A deep learning framework for inference of single-trial neural population activity from calcium imaging with sub-frame temporal resolution

Abstract

In many brain areas, neural populations act as a coordinated network whose state is tied to behavior on a moment-by-moment basis and millisecond timescale. Two-photon (2p) calcium imaging is a powerful tool to probe network-scale computation, as it can measure the activity of many individual neurons, monitor multiple cortical layers simultaneously, and sample from identified cell types. However, estimating network state and dynamics from 2p measurements has proven challenging because of noise, inherent nonlinearities, and limitations on temporal resolution. Here we describe RADICaL, a deep learning method to overcome these limitations at the population level. RADICaL extends methods that exploit dynamics in spiking activity for application to deconvolved calcium signals, whose statistics and temporal dynamics are quite distinct from electrophysiologically-recorded spikes. It incorporates a novel network training strategy that capitalizes on the timing of 2p sampling to recover network dynamics with high temporal precision. In synthetic tests, RADICaL infers network state more accurately than previous methods, particularly for high-frequency components. In real 2p recordings from sensorimotor areas in mice performing a "water grab" task, RADICaL infers network state with close correspondence to single-trial variations in behavior, and maintains high-quality inference even when neuronal populations are substantially reduced.