4.2 Neuroscience 68
Untangling the interactions between brain waves of different frequencies Collaborators: Raul Vicente 1,2 , Juhan Aru 3 , Jaan Aru 1,2 , Michael Wibral 4 , Viola Priessemann 5 , Gordon Pipa 6 , Luiz Lana 1 , Wolf Singer 1,2 1 Dept. Neurophysiology, Max-Planck <strong>Institute</strong> <strong>for</strong> Brain Research, <strong>Frankfurt</strong>, 2 <strong>Frankfurt</strong> <strong>Institute</strong> <strong>for</strong> <strong>Advanced</strong> <strong>Studies</strong>, 3 Center <strong>for</strong> Research and Interdisciplinarity, Faculty of Medicine, Paris Descartes University, Paris, France 4 MEG Unit, Brain Imaging Center, Goethe University, <strong>Frankfurt</strong>, 5 Neural Systems and Coding, Max- Planck <strong>Institute</strong> <strong>for</strong> Brain Research, <strong>Frankfurt</strong>, 6 <strong>Institute</strong> of Cognitive Science, University of Osnabrück One of the central questions in neuroscience is how neural activity is organized across different spatial and temporal scales. Brain waves of slow frequencies extend over large regions of cortex as opposite to faster oscillations which are coherent over smaller patches. Thus, the interaction between oscillations at different frequencies has been proposed to facilitate flexible coordination of neural activity simultaneously in time and space. Although various experiments have revealed amplitude-toamplitude and phase-to-phase coupling, the most common and most celebrated result is that the phase of the lower frequency component modulates the amplitude of the higher frequency component. Over the recent 5 years the amount of experimental works finding such phase-amplitude and associating it to behavioral and pathological states has been tremendous. In this research project, we evaluate the mathematical foundations of cross frequency analysis and provide evidence that current methods might overestimate physiological cross-frequency coupling (CFC) actually evident in the signals of LFP, ECoG, EEG and MEG. In particular, we have pointed out three conceptual problems in assessing the components and their correlations of a time series. Although we focus on phase-amplitude coupling, most of our argument is relevant <strong>for</strong> any type of coupling. Our results are based on mathematical considerations, numerical simulations, and analysis of neurophysiologic data. The problems are associated with: 1) the ambiguous identification of physiological oscillatory components via filtering and Fourier decomposition, 2) the origin of the correlation between oscillatory components (effects of non-linearity and non-stationarity), and 3) the statistical evaluation of the measures of CFC. Figure 1: A) Recordings at early sensory (blue), and higher (red) cortical areas. B) While inputs to the early cortical areas can be easily controlled or monitored, inputs to higher areas are difficult to evaluate. C) The non-stationarities associated to input arrival can generate patterns that can be confounded with crossfrequency interactions. Un<strong>for</strong>tunately, without control of inputs the ambiguity between oscillatory interactions and input-related patterns cannot be easily disentangled. Thus, in this research project we study which conditions need to be tested to solve some of the ambiguities with the hope that knowing such conditions will be helpful <strong>for</strong> the advancing of crossfrequency analysis of brain activity. [1] J. Aru, J. Aru, M. Wibral, V. Priesemann, G. Pipa, L. Lana, W. Singer, R. Vicente, Untangling cross-frequency analysis in neuroscience, submitted. 69