Computational Modeling of Normal and Abnormal Cortical Processing

Computational neuroscience is one of the most outstanding and challenging fields of scientific endeavor, with potential benefits not only for highly enhanced theoretical understanding, but also for important practical applications such as in mental-health studies. The Computational Modeling of Normal and Abnormal Cortical Processing project proposes the establishment of a computational neuroscience research group at NYU Abu Dhabi, closely tied to experimental groups at NYU and around the world, with the goal of contributing to the understanding of how the brain operates in both healthy and abnormal states.  The project will be divided into two parts.

Project Part I

Project I will focus on the functioning of the early visual pathway, in particular, the primary visual cortex. A large-scale computational model pioneered by members of the group will be further developed and applied to modeling both healthy and dysfunctional visual processing. For the latter, the aim is to understand the effects of the NMDA receptor hypofunction on the early stages of visual perception, with an eye on this hypofunction being one of the possible neuronal disorders underlying schizophrenia. The emergence of a number of salient phenomena observed in schizophrenic patients, including reduced contrast sensitivity and reduced presence of higher harmonics in the EEG signal, will be modeled using the hypothesis of reduced NMDA receptor efficiency and its potential functional consequences such as disinhibition of excitatory neurons and network sparsification. The compatibility of these paradigm phenomena with the NMDA hypofunction hypothesis will be examined.

Project Part II

Project II will explore the theoretical foundation of the nonlinear dynamical analysis and functional connectivity of brain processing, and, in particular, its relationship to noninvasive measurements such as the EEG. Particularly intriguing questions concern the reduction of the chaotic dynamics and dimensionality and the loss of small world attributes in the EEG signals measured in Alzheimer’s disease patients. A theoretical framework of an effective theory of EEG as well as information how formalism will be established through coarse-graining of a hierarchy of neuronal networks over many areas of the brain. The origin of the possible chaotic signature in EEG recordings and the loss of small-world attributes of graph analysis of EEG signals in Alzheimer’s patients will be examined in this framework after incorporating acetylcholine neurotransmitter. The impact of the acetylcholine deficiency and loss of connectivity of local neuronal networks on the complexity of EEG signals will be examined.

A central feature of the proposed research is to use computational modeling of cortical dynamics to rationalize experimental observations in subjects that cannot be easily studied by invasive methods under normal conditions. The proposed research aims to push the computational approach in neuroscience to a new level so that it will become a truly powerful methodology capable of unifying invasive and noninvasive observations in healthy and abnormal brain states.


Principal Investigator David Cai
Professor of Mathematics and Neural Science, Courant NYU
Co-Principal Investigator Aaditya Rangan
Assistant Professor of Applied Mathematics, Courant NYU

Co-Principal Investigator Gregor Kovacic
Associate Professor, Rensselaer Polytechnic Institute
Co-Principal Investigator Douglas Zhou
Professor, Shanghai Jiao Tong University

Co-Principal Investigator Dan Hu
Professor, Shanghai Jiao Tong University
Postdoctoral Associate Zhiqin Xu
NYU Abu Dhabi