Dual-channel Sub-millisecond Resolution Neural Imaging System

Physical Sciences, Inc, Andover MA

Investigators

Abstract

Project Summary/Abstract The investigation of the complex neural dynamics and the cognitive functions of the brain requires non- invasive recording tools with high spatio-temporal resolution. Sub-cellular resolution fluorescence imaging/microscopy, based on either voltage indicators or calcium-sensitive fluorescent proteins, enables parallel recording of the activity of spatially distributed neuron populations. However, the speed of current fluorescence imaging techniques is largely limited to a few hundred frames per second (excluding the use of some extremely sophisticated scientific cameras of substantially high noise and prohibitively high cost). This speed limit originates from not only instrumentation/technology constrains (such as the slow scan in scanning microscopes or the low frame rate of the cameras used in wide-field microscopes), but also the fundamental photon stochastic noise of the low level fluorescent signals. The milliseconds or slower temporal resolution substantially precludes measuring the precise timing of the generation and propagation of neuron spikes or spike trains, which is the key component of neural signaling. Physical Sciences Inc. (PSI) and Massachusetts Institute of Technology (MIT) propose to develop and demonstrate a unique fluorescence imaging system that will enable parallel recording of multiple neuron populations with sub-cellular spatial resolution and sub-millisecond temporal resolution. This novel neural recording system will have two fluorescence detection channels: one to record the slow dynamics of all the neurons in a given field of view with a sub-cellular spatial resolution, and a second one with sub- millisecond temporal resolution to study fast spiking dynamics of neurons, which can be arbitrarily selected under the guidance of the first channel. This high-speed parallel neural recording function is enabled by an innovative imaging concept that uses a high-sensitivity single-point detector to collect fluorescence signals generated at multiple locations (or neuron populations), while the spatial information is encoded by modulating the excitation light at each location at a unique temporal frequency. In Phase I, we will develop and optimize a bench-top microscope at PSI facility, and then move it to MIT for testing using cultured neurons. The small form factors of all the components will make it possible to develop a device that is compact enough to be head-mounted on small animals (such as rats or mice). Development and evaluation of head-mounted devices will occur during a subsequent Phase II program. This R&D project will lead to a low-costs solution for non-invasive recording of large neuron populations with high spatio- temporal resolution, providing an important enabling tool for cutting-edge brain studies and research.