Enhancement and Segmentation Workflow for the Developing Zebrafish Vasculature.
Elisabeth C KuglerKaren PlantTimothy ChicoPaul ArmitagePublished in: Journal of imaging (2019)
Zebrafish have become an established in vivo vertebrate model to study cardiovascular development and disease. However, most published studies of the zebrafish vascular architecture rely on subjective visual assessment, rather than objective quantification. In this paper, we used state-of-the-art light sheet fluorescence microscopy to visualize the vasculature in transgenic fluorescent reporter zebrafish. Analysis of image quality, vascular enhancement methods, and segmentation approaches were performed in the framework of the open-source software Fiji to allow dissemination and reproducibility. Here, we build on a previously developed image processing pipeline; evaluate its applicability to a wider range of data; apply and evaluate an alternative vascular enhancement method; and, finally, suggest a work-flow for successful segmentation of the embryonic zebrafish vasculature.
Keyphrases
- deep learning
- convolutional neural network
- image quality
- single molecule
- computed tomography
- electronic health record
- high resolution
- crispr cas
- magnetic resonance imaging
- artificial intelligence
- machine learning
- quantum dots
- optical coherence tomography
- systematic review
- depressive symptoms
- high throughput
- meta analyses