摘要: Although it is already four decades since the very first magnetoencephalographic (MEG) measurement, this non-invasive neuroimaging modality still advancing at a considerable speed thanks to continuous developments in instrumentation and analysis methods but also experimental designs interpretation of data (see e.g. Hari et al., 2010). The targets early MEG studies, such as evoked responses sensory systems, are strong field new ways looking have emerged. Functional connectivity particularly appropriate object study unique combination high temporal resolution decent spatial makes well-suited for tapping on, e.g., oscillatory brain activity which considered mediate connectivity. can be used investigate intrinsic mechanisms memory attention even characterize consciousness. In talk, I will illustrate important factors design that pertain studies on cognitive neuroscience (Salmelin Parkkonen, show examples use (Nishitani Hari, 2002), spontaneous oscillations (Caetano 2007) temporally structured, or tagged, stimuli (Parkkonen 2008). Recently, real-time on-line feedback subject become feasible they processes novel interesting manner (Sudre 2011). For example. visual modulates distribution posterior alpha oscillations, exploited brain– computer interface (Bahramisharif We studied using frequency-tagged feedback; present our preliminary results. Applying machine learning algorithms may allow decoding stimulus class from activity. Such an approach has been fruitful fMRI without any information about Decoding not yet widely applied one could speculate should excel because wealth data. results single-trial low-level features subjective awareness.