Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Master's Award: Canada Graduate Scholarships

McGill University

$17,500.00

Mechanisms of functional-MRI-based decoding of information conveyed by cortical columns

EF 2021-22

Résumé
Previous studies demonstrated that multivariate machine learning algorithms could decode visual stimuli from fMRI data. This result is surprising given the large size of the voxels relative to the mean cycle length of columns. The mechanism by which low-resolution imaging decodes information represented at a fine-scale relative to the voxel size is unclear. We hypothesize that the successful decoding of information conveyed by cortical columns relies on a combination of different mechanisms. The objective of our study is to determine the extent to which components of different spatial frequencies of the hemodynamic response in gray matter regions, venules, and veins carry information on neuronal responses organized in cortical columns. We will evaluate the differential contrasts and component-specific decoding performances using large voxels. Two marmosets will be trained to fixate on a fixation spot while oriented gratings stimuli are presented. We will image the responses in marmoset primary visual area to oriented gratings using multiple illumination wavelengths and compute relative changes in the content of deoxy-Hb, which is a reliable marker of BOLD fMRI responses. We will analyze separate and combined contributions to decoding from irregularities in the columnar organization (high-pass filtered data), large-scale organization (low-pass filtered data) from gray matter regions, and from venules and veins. The results of our experiments and modeling will dissect the mechanisms underlying fMRI-based decoding of information conveyed in cortical columns based on an experimental model with a known columnar organization.
Programme
Master's Award: Canada Graduate Scholarships
Bénéficiaire
Emplacement
Quebec, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
CIHR Funding Decisions Database

Tel que publié

Id
466894
Name
Griffiths Harold G
Region
Québec
Country
Canada
Orgtype
Institution Paid; Research Institution
Prcname2
Special Cases - Awards Programs
Projecttitle
Mechanisms of functional-MRI-based decoding of information conveyed by cortical columns
Programtype2
Studentship Programs
Orgnameinp2
McGill University
Cihramount2
$17,500
Cihrequipment2
$0
Programname2
Master's Award: Canada Graduate Scholarships
Keyworddelim
Brain; Cerebral Cortex; Cortical Columns; Decoding; Funtional Mri; Machine Learning; Optical Imaging; Orientation Columns; Support Vector Machine; Visual Cortex
Abstract
Previous studies demonstrated that multivariate machine learning algorithms could decode visual stimuli from fMRI data. This result is surprising given the large size of the voxels relative to the mean cycle length of columns. The mechanism by which low-resolution imaging decodes information represented at a fine-scale relative to the voxel size is unclear. We hypothesize that the successful decoding of information conveyed by cortical columns relies on a combination of different mechanisms. The objective of our study is to determine the extent to which components of different spatial frequencies of the hemodynamic response in gray matter regions, venules, and veins carry information on neuronal responses organized in cortical columns. We will evaluate the differential contrasts and component-specific decoding performances using large voxels. Two marmosets will be trained to fixate on a fixation spot while oriented gratings stimuli are presented. We will image the responses in marmoset primary visual area to oriented gratings using multiple illumination wavelengths and compute relative changes in the content of deoxy-Hb, which is a reliable marker of BOLD fMRI responses. We will analyze separate and combined contributions to decoding from irregularities in the columnar organization (high-pass filtered data), large-scale organization (low-pass filtered data) from gray matter regions, and from venules and veins. The results of our experiments and modeling will dissect the mechanisms underlying fMRI-based decoding of information conveyed in cortical columns based on an experimental model with a known columnar organization.
Competitiondate
202112
Approvedterm2
1 yr 0 mth
Pinamesdelim
Griffiths, Harold G
Approvedterm
1 an 0 mois
Cihrcontribution2
~ $17,500~ $0~