Evolution has led living beings to develop a series of survival strategies. The evolutionary success of mammals has been at the expense of such brain development as to allow them to combine the information received through external stimuli with the ability to anticipate the future through memory accumulated through experience, with the ultimate goal of adapting their behaviour as favourably as possible.  

Whole-brain computational modelling enables studying the effect of perturbations in input signals, in the integration and segregation of information, thus managing to discern between different states of consciousness, as well as between brain health and disease.

These are the main conclusions of a study published by Nature Reviews Neuroscience on 17 June, coordinated by Gustavo Deco , ICREA researcher of the Department of Information and Communication Technologies (DTIC) and director of the Center for Brain and Cognition (CBC) at Pompeu Fabra University, which involved Danish and north American researchers. The article reviews how the brain must regulate the information flow of incoming stimuli to enable cognition and the adaptation of behaviour.

We know that a balance exists between the segregation and integration of stimuli that reach the brain. The topological characteristics of the brain networks observed, mainly network communities and connecting nodes, give support to this segregation and integration of signals, but do not provide information about brain dynamics or how external impulses are processed in time.   

For example, formally, information integration is understood as the phenomenon by which the system acquires more information than it has available based on the sum of its parts. This integration has been linked with consciousness, even though this same state can be achieved non-consciously, and the underlying principles of this fundamental process, on the other hand, are not known.

Until now, one of the strategies used to obtain information on brain function were experiments based on the control and monitoring of the selective input of stimuli to the brain; but these studies have had limited success.

Computational modelling and the perturbations induced on the model

The authors of the study published in Nature Review Neuroscience argue that whole-brain computational modelling, performed by means of the data obtained by neuroimaging, provides totally new knowledge about the segregation and integration of stimuli, and in their work they give a description of the progress achieved by applying mathematical graph theory in their model.

Deco et al.  show that what in recent years has led to the greatest advances in knowledge, architecture and functionality of the brain have been computational modelling methods. Through this strategy it has been possible to assess how the perturbations of input signals affect the dynamics of the brain.

Having formulated the computational model, the neuroscientists introduced perturbations, whereby "we have been able to track brain dynamics and better understand the processes of segregation and integration of information over time", said Deco. "And," he added, "the most impor tant thing is that this method, which allows us to measure the segregation and integration of information, enables us to distinguish between different states of consciousness, of brain health, of neuropsychiatric diseases and disorders".

Gustavo Deco, in addition being recognized by the European Research Council with an advanced grant, is one of the principal investigators of the European project FET Flagship Human Brain Project (HBP) whose main mission is to build an innovative future ICT infrastructure that incorporates neuroscience, medicine and computation in order to understand the functioning of the human brain and its diseases, as well as be able to emulate its functions on a computational level.

Reference work:

Gustavo Deco, Giulio Tononi, Melanie Boly and Morten L. Kringelbach (2015), " Rethinking segregation and integration: contributions of whole-brain modelling",  Nature Review Neuroscience, 16, 430-439, doi:10.1038/nrn3963, 17 June.

Other related research e-news:

Computational neuropsychiatry: an emerging discipline that studies neural networks in the diseased brain