Data, Data mining and Modelling of Brain Activity for Understanding Attentional Mechanisms

Four lectures, each 20 minutes long + 5 minutes question/transition, followed up at the end by a 20 minute round table/general discussion

1. Andreas A. Ioannides and Vahe Poghosyan "When and where spatial and non-spatial attention influence brain processing"
2. Nikolaos A. Laskaris "Using Graphs to Mine Brain Dynamics: current practice and future trends"
3. Michalis Zervakis "Dynamic Synchronization of Independent Components in Evoked Response Potentials"
4. John G. Taylor "Modelling attention and its breakdowns"

• What has been explained, what remains to be done?
• What improvements or new tools we need?

Brief summary: Attention always seemed important, but it was not always considered a "proper" scientific topic because it was difficult to correlate its effects with hard scientific data. In the last couple of decades however attention has gained prominence thanks to both the availability of hard data from animal recordings, improvements in human electrophysiology (EEG and MEG) and the introduction of new neuroimaging methods (PET and fMRI). The first presentation of the session will review the evidence for when and where in the brain attentional modulations are seen, emphasizing recent results demonstrating very early attentional effects (Ioannides and Poghosyan). The second presentation will focus on graph-based methods that promise to make easier the contrast of brain activations in attended and unattended stimuli and thus empirically model possible attentional mechanisms (Laskaris et al). The third presentation will consider attentional effects as modulations of synchronization between different cell assemblies and use ICA to study them in normal subjects and AD patients (Zervakis). The last presentation will describe how attention and its breakdown can be described by the CODAM model of attention.