[BIC-announce] Séminaire UNF _ Dr Gaël Varoquaux - lundi 20 août 1pm

[Claude Godbout]

LES SÉMINAIRES DE L’UNF / SEMINAR UNF SERIES


Présentateur/ Speaker:

Gaël Varoquaux, Ph.D.


Titre/  Title:

Mapping brain function with decoding: brain structures to predict mental
states.


Endroit/ Where:

CRIUGM – Local E1910 ( <http://www.criugm.qc.ca/en/contact.html>
http://www.criugm.qc.ca/en/contact.html)


Date/ When:

Lundi 20 août, 13h-14h/ Monday, August 20th 1pm-2pm


 

*La conférence sera présentée en anglais/The seminar will be presented in
English

 

Dr. Gaël Varoquaux is a tenured computer-science researcher at INRIA. His
research develops statistical learning tools for functional neuroimaging
data with application to mapping brain of cognition and pathologies. In
addition, he is heavily invested in software development for data science,
as project-lead for scikit-learn, one of the reference machine-learning
toolboxes, and on joblib, Mayavi, and nilearn. Varoquaux has contributed key
methods for functional brain atlasing, extracting brain connectomes,
population studies, as well as efficient models for high-dimensional
data-scarce machine learning beyond brain imaging. He has a PhD in quantum
physics and is a graduate from Ecole Normale Superieure, Paris.

 

Abstract:

Brain decoding, predicting mental process recruited based on observed brain
activity, is a central tool to characterize the functions of a brain
structure. Such conclusion based on a standard analysis of an fMRI
experiment would indeed be an invalid reverse inference. I will describe our
progress using decoding for brain mapping.

To come to general conclusions on the function of a brain region, it is also
necessary to explore a vast repertoire of cognitive function. Doing so
requires joint analysis of many studies. Such an analysis enables going
beyond the idiosyncrasies of a study, but it requires describing a huge
variety tasks with a common vocabulary. On 30 different fMRI studies, we
describe experiment conditions with 20 different experimental-psychology
notions [Schwartz NIPS 2013]. Using a predictive model that leverages
contrasts between conditions with similar stimuli, we can describe a
completely new experiment with these notions.

Importantly, prediction is perform on unseen studies, thus validating that
the decoder has learned maps specific to the notion of interest, and not to
the experimental paradigm. A precious outcome is a "reverse inference" atlas
of the predictive regions for these notions.

>From a statistical modeling standpoint, the challenge is that machine
learning models are not only used for their predictions, but also in an
inverse problem settings, to recover discriminant regions. Given the
scarcity of data and the dimensionality of the brain images, the problem is
very ill-posed and requires injecting various priors. I will summarize five
years of research in developing and validating regularizations that encode
efficiently priors well suited for neuroimaging, in particular to segment
and outline brain regions. Specifically, these draw ideas from
total-variation [Michel TMI 2011] or randomized-clustering [Varoquaux ICML
2012] approaches, and have resulted in stable, fast, and powerful decoders
[Hoyos-Idrobo 2017].

 

I hope to see you there / En espérant vous voir nombreux.

 

 

Pierre Bellec 

Assistant professor/Professeur adjoint sous octroi, Département
d'informatique et de recherche opérationnelle ( <http://diro.umontreal.ca/>
DIRO)

Director/Directeur, Unité de Neuroimagerie Fonctionnelle (
<http://unf-montreal.ca> UNF)

Researcher/Chercheur Centre de recherche de l'institut Universitaire de
gériatrie de Montréal ( <http://www.criugm.qc.ca/> CRIUGM)

Université de Montréal, Montréal, Canada

Phone +1 514 713 5596  

 <http://simexp-lab.org/brainwiki/doku.php?id=pierrebellec> Coordinates/c
<http://simexp-lab.org/brainwiki/doku.php?id=pierrebellec> oordonnées.

 <http://simexp-lab.org> Laboratory/Laboratoire SIMEXP

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.bic.mni.mcgill.ca/pipermail/bic-announce/attachments/20180817/79b25e53/attachment-0001.html>