[BIC-announce] New course now available : BMDE610 - Functional neuroimaging fusion - Registration is open

[Christophe Grova]

Dear all,

Just a friendly reminder that the new course I propose is open for registration.
There was a small technical problem that was blocking some registrations, it should be solved now.
Don't hesitate to contact me if you have any questions (regarding prerequisites notably) or problem to register.

I am happy to announce you that a new graduate course BMDE610, entitled Functional neuroimaging fusion will be offered in Winter 2012 in Biomedical Engineering Dpt

Most of the lectures will be done by myself  + few invited lecturers.

http://www.mcgill.ca/global/php/coursepopup.php?Course=BMDE%20610<UrlBlockedError.aspx>

Registration is now open on Minerva !

Prerequesites are ECSE 305 and MATH 223, or equivalent to be discussed with myself.
The main prerequesites actually consist in basic notions of Probablity and Linear Algebra.

Course description:
BMDE610: Biomedical engineering: Multimodal data fusion of electrophysiology and functional neuroimaging data, including: detailed description of source localization methods for Electro- and MagnetoEncephaloGraphy data, analysis of brain hemodynamic activity through simultaneous recordings with electrophysiology, analysis and reconstruction of Near Infra-RedSpectroscopydata, modelling of the neurovascular coupling, validation methodology.

Objectives of the course
The main objectives of this new course BMED 610 is to address specific issues related to data fusion of functional neuroimaging data. In this context, we will define “multimodal fusion” as the use of more than one modality to analyze functional neuroimaging data. For instance, fusion could be considered when anatomical information extracted from Magnetic Resonance Imaging (MRI) data is used as prior information to guide source localization of Electro- or Magneto-EncephaloGraphy (EEG vs MEG) data. Data fusion will also be considered when EEG markers are used to analyse functional MRI data, when studying for instance the hemodynamic response to spontaneous epileptic discharges detected on scalp EEG.

The objectives of the course are to introduce methods necessary to address the difficult problem of functional neuroimaging data fusion. The first part of the course will be dedicated to a detailed presentation of methods for which source localization has to be inferred from data acquired on the scalp: (1) in electrophysiology when using EEG or MEG, and (2) for hemodynamic activity when using Near Infra Red Spectroscopy (NIRS) data. The second part of the course will concern the link between bioelectrical activity and hemodynamic processes, the so-called neurovascular coupling. Applications for which simultaneous recordings of more than one modality are required will be presented, as well as related technical and methodological difficulties. The course will end with a state of the art of the models proposed to characterize the neurovascular coupling.

After completing this course, the students will be able to perform source localization from EEG and MEG data, which is far from trivial for most available softwares. They will also acquire sufficient knowledge to start analyzing hemodynamic data (fMRI, NIRS) and to studying the neurovascular coupling.

Evaluation of the course:
- 30%: Mid-term exam: theoretical questions related to the course (1h).
- 70%: Final project: detailed analysis of an article or a particular application of neuroimaging data fusion, with specific emphasis on validation methodology. This project will consist in a report (35%) and will be followed by a 15min oral presentation (35%) during the last sessions of the course. The objective of the project is to present, in details, the added value of using data fusion in a specific application context.


Preliminary syllabus
Week 1: Basis of electrophysiology (EEG, MEG) and hemodynamic signals (fMRI, NIRS)
Week 2: Forward model in EEG/MEG and in NIRS modelling
Week 3: Validation methodology in neuroimaging
Week 4: Source localization: equivalent current dipole and dipole scanning approaches
Week 5: Source localization: distributed models and Bayesian approaches
Week 6: Source localization: hierarchical Bayesian models, entropic framework
Week 7: Source localization: time-frequency based source localization
Week 8: Simultaneous EEG/fMRI: acquisition, analysis, and interpretation
Week 9: Multimodal fusion and neurovascular coupling modelling
Week 10: NIRS : data acquisition, data analysis and inverse problem
Weeks 11-12: oral presentations of the final projects



Feel free to contact me if you need more information

Christophe Grova

***************************
Christophe Grova, PhD
Assistant Professor
Biomedical Engineering Dpt
Neurology and Neurosurgery Dpt

Montreal Neurological Institute
Centre de Recherches en Mathématiques

Biomedical Engineering Department - Room 304
McGill University
3775 University Street, Montreal, Quebec, Canada, H3A 2B4
email : christophe.grova at mcgill.ca<UrlBlockedError.aspx>
tel : (514) 398 2516
fax : (514) 398 7461

web:
http://apps.mni.mcgill.ca/research/gotman/members/christophe.html<UrlBlockedError.aspx>
http://www.bmed.mcgill.ca/<UrlBlockedError.aspx>
***************************