Teams at Computational Neuroscience Lab
We are developing a unified model of the basal ganglia that explain a wide variety of motor and cognitive symptoms of Parkinson's disease (PD). We are also currently working to develop quantitative and optimal clinical therapies for PD.
India is a land of a large number of languages. There are ten major scripts that are used to write most of the major languages of India. Bharati is being proposed as a common script for India. The Roman script is used as a common script for many European languages (English, French, German, Italian etc.), which facilitates communication across nations that speak and write those languages. Likewise, a common script for the entire country is hoped to bring down many communication barriers in India.
We particularly focus on modeling the visual motion perception using machine learning tools. Our proposed neural field model of self-organization of brain function is the plausible mechanism for the development of cortical columns, direction selectivity, pattern selectivity and higher order flow selectivity.
What we usually think of as "Memory" in day-to-day usage is actually long-term memory, but there is also important short-term memory, which must be worked through before a long-term memory can be established. This establishment of long-term memory is done through the consolidation of short-term-memory by hippocampus. Our aim is to understand the role of Hippocampus in the formation and consolidation of Human Memory.
Although the electrical activity of a single neuron seems like train of spikes, the activity of a population of neurons at various scales measured in terms of LFP, BOLD, fMRI or EEG signals looks like continuous signals with distinct oscillatory components associated with various frequency bands (delta, theta, gamma, alpha etc). The aim of our team is to develop general trainable networks of nonlinear oscillators that can be applied to model brain dynamics and also artificial engineering applications.
The neuromotive team is dedicated for creating brain- inspired architectures for Automated Driver Assist Systems (ADAS). We developed attentional search network which is inspired by the searching technique of human brain for a target in the real time system. The attentional search network is able to outperform the state-of-the-art object detector like YOLO and Faster R-CNN.
We work together to bring out a well defined simplified model of neurovascular interaction to explain various changes in cerebrovascular functions in many neurological disorders. Develop a network level model to capture neuro-glial-vascular interaction with a simplified computational model. Redefine the neurovascular interaction as a form of efficient information transfer between neural network and vascular network using non linear dynamical systems modelling.
We focus on studying how the external spatial world is mapped on to the brain in the form of spatial cell representations like the place cells, grid cells, border cells etc. We developed a unified network model of spatial cells that demonstrates emergence of grid cells, place cells, border cells, and corner cells. It models navigation in both 2D (rats/mice) and 3D (bats).
Upon the occurrence of Stroke in a person, the location and extent of damage are measured using brain imaging (fMRI etc.) and by assessing functionality using motor assessment scores (FMA etc). The goal of our team is to develop a computational model which can simulate both neuroimaging and behavioral data and prescribe a rehabilitation protocol that is customized and optimal for that patient. The chosen protocol will be administered to the patient in the form of games in a Virtual Reality Environment. The performance of the patient will be tracked (again by fMRI and FMA data) and the therapy will be modified as and when required.
