In the course of the tutorial, Alexandre Gramfort gave an introduction to Python and NumPy, as well as an intorduction to machine learning with scikit-learn + hands-on session on MEG data. Finally, he introduced MNE/mne-python - a tool for analyzing MEG data and, and guided the participants through a hands-on session on MEG sensor vs. source space analysis.
You are here
The purpose of this visit is to foster the collaboration between Anna Gardumi (Maastricht University) and Emanuele Olivetti (NILab) on the analysis of a 7T fMRI dataset replicating the study of (Formisano et al., Science, 2008).
Andrea Bertana, Master in Cognitive Neuroscience Program UNITN joins NILab for his internship/thesis
"Handed Brain", installed at the Natural Science Museum (Trento), presented the research on brain mapping, carried out at NI lab, where all areas related to hands and their movements were mapped. The stand hosted videos of brain tracts animation, videos of movie clips related to hand movements from Skyfall (synchronized to the brain), snapshot of all segmented tracts, in three projections each.
COPEM was held in Prague, Czech Republic on September 27, 2013 as a workshop of the ECML-PKDD conference. The workshop was organized by Ioannis Katakis, Daniel Hernandez-Lobato, Gonzalo Martinez-Munoz and Ioannis Partalas and the main goal was to discuss about ensemble strategies that solve difficult machine learning problems. The paper "Clustering ensemble on reduced search spaces" was presented by Sandro Vega-Pons as an oral presentation during the afternoon session.
The authors propose a seamless integration of the local and remote environments for neuroscientific data storage, management and analysis. he main advantage is to avoid metadata annotation coming from external frameworks in data - management. An additional benefit is to preserve the flexibility of pipeline deployment in a user friendly interface to HPC infrastructure.
The challenge of longitudinal studies analysis is to disentangle plain session effect from the plasticity one on BOLD response as the former is of interest to see how the patterns of activation differ through time. In the paper, the authors propose the method to reduce the bias of the coregistration error related to the alignment of two sessions. They show how combining functional hyperalignment and multivariate pattern analysis, it is possible to summarize a single brain map that accounts for plasticity effect without the session effect.
Talk: Neurodegenerative Assessment of Corticospinal Tract in ALS with Deterministic Tractography
joint work of Paolo Avesani (FBK and CIMeC), Francesca Maule (CIMeC), Nivedita Agarwal (APSS and CIMeC)
presented at AINR Conference at L'Aquila, Italy, 18-21.09.13
Dr. ir. Robert P.W. Duin from the Pattern Recognition Lab of Delft University of Technology (The Netherlands) gave a seminar entitled "Non-Euclidean Problems in Pattern Recognition". Robert Duin started investigating alternative object representations for classification some years ago, and thereby became interested in dissimilarity-based pattern recognition. He is actively working on this topic and its application in real-world problems.
The INCF Latin American Summer School on Neuroinformatics was held in Havana, Cuba from September 16th to 21th, 2013. Several talks and hands-on sessions were given by leading researchers on different topics related to Neural modeling, Data Analysis and Databases.