Computational mapping of brain networks

Magnetic resonance imaging (MRI) has developed into an indispensible diagnostic tool in medicine. MRI has also demonstrated immense potential for researchers who are making progress in every aspect of this modality expanding its applications into uncharted territories. Computational techniques have made major contributions to MRI enabling detection of minute signals from human brain. Functional MRI (fMRI) offers imaging of the mind as well as the brain in the same session. Complex computational tools are used to visualize brain networks that offer a new powerful tool to study the brain and its disorders. Functional connectivity (fc) maps using resting state fMRI (rsfMRI) is computed by detecting temporal synchronicity of neuronal activation patterns of anatomically separated brain regions. But, a great deal of technological advancement, both in hardware and software, had to be made to make computation of brain networks possible. The critical technologies that made computational modeling of functional brain networks possible were high quality gradients for implementation of distortion free fMRI, faster pulse sequences and radio frequency (RF) coils to capture the fluctuation frequency of neuronal activity, and complex post processing computation of brain networks. rsfMRI is capable of detecting brain function that mediate high cognitive processes in normal brain. We aim to ultimately detect the disruption of this mediation in psychiatric patients. We have already obtained functional connectivity in normal subjects using fMRI data during resting state. We did this as a function of spatial resolution to explore the required computational sources and susceptibility effects on the sensitivity of fMRI to anatomic specialization. We provide a conceptual summary of the role of computational techniques in fMRI data analysis. In exploring this question, ultimately MRI's capability in accessing information at the neuronal level comes to surface. We use latest computational tools for analysis of data from human brain and offer a vision for future developments that could revolutionize the use of computational techniques in making neuropsychiatry a quantitative practice.