Abstract

In this talk, I will describe our efforts to use light to understand information processing in the brain. I will start with an established tool of using light for 3D brain imaging via multi-photon microscopy, where within fundamental optical limits, provides a good spatial range for studying single neurons up to interconnected neurons in a circuit. Within this spatial range, I will then talk about using patterned light to photo-induce synaptic inputs and analyze how the spatio-temporal organization of these inputs cause the neuron to fire an output. We use a dynamically programmable hologram to produce 3D light patterns that can induce targeted and highly localized synaptic inputs along the dendritic tree of a neuron. Such technique can also be used for light-based recording of responses from multiple neurons – leading to an all-optical method for stimulating and recording neuronal activity. To target neurons deep within the brain tissue, the hologram can be added with an adaptive phase-pattern to correct for optical aberrations caused by the brain tissue. Towards the end, I will discuss preliminary results on using non-linear light-tissue interactions to dynamically prune the neuronʼs dendritic tree. The neuronʼs output and overall function in a neuronal circuit are said to be dependent on the spatial extent of its dendritic tree. Hence, laser dendrotomy allows us to study morphology-dependent neuronal function. These experiments are among many light-neuron interactions that facilitate our understanding of how the brain works.

© 2014 Japan Society of Applied Physics, Optical Society of America