He works with James Haxby at the CIMeC, and he is currently supported by the ATTEND
project. His research focuses on the cognitive and neural representation of actions and attention. In this line of research data is acquired with functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG) and analyzed using multi-variate pattern analysis (MVPA). Besides emperical work, He is also involved in the development of the PyMVPA (Python) and CoSMoMVPA (Matlab) toolboxes.
Nikolaas Oosterhof
PRNI 2016 - TUTORIAL 3
CoSMoMVPA: A multi-modal/variate pattern analysis toolbox in Matlab
 |  |  |  |
|---|---|---|---|
 |  |  |  |
 |  |  |  |
After an introductory presentation, it starts with basic operations of reading, writing, selecting, and aggregating dataset structures. This is followed by MVPA correlation and classficiation analysis of fMRI data in a region of interest. Subsequently, this is extended to exploratory searchlight analysis, representational similarity analysis, and MEEG analysis in the space and time dimensions. Finally approaches to multiple comparison are discussed.
Note: although MEEG analysis is covered only on day 2, basic concepts and functionality for MEEG analysis is discussed on day 1. Also for those who are mainly interested in MEEG analysis (and less so in fMRI) it is still recommended to attend both days.
MONDAY, JUNE 20th
08:30 :: Registration
09:00 :: Introduction
11:00 :: Get your computer ready to run CoSMoMVPA; practical exercises: fMRI basic dataset operations
10:30 :: Coffee Break
12:30 :: Lunch Break
13:30 :: Correlation Analysis
15:00 :: Coffee Break
15:30 :: Classification Analysis
TUESDAY, JUNE 21st
09:00 :: Neighborhoods, Searchlight
11:00 :: MEEG basics, classification, searchlight
10:30 :: Coffee Break
12:30 :: Lunch Break
13:30 :: MEEG time generalization; fMRI RSA
15:00 :: Coffee Break
15:30 :: Multiple comparison correction; concluding remarks
Setting up your system
It is recommended, prior to the tutorial, to:
-
read the CoSMoMVPA manuscript: http://dx.doi.org/10.1101/047118
-
have the most recent CoSMoMVPA code http://www.cosmomvpa.org/download.html
-
have a recent version of the tutorial data
-
have set paths properly in .cosmomvpa.cfg (described :ref:`here <set_cosmovmpa_cfg>`)
-
have tested your local setup that you can load and save data from and to the paths in .cosmomvpa.cfg.
Please visit the repository of tutorial materials at http://bit.ly/prni2016cosmomvpa
Goals of the tutorial
Learn how to use CoSMoMVPA to perform these analyses:
-
Understand the dataset structure to represent both the data itself (e.g. raw measurements or summary statistics) and its attributes (e.g. labels of conditions (targets), data acquisition run (chunks).
-
See how parts of the data can be selected using slicing and splitting, and combined using stacking.
-
Introduce measures that compute summaries of the data (such as correlation differences, classification accuracies, similarity to an a prior defined representational simillarity matrix) that can be applied to both a single ROI or in a searchlight.
Objectives not covered by this tutorial
-
Preprocessing of fMRI / MEEG data
-
Learning to use Matlab / Octave
-
Dataset types other than volumetric fMRI data and MEEG time-locked data
(Not covered: surface-based fMRI, source-space MEEG) -
How to become a CoSMoMVPA developer
For additional information you may contact tutorial3@prni.org