M. Zanin and D. Papo
Chaos, 27:047403 (2017).
In biological systems such as the brain, interactions can be fast and short-lived. However, robust causal relationships are usually quantified over time-windows much larger than functionally meaningful time scales. We propose a method to overcome this drawback and quantify causal dynamical connectivity in which every possible time window is considered and non-overlapping ones in which the causality is strongest eventually selected. Our results show that transient but not classical time-averaged causality estimations can discriminate between the electroencephalographic activity of a small sample of alcoholic subjects and a matched healthy control group. Differences between groups appear to be a local rather than a global network property. The implications of these results for the modelling of brain functioning and pathology are briefly discussed.
[Read more in Chaos]
M. Zanin and D. Papo
Entropy, 16:5655-5667 (2014).
Motifs are small recurring circuits of interactions which constitute the backbone of networked systems. Characterizing motif dynamics is therefore key to understanding the functioning of such systems. Here we propose a method to deﬁne and quantify the temporal variability and time scales of electroencephalogram (EEG) motifs of resting brain activity. Given a triplet of EEG sensors, links between them are calculated by means of linear correlation; each pattern of links (i.e., each motif) is then associated to a symbol, and its appearance frequency is analyzed by means of Shannon entropy. Our results show that each motif becomes observable with different coupling thresholds and evolves at its own time scale, with fronto-temporal sensors emerging at high thresholds and changing at fast time scales, and parietal ones at low thresholds and changing at slower rates. Finally, while motif dynamics differed across individuals, for each subject, it showed robustness across experimental conditions, indicating that it could represent an individual dynamical signature.
[Read more in Entropy]
D. Papo, M. Zanin, J.A. Pineda-Pardo, S. Boccaletti, and J.M. Buldú
Philosophical Transactions of the Royal Society B 369:20130525 (2014).
Many physical and biological systems can be studied using complex network theory, a new statistical physics understanding of graph theory. The recent application of complex network theory to the study of functional brain networks generated great enthusiasm as it allows addressing hitherto non-standard issues in the field, such as efficiency of brain functioning or vulnerability to damage. However, in spite of its high degree of generality, the theory was originally designed to describe systems profoundly different from the brain. We discuss some important caveats in the wholesale application of existing tools and concepts to a field they were not originally designed to describe. At the same time, we argue that complex network theory has not yet been taken full advantage of, as many of its important aspects are yet to make their appearance in the neuroscience literature. Finally, we propose that, rather than simply borrowing from an existing theory, functional neural networks can inspire a fundamental reformulation of complex network theory, to account for its exquisitely complex functioning mode.
[Read more in Philosophical Transactions] [Read more in ArXiv] [Read interview in Phil. Trans Blog] [Listen to podcast in Nature]
D. Papo, M. Zanin and J.M. Buldú
Frontiers in Human Neuroscience, 8:107 (2014).
Both at rest and during the executions of cognitive tasks, the brain continuously creates and reshapes complex patterns of correlated dynamics. Thus, brain functional activity is naturally described in terms of networks, i.e. sets of nodes, representing distinct subsystems, and links connecting node pairs, representing relationships between them. Recently, brain function has started being investigated using a statistical physics understanding of graph theory, an old branch of pure mathematics (Newman, 2010). Within this framework, networks properties are independent of the identity of their nodes, as they emerge in a non-trivial way from their interactions. Observed topologies are instances of a network ensemble, falling into one of few universality classes and are therefore inherently statistical in nature. Functional network reconstruction comprises various steps: first, nodes are identified; then, links are established according to a certain metric. This gives rise to a clique with an all-to-all connectivity. Deciding which links are significant is done by choosing which values of these metrics should be taken into account. Finally, network properties are computed and used to characterize the network. Each of these steps contains an element of arbitrariness, as graph theory allows characterizing systems once a network is reconstructed, but is neutral as to what should be treated as a system and to how to isolate its constituent parts. Here we discuss some aspects related to the way nodes, links and networks in general are defined in system-level studies using noninvasive techniques, which may be critical when interpreting the results of functional brain network analyses.
[Read more in Frontiers in Human Neuroscience]
Marco Buiatti, David Papo, Pierre-Marie Baudonnière, and Carl Van Vreeswijk
Neuroscience, 146:1400-1412 (2007).
We used the electroencephalogram (EEG) to investigate whether positive and negative performance feedbacks exert different long-lasting modulations of electrical activity in a reasoning task. Nine college students serially tested hypotheses concerning a hidden rule by judging its presence or absence in triplets of digits, and revised them on the basis of an exogenous performance feedback. The scaling properties of the transition period between feedback and triplet presentation were investigated with detrended fluctuation analysis (DFA). DFA showed temporal scale-free dynamics of EEG activity in both feedback conditions for time scales larger than 150 ms. Furthermore, DFA revealed that negative feedback elicits significantly higher scaling exponents than positive feedback. This effect covers a wide network comprising parietooccipital and left frontal regions. We thus showed that specific task demands can modify the temporal scale-free dynamics of the ongoing brain activity. Putative neural correlates of these long-lasting feedback-specific modulations are proposed.
[Read more in Pubmed] [Read more in Neuroscience]
Papo, D., Douiri, A., Bouchet, F., Bourzeix, J.-C., Caverni, J.-P., & Baudonnière, P.-M.
Cerebral Cortex, 17:1314-1322 (2007).
The neural correlates of the response to performance feedback have been the object of numerous neuroimaging studies. However, the precise timing and functional meaning of the resulting activations are poorly understood. We studied the electroencephalographic response time locked to positive and negative performance feedback in a hypothesis testing paradigm. The signal was convoluted with a family of complex wavelets. Intracranial sources of activity at various narrow-band frequencies were estimated in the 100- to 400-ms time window following feedback onset. Positive and negative feedback were associated to 1) early parahippocampo-cingular sources of alpha oscillations, more posteriorly located and long lasting for negative feedback and to 2) late partially overlapping neural circuits comprising regions in prefrontal, cingular, and temporal cortices but operating at feedback-specific latencies and frequencies. The results were interpreted in the light of neurophysiological models of feedback and were used to discuss methodological issues in the study of high-level cognitive functions, including reasoning and decision making.
[Read more in Cerebral Cortex]
Papo D, Baudonnière PM, Hugueville L, Caverni JP
Journal of Cognitive Neuroscience, 15:508-522 (2003).
We used event-related potentials (ERPs) to probe the effects of feedback in a hypothesis testing (HT) paradigm. Thirteen college students serially tested hypotheses concerning a hidden rule by judging its presence or absence in triplets of digits and revised them on the basis of an exogenous performance feedback. ERPs time-locked to performance feedback were then examined. The results showed differences between responses to positive and negative feedback at all cortical sites. Negative feedback, indicating incorrect performance, was associated to a negative deflection preceding a P300-like wave. Spatiotemporal principal component analysis (PCA) showed the interplay between early frontal components and later central and posterior ones. Lateralization of activity was selectively detectable at frontal sites, with a left frontal dominance for both positive and negative feedback. These results are discussed in terms of a proposed computational model of trial-to-trial feedback in HT in which the cognitive and emotive aspects of feedback are explicitly linked to putative mediating brain mechanisms. The properties of different feedback types and feedback-related deficits in depression are also discussed.
[Read more in Journal of Cognitive Neuroscience]