实验室）

报告摘要：

Complex sensory stimuli, such as chords in a song melody, can be reliably recognized. However, it is unclear how these complex sensory items are represented by single neurons in the brain. Here, we trained mice to associate sound stimuli with voluntary licking and investigated learning-induced transformation of neurons in primary auditory cortex. We demonstrate that learning transformed a small fraction of neurons from the common singlet firing to a bursting mode that was associated with large Ca2+ transients. This transformation was reversed by behavioral extinction. Strikingly, in mice that learned to associate licking with chords consisting of 4 pure tones, we encountered not only neurons responding to the constituent pure tones but also ‘holistic’ neurons responding to just the chord but not to the individual components. This reliable and specific representation of the learned chords by single learning-transformed ‘holistic’ bursting neurons reveals a high-level computation in a primary sensory cortex.

报告人简介：

廖祥，男，重庆大学神经智能研究中心副研究员。目前开展神经科学和智能技术的交叉研究，通过光学成像技术，研究皮层神经网络功能，并建立成像数据的自动化分析技术。前期在Cerebral Cortex, Brain Structure and Function, Biomedical Optics Express，IEEE Transactions on Biomedical Engineering等国际期刊上共发表SCI论文15篇，其中第一或通讯作者7篇，1区6篇。