DIAS, Álvaro Machado. Computational models and analyses in neuroscience: systematic review. []. , 10, 2, pp.530-552. ISSN 1808-4281.

Computational models of the human brain may be classified as ‘data treatment models’ (brain maps) and ‘operational-inspired models’ (connectionist models). The literature on both is extensive and varied, rendering difficulties to systematic reviews. This paper aims to contribute to outfit such demand, and to provide an overview of the most exciting projects currently under development. Method: besides traditional revising methods, a data mining program used to acknowledge patterns in complex databases (e.g. genomes) was used. Results: among the 2749 studies, learning and memory were the most recurrent topics, modeled in 0.14 and 0.9 of the studies. Epilepsy, schizophrenia and Parkinson are the most computationally simulated diseases; Alzheimer is the fourth, and more correlated to MRI (r=0.5435) than to MEG 9r=0.2033), PET (r=0.4718) and EEG (r=0.1766). Among the 87 2007-2009 studies on ‘operational-inspired models’, small world architectures is the most prominent concept. The most exciting project under development is the Blue Brain Project, which aims to provide a computational model of the whole brain.

: Computational models; Human brain; Neuroanatomy; Neurophysiology; Systematic review.

        · |     · |     · ( pdf )