data.rst

Data

The data module gives access to a set of publicly available WSIs, stained with different techniques (H&E and IHC). In particular, slides in the data module are retrieved from the following repositories:

• The Cancer Genome Atlas (TCGA): as detailed in the methods docstring, for each WSI, we access the URL pointing to the corresponding location within the portal, e.g. https://portal.gdc.cancer.gov/files/9c960533-2e58-4e54-97b2-8454dfb4b8c8, to retrieve the WSI;
• OpenSlide, a repository of freely-distributed test slides from different scanner vendors;
• Image Data Resource (IDR): the WSIs are selected from the data collection provided by Schaadt et al. [1] and available at IDR under the accession number idr0073.

Note

We use Pooch under the hood, which is an optional requirement for histolab and needs to be installed separately with:

pip install pooch

Set of downloadable WSIs.

Tissue	Dimensions (wxh)	Size (MB)	Repository	Staining
Aorta	15374x17497	63.8	OpenSlide	H&E
CMU small sample	2220x2967	1.8	OpenSlide	H&E
Breast	96972x30682	299.1	TCGA-BRCA	H&E
Breast (black pen)	121856x94697	1740.8	TCGA-BRCA	H&E
Breast (green pen)	98874x64427	719.6	TCGA-BRCA	H&E
Breast (red pen)	60928x75840	510.9	TCGA-BRCA	H&E
Breast (IHC)	99606x7121	218.3	IDR	IHC
Heart	32672x47076	289.3	OpenSlide	H&E
Kidney	5179x4192	66.1	IDR	IHC
Ovary	30001x33987	389.1	TCGA-OV	H&E
Prostate	16000x15316	46.1	TCGA-PRAD	H&E

TCGA-BRCA: TCGA Breast Invasive Carcinoma dataset; TCGA-PRAD: TCGA Prostate Adenocarcinoma dataset; TCGA-OV: Ovarian Serous Cystadenocarcinoma dataset.

.. toctree::
   :caption: API Reference
   :maxdepth: 2

.. automodule:: histolab.data
    :members:

.. toctree::

References

[1]	Schaadt NS, Schönmeyer R, Forestier G, et al. "Graph-based description of tertiary lymphoid organs at single-cell level." PLoS Comput Biol. (2020)