Talk by Prof. Hideaki Shimazaki (CSN Virtual Seminar)

We hereby announce the next talk in 'CSN Virtual Seminar' series

Deciphering hidden circuits from higher-order statistics of neural activity

Speaker: Prof. Hideaki Shimazaki, Kyoto University

Abstract
Identifying neural circuitry from living animals' neural activity is fundamental to neuroscience. Traditional methods infer such connectivity via correlation analysis between pairs of observed neurons. However, the observed neurons' co-activities reflect influences of common inputs from unobserved neurons. As such, it is critical to consider if we can identify the input structure from unobserved neurons from the collective activity of observed ones. The activity of a neural population depends not only on the circuit structure but also on the individual neuron's nonlinear input-output relationship. Modelling this relationship might enable circuit structure inference from the statistical structure of the population activity. In this talk, I will introduce our recent study [1] led by Dr Shomali, who constructed a method for inferring hidden common-input architecture from higher-order statistics of neural activities using an in-vivo neuron model under the balanced excitation/inhibition conditions. Application of the method to mammalian visual neurons suggests that an excitatory-to-pairs input structure ubiquitously underlies the sparse population activity.

[1] Shomali SR, Ahmadabadi MN, Rasuli SN, Shimazaki H. Uncovering hidden network architecture from spiking activities using an exact statistical input-output relation of neurons. (2023) Communications biology, 6, 169 s42003-023-04511-z (access)

You can dial in using the following zoom room information:
https://zoom.us/j/99314955130?pwd=YXkvaW1wY2s2OHJsMFRWVDNtcERiUT09
Meeting ID: 993 1495 5130
Passcode: 4pM2Z8