From activity fluctuations to points: compressing fMRI recordings into a discrete point-process allows insights on the dynamical regime of the brain.

ENZO TAGLIAZUCCHI

Dresden

Simposio; Physical origins of correlated extreme events (ExEv14); 2014

Max Planck for the Physics of Complex Systems

Brain science has enormously benefited from the use of non-invasive methods to quantify neural activity over time. One of the most prominent examples is functional Magnetic Resonance Imaging (fMRI), which yields continuous signals indirectly encoding neural activity levels over time. In this talk I will present evidence favoring an interpretation and origin of those signals which largely departs from the usual paradigm. I will show that activity fluctuations can be decomposed and compressed into a set of discrete events (a spatio-temporal point-process) and that this approach allows novel insights on the dynamical regime of the human brain. These include the presence of large-scale avalanches and the introduction of control and order parameters, thus establishing a link between the problem of large-scale brain complexity and the methodology of statistical mechanics. I will conclude with a preliminary discussion on the relationship between the dynamical regime of the brain and the level of conscious awareness, using fMRI recordings in the descent from wakefulness to deep sleep as a paradigmatic example of vanishing awareness.