z-bio 2025-04-28 92

　　Schematic diagram of the effect of visual stimuli on the topological structure of the visual cortex functional network in mice

　　Supported by the National Natural Science Foundation of China (Grant No. 92370116) and other grants, the team led by Associate Professor Xiaoxuan Jia from Tsinghua University, along with Assistant Professor Hannah Choi from Georgia Institute of Technology and Associate Professor Joel Zylberberg from York University, conducted in-depth research on the dynamic processing mechanism of mouse visual cortex networks in response to various types of visual stimuli. The research results, titled "Stimulus type shapes the topology of cellular functional networks in mouse visual cortex", were published online on July 9, 2024 in the international academic journal Nature Communications. The paper link is: https://www.nature.com/articles/s41467-024-49704-0 .

　　The brain is composed of millions of neurons forming a complex network, and the interactions between these neurons determine how information is extracted from the external environment and used to guide behavior. In a short time scale, the anatomical connections between neurons usually remain stable, while the functional networks that reflect the interaction between neurons can quickly and dynamically adjust with different input stimuli. How to systematically reveal the regulatory mechanism of visual input on neuronal functional connections is a key scientific issue in visual system research.

　　Most existing studies have found it difficult to achieve single neuron resolution or have not fully considered the neural activity of multiple stimulus types and visual cortex regions, thus unable to accurately reveal the dynamic characteristics of visual networks in input changes. In response to this challenge, the research team utilized the electrophysiological neural signal data from the Allen Institute to predict the possible positive and negative, bidirectional, and multi synaptic connections between neurons through directed edge analysis based on neuron discharge time series. They proposed a randomized model to preserve the positive and negative directed edge distribution of the real network, and systematically studied the non random characteristics of functional networks from multiple topological scales.

　　The research results show that specific triadic connectivity patterns are key information processing units in the visual cortex functional network. When facing different types of visual inputs, neurons can dynamically reassemble functional networks with similar local topological structures but different global characteristics, thereby efficiently processing diverse tasks. This discovery not only deepens the understanding of the dynamic information processing mechanism of the visual system, but also provides important inspiration for the development of next-generation interpretable artificial intelligence with high adaptability and memory ability.