Optical segmentation-based compressed readout of neuronal voltage dynamics

Abstract

AbstractFunctional imaging of biological dynamics generally begins with acquiring time-series images, followed by quantifying spatially averaged intensity traces for the regions of interest (ROIs). The conventional pipeline discards a substantial portion of the acquired data when quantifying intensity traces, indicative of inefficient data acquisition. Here we propose a conceptually novel acquisition pipeline that assigns each ROI to a single pixel in the detector, enabling optimally compressed acquisition of the intensity traces. As a proof-of-principle, we implemented a detection module composed of a pair of spatial light modulators and a microlens array, which segments the original image into multiple subimages by introducing distinct angular shifts to each ROI. Each subimage exclusively encodes the signal for the corresponding ROI, facilitating the compressed readout of its intensity trace using a single pixel. This spatial compression allowed for maximizing the temporal information without compromising the spatial information on ROIs. Harnessing our novel approach, we demonstrate the recording of circuit-scale neuronal voltage dynamics at over 5 kHz sampling rate, revealing the individual action potential waveforms within subcellular structures, as well as their submillisecond-scale temporal delays.