Skip to Main content Skip to Navigation

Source separation analysis of visual cortical dynamics revealed by voltage sensitive dye imaging

Abstract : The work presented in this thesis consists of a statistical source separation analysis of VSD imaging recordings in order to 1) separate the neural-related signal from the artefacts, and 2) investigate the dynamics of the spatio-temporal patterns of activity created by populations of cells that share common statistics. For this purpose, we performed VSD imaging of areas 17 and 18 in the anaesthetized and paralyzed cat. Full field sinusoidal luminance grating stimulations whose orientation was chosen among 4 or 6 cardinal values were presented moving each at two opposite directions. We developed an analysis method for VSD imaging that separates the neural response from the artefacts on single trials, and extracts the spatio-temporal patterns of activity that are related to functionally differentiable populations. Our denoising method permits to perform single trial analysis, for instance to study trial-to-trial variability of signal-to-noise ratio in neuronal population activity. Furthermore, we could also separate neuronal populations that show different response profiles to orientation and/or to direction of the drifting grating stimuli by using PCA and ICA. The compression and projection of V1 maps in response to drifting gratings on three spatiotemporally structured components allowed us to investigate cortical state dynamics and demonstrate the separation of state trajectories as a function of orientation. The sensitivity of our method appears much higher than that gained with conventional analysis of time courses of individual pixels. We hope in the future to provide a measurement framework to quantify and interpret the responses to more naturalistic stimuli
Document type :
Complete list of metadata
Contributor : Theses Bupmc Connect in order to contact the contributor
Submitted on : Friday, June 21, 2013 - 5:42:12 PM
Last modification on : Sunday, June 26, 2022 - 9:58:50 AM
Long-term archiving on: : Sunday, September 22, 2013 - 4:11:21 AM


  • HAL Id : tel-00836931, version 1


Esin Yavuz. Source separation analysis of visual cortical dynamics revealed by voltage sensitive dye imaging. Neurons and Cognition [q-bio.NC]. Université Pierre et Marie Curie - Paris VI, 2012. English. ⟨NNT : 2012PAO6551⟩. ⟨tel-00836931⟩



Record views


Files downloads