Bioimage informatics
Harshavardhan Khare*, Nathaly Dongo Mendoza*, Chiara Zurzolo
Abstract
The implementation of computational tools for analysis of microscopy images has been one of the most important technological
innovations in biology, providing researchers unmatched capabilities to comprehend cell shape and connectivity. While numerous tools exist for
image annotation and segmentation, there is a noticeable gap when it comes to morphometric analysis of microscopy images. Most existing
tools often measure features solely on 2D serial images, which can be difficult to extrapolate to 3D. For this reason, we introduce CellWalker, a
computational toolbox that runs inside Blender, an open-source computer graphics software. This add-on improves the morphological analysis
by seamlessly integrating analysis tools into the Blender workflow, providing visual feedback through a powerful 3D visualization, and leveraging
the resources of Blender’s community. CellWalker provides several morphometric analysis tools that can be used to calculate distances, volume,
surface areas and to determine cross-sectional properties. It also includes tools to build skeletons, calculate distributions of subcellular organelles.
In addition, this python-based tool contains ‘visible-source’ IPython notebooks accessories for segmentation of 2D/3D microscopy images
using deep learning and visualization of the segmented images that are required as input to CellWalker. Overall, CellWalker provides practical
tools for segmentation and morphological analysis of microscopy images in the form of an open-source and modular pipeline which allows a
complete access to fine-tuning of algorithms through visible-source code while still retaining a result-oriented interface.
More information at https://doi.org/10.1093/bioinformatics/btad710
Commenti