3DFM_AC and 3DFM_BCPop Datasets

Published: 17 January 2023| Version 1 | DOI: 10.17632/f6kk4364p4.1
, Alison Noble


Two fluorescent microscopy image datasets of retinal neurons and neural populations. 3DFM_AC: A dataset of 22 3D fluorescent microscopy stacks acquired of individual EYFP-expressing Starburst Amacrine Cells (SACs) from the mouse retina and their corresponding ground truths. Each SAC was acquired 8 times with 5 differing laser powers (0.81uW, 1.85uW, 3.84uW, 7.71uW, 15.70uW). The xy-resolution and z-resolution of the acquired stacks were 0.621um and 0.100um respectively. 3DFM_BCPop: A dataset of 2 fluorescent microscopy stacks acquired of GFP-expressing Bipolar Type 2 Cell populations from the mouse retina and their corresponding ground truths. Each Bipolar Cell population was acquired by averaging 8 frames with a laser power of 15.70uW. The xy-resolution and z-resolution of the acquired stacks were 0.155um and 0.100um respectively.


Steps to reproduce

Animal procedures were conducted according to institutional guidelines and approved by the NINDS Animal Care and Use Committee (ASP-1344). For our 3DFM_AC dataset, ChAT-Cre mice (Ivanova et al. (2010)) were intravitreally injected with 2uL of AAV-7m8-EF1a-BbTagBY virus with a ∽0.8×10e12 vg/mL titration (Cai et al. (2013)). Retinas were then dissected 3-4 weeks after injection, fixed with 4% paraformaldehyde for 45 minutes, and washed three times. A confocal microscope (Zeiss LSM 510) with a Plan-Neofluar 40x/1.3 Oil objective, a 488 laser line, and a BP505-530 filter was then used to acquire stacks of the fixed retinas. To acquire the 3DFM_BCPop dataset, the retina of a Syt2 mouse was dissected, fixed, and imaged using the same protocol with only a laser power of 15.70uW and a magnification of 0.25. Ground truths were created for both datasets by inputting the highest quality available image into a Frangi filter and then binarizing the output of that filter. References: [1] Ivanova, E., Hwang, G.S., Pan, Z.H., 2010. Characterization of transgenic mouse lines expressing cre recombinase in the retina. Neuroscience 165, 233–243. [2] Cai, D., Cohen, K.B., Luo, T., Lichtman, J.W., Sanes, J.R., 2013. Improved tools for the brainbow toolbox. Nature methods 10, 540–547.


National Institute of Neurological Disorders and Stroke, University of Oxford


Microscopy, Neuron, Image Analysis