Blind Sparse Deconvolution For Spike Inference From Fluorescence: Larval Zebrafish Fluorescence Recordings

This repository contains the two fluorescence imaging datasets used in the publication "Blind Sparse Deconvolution for Spike Inference from Fluorescence Recordings". Each dataset is a T X N DF/F matrix, where T is the number of fluorescence measurement and N is the number of neurons. They were obtained from light-sheet fluorescence microscopy. Both recordings correspond to spontaneous activity. The first file "DFF_20Hz_GCaMP3.mat" was obtained by a 2D recording acquired at 20 frame/second for 20 minutes of a 5dpf-old zebrafish larva expressing the genetically encoded indicator GCaMP3 (elavl3:GCaMP3). The images were parsed into 8082 neural traces. The second file "DFF_1Hz_GCaMP5.mat" was obtained by a 3D, whole-brain recording acquired at 20 frame/second and 20 stacks, (1 measurement/voxel/s) of a 5dpf-old zebrafish larva expressing the genetically encoded indicator GCaMP5. After segmentation, 255463 fluorescence traces encompassing the brain volume are computed independently.