BigImage is an extension of 3D Slicer. It is used for large scale microscopic image viewing and computing.
Computational histopathology is a fast emerging field which converts the traditional glass slide based department to a new examination platform. Such a paradigm shift also brings the in silico computation to the field. Much research have been presented in the past decades on the algorithm development for pathology image analysis. On the other hand, a comprehensive software platform with advanced visualization and computation capability, large developer community, flexible plugin mechanism, and friendly transnational license, would be extremely beneficial for the entire community.
BigImage is an open platform for whole slide histopathology image computing based on the highly successful 3D Slicer.
In addition to viewing the giga-pixel whole slide image viewing, currently, the extension also offers several specific analytical modules for qualitative presentation, nucleus level analysis, tissue scale computation, and 3D pathology. Thanks to the openess of Slicer, the BigImage extension could also be further extended by you.
On Windows, all required dependencies are automatically installed. On Linux and macOS OpenSlide must be installed manually as described below.
You will need to manually install the openslide library to your OS. This can be done with, e.g.,
sudo apt install libopenslide0
This extension currently has two modules:
- BigImageViewer: This module loads and views the WSI images.
- ColorDecomposition: This module performs the color/staining decomposition of the image.
Their detailed usages are listed below. More modules including nucleus segmentaiton and others will be uploaded soon.
Example large scale wholse slide image can be downloaded at, e.g., the OpenSlide website
Swith to the "BigImageViewer" module in the BigImage category. As shown in the module panel, select the WSI file to view in the "Select WSI" box. Then click "Load WSI" below.
One can then view different region/scale of the image, using mouse dragging and mosue wheeling, as shown below:
Histopathology images are often stained using different dyes. When a WSI is stained using multiple dyes, the different stains can be computationally separated using the module "ColorDecomposition", whose panel is shown below.
There are various options for color decomposition. This particular WSI is stained with H-E. Its original appearance is:
After decomposition, the hematoxylin content is shown in gray-scale as:
where the dark regions corresponding to the high hematoxylin content.
If the eosin chanel is wanted, one can switch the output chanel in the module panel to the 2nd chanel, and the result will be like:
The extension contains an experimental module (NgffImageIO)
for reading OME-NGFF file format. Currently, only a simple image array can be saved and loaded
in Zarr format (with the .zarr file extension, with ZipStorage class), but we do not follow the NGFF specification yet.
This module may be used in the future instead of OpenSlide to make dependencies simpler, and to store more complete metadata.
If you find this extension helpful please cite this paper:
Xiaxia Yu, Bingshuai Zhao, Haofan Huang, Mu Tian, Sai Zhang, Hongping Song, Zengshan Li, Kun Huang, Yi Gao, "An Open Source Platform for Computational Histopathology," in IEEE Access, vol. 9, pp. 73651-73661, 2021, doi: 10.1109/ACCESS.2021.3080429.