Are EEG functional networks really describing the brain? A comparison with other information-processing complex systems
Authors:
Sofia Gil-Rodrigo,
Raúl López-Martín,
Görsev Yener,
Jan R. Wiersema,
Bahar Güntekin,
Massimiliano Zanin
Abstract:
Functional networks representing human brain dynamics have become a standard tool in neuroscience, providing an accessible way of depicting the computation performed by the brain in healthy and pathological conditions. Yet, these networks share multiple characteristics with those representing other natural and man-made complex systems, leading to the question of whether they are actually capturing…
▽ More
Functional networks representing human brain dynamics have become a standard tool in neuroscience, providing an accessible way of depicting the computation performed by the brain in healthy and pathological conditions. Yet, these networks share multiple characteristics with those representing other natural and man-made complex systems, leading to the question of whether they are actually capturing the uniqueness of the human brain. By resorting to a large set of data representing multiple financial, technological, social, and natural complex systems, and by relying on Deep Learning classification models, we show how they are highly similar. We specifically reach the conclusion that, under some general reconstruction methodological choices, it is as difficult to understand whether a network represents a human brain or a financial market, as to diagnose a major pathology. This suggests that functional networks are describing information processing mechanisms that are common across complex systems; but that are not currently defining the uniqueness of the human mind. We discuss the consequence of these findings for neuroscience and complexity science in general, and suggest future avenues for exploring this interesting topic.
△ Less
Submitted 23 May, 2025; v1 submitted 3 November, 2024;
originally announced November 2024.
