Abstract
Fluorescence microscopy and genetically encoded calcium indicators help understand brain function by recording large-scale in vivo videos in assorted animal models. Extracting the fluorescent transients that represent active periods of individual neurons is a key step when analyzing imaging videos. Non-specific calcium sources and background adjacent to segmented neurons contaminate the neurons’ temporal traces with false transients. We developed and characterized a novel method, temporal unmixing of calcium traces (TUnCaT), to quickly and accurately unmix the calcium signals of neighboring neurons and background. Our algorithm used background subtraction to remove the false transients caused by background fluctuations, and then applied targeted non-negative matrix factorization to remove the false transients caused by neighboring calcium sources. TUnCaT was more accurate than existing algorithms when processing multiple experimental and simulated datasets. TUnCaT’s speed was faster than or comparable to existing algorithms.
Highlights
The brain has many neurons that coordinate their activity to support complex dynamics and behaviors
We compared the performance of temporal unmixing of calcium traces (TUnCaT) with the performance of other unmixing algorithms using three datasets: experimental two-photon videos from Allen Brain Observatory (ABO), simulated two-photon videos using Neural Anatomy and Optical Microscopy (NAOMi), and experimental one-photon videos
We decontaminated the false transients arising from neighboring neurons, axons, and dendrites by using nonnegative matrix factorization (NMF) unmixing
Summary
The brain has many neurons that coordinate their activity to support complex dynamics and behaviors. We varied multiple parameters in simulation: recording frame rate, video length, number of neurons, laser power, electronics base noise variance, and GCaMP type. Our default parameter set specified the frame rate as 30 Hz, the simulated video length as 120 s, the number of neurons in the volume as 200, the laser power as 100 mW, the electronics base noise variance as 2.7, and the sensor as GCaMP6f. Some simulation parameters for GCaMP7 series sensors were different to accommodate the sensors’ slower kinetics For these sensors, we set the frequency to 3 Hz, the simulated video length to 1100 s, the spiking rate to 0.01/s, and the initial transient period to 100 s.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
