Skip to content

Latest commit

 

History

History
43 lines (20 loc) · 880 Bytes

README.md

File metadata and controls

43 lines (20 loc) · 880 Bytes

mPCA_detect

binary cancer detection based on resnet50.

input: openslide compatible WSI

output: qupath json, probability map

usage: python detection_inference_cluster.py

[required] users should pass one of the following arguments:

--by_csv "/path/to/file.csv" containing full filepaths to WSI (see example)

--by_folder "/path/to/folder" folder path to WSI files, does not search recursively

--single_image "/path/to/image.tif" full filepath for openslide compatible WSI

[optional]

--save_location "/path/to/save/outputs" will default to "./output"

creating environment

conda create -n mpca_detect python=3.6 numpy scipy

conda install -c bioconda openslide

pip install openslide-python

conda install -c fastchan fastai

conda install scikit-image

conda install -c conda-forge opencv

pip install Shapely

pip install geojson