Research

Primarily we use a combination of modeling and data analysis to understand neural activity. We are also starting to do some experimental work with human EEG, and other interesting kinds of analyses of complex systems. Some broad directions of work in the lab are as follows.

Spontaneous activity, response variability and brain state dependence of neural activity

The dominant mode of studying neural activity is to provide an external perturbation and measuring the response of the brain. However, our brains are active at all times, even in the absence of external stimuli (for example, waves of actiation have been observed in the retina and visual cortex even before birth in animals; activity during dreaming and sleep etc.). Secondly, even the responses to stimuli are not only dependent on the stimuli, but on the spontaneous background activity of the neurons. We are therefore interested in understanding this spontaneous activity. In particular:

• Characterizing regular patterns of spiking / LFPs in spontaneous brain activity during waking, sleep or anesthesia (ML)
• Modelling to understand the network architectures that could give rise to such spontaneous patterns (mostly using RNN type models)
• Characterizing the variability of neural responses (spiking, LFP, EEG) to repeated stimuli and determining to what extent this variability can be explained by emergent variables such as cortical oscillations, functional connectivity and behavior (ML, GLMs)

Neural effects of technology, cognition outside the lab

We have only begun to do some fun experiments with EEG to understand the effects of technology (over)use on the brain. Going forward, we are interested in developing wearables and techniques to assess neural activity in the ‘wild’ ie outside the lab.

Developing tools and methods for analyzing large scale electrophysiological recordings

A huge amount of high quality electrophysiological data is being made publicly available from several initiatives, such as the Allen Institute, International Brain Laboratory etc. In addition to using these data for the above mentioned problems, these datasets offer many opportunities for developing novel analyses techniques, that we pursue.

Minor pursuits

Embedded systems, wearables

… and more.