Vasculature Segmentation Validation Dataset - Part II - Decreasing Contrast-To-Noise Ratio

Dataset to allow exploration of data enhancement, segmentation, and validation for: https://www.biorxiv.org/content/10.1101/2020.07.21.213843v1 and associated future publications

Dataset description:

• LaserForNoise: Example data of the head vasculature of 4dpf of zebrafish. The data were consecutively acquired with 1.2%, 0.8%, and 0.4% laser power (LP), respectively. This allowed us to produce data with decreasing contrast-to-noise ratio, and thus study robustness over a range of signal distributions.

Data acquisition:
Experiments were performed according to the rules and guidelines of institutional and UK Home Office regulations under the Home Office Project Licence 70/8588 held by TC. Maintenance of adult zebrafish Tg(kdrl: HRAS-mCherry)s916 (Chi et al., 2008) was performed as described in standard husbandry protocols (Westerfield, 1993). Embryos, obtained from controlled mating, were kept in E3 medium buffer with methylene blue and staged as previously described (Kimmel et al., 1995).

Embryos were embedded in 2% LMP-agarose with 0.01% Tricaine in E3 (MS-222, Sigma). Data were acquired using a Zeiss Z.1 light sheet microscope, Plan-Apochromat 20x/1.0 Corr nd=1.38 objective, dual-side illumination with online fusion, activated Pivot Scan, image acquisition chamber incubation at 28°C, with a scientific complementary metal-oxide-semiconductor (sCMOS) detection unit. Data properties can be summarised as: 16bit image depth, voxel dimensions in x, y and z of 1920 x 1920 x 400-600, respectively, giving a voxel size of 0.33 x 0.33 x 0.5 µm). 

Contact: kugler.elisabeth[at]gmail.com

Useful code links: 

• https://github.com/ElisabethKugler/ZFVascularQuantification
• https://github.com/ElisabethKugler/Matlab3D-ImageAnalysis