Efficient and accurate extraction of in vivo calcium signals from microendoscopic video data

Abstract

In vivo calcium imaging through microendoscopic lenses enables imaging of previously inaccessible neuronal populations deep within the brains of freely moving animals. However, it is computationally challenging to extract single-neuronal activity from microendoscopic data, because of the very large background fluctuations and high spatial overlaps intrinsic to this recording modality. Here, we describe a new constrained matrix factorization approach to accurately separate the background and then demix and denoise the neuronal signals of interest. We compared the proposed method against previous independent components analysis and constrained nonnegative matrix factorization approaches. On both simulated and experimental data recorded from mice, our method substantially improved the quality of extracted cellular signals and detected more well-isolated neural signals, especially in noisy data regimes. These advances can in turn significantly enhance the statistical power of downstream analyses, and ultimately improve scientific conclusions derived from microendoscopic data.

Encoding of danger by parabrachial CGRP neurons

“…To ensure that these effects were not due movement artifacts, we analyzed fluorescence in regions of interest lacking neurons and observed no changes before or during feeding (data not shown). We also analyzed calcium recordings using constrained nonnegative matrix factorization (CNMF) 21 . Individual traces extracted from CNMF analysis show that calcium transients are not synchronized, but all neurons are similarly inhibited upon presentation of food and before bites (Fig.…”

“…Raw traces were converted to ΔF/F (F-F baseline average /F baseline average), where F was the fluorescence at any given time-point and F baseline average was the average fluorescence during a designated baseline period. Additional analyses were conducted using CNMF for microendoscopic data 21 .…”

MIN1PIPE: A Miniscope 1-Photon-Based Calcium Imaging Signal Extraction Pipeline

“…In single-photon miniscope imaging, existing automatic signal extraction methods mainly include principal-component analysis/independent component analysis (PCA/ICA) (Mukamel et al, 2009) and constrained nonnegative matrix factorization-extended (CNMF-E) (Zhou et al, 2018). PCA/ICA (Mukamel et al, 2009) was the first attempt to automatize the signal extraction process from miniscope imaging data using PCAfollowed by ICAto extract signals from background and involving manual annotations of regions of interest (ROIs).…”

“…PCA/ICA (Mukamel et al, 2009) was the first attempt to automatize the signal extraction process from miniscope imaging data using PCAfollowed by ICAto extract signals from background and involving manual annotations of regions of interest (ROIs). A latest method, CNMF-E, adapts the constraint matrix factorization framework (Zhou et al, 2018). Specifically, because the data from single-photon imaging are dominated by a noisy, uneven, and fluctuating background, CNMF-E uses a sophisticated background fitting model to better characterize the non-neuronal dynamics and is, thereby, better at identifying neuron-like ROIs (Zhou et al, 2018).…”

“…A latest method, CNMF-E, adapts the constraint matrix factorization framework (Zhou et al, 2018). Specifically, because the data from single-photon imaging are dominated by a noisy, uneven, and fluctuating background, CNMF-E uses a sophisticated background fitting model to better characterize the non-neuronal dynamics and is, thereby, better at identifying neuron-like ROIs (Zhou et al, 2018). However, both PCA/ICA and CNMF-E have false-negative and false-positive issues in ROI identification, as we will illustrate in the Results section.…”