Python Package: PyHistopathology

Read our documentation at https://pyhistopathology.readthedocs.io/en/latest/

Command line tool:

python3 WSI_PATCH_Extraction.py -args Mandatory args

• -i: input svs file path
• -o: output folder path
• -f: input folder for path
  • Note: you should use either -i or -f, cannot use both. Additional args
• -c: criteria
  • criteria: Random or None, Default is None
• -s: patch size
  • Size of the patch to extract, default is (256,256)
• -n: number of patches
  • Only should be given for -c Random. Default value is 2000.
• -a: input xml
  • if annotations are provided annotations file path should be given. Otherwise don't use this arg.

Package Usage:

Reading WSI

Description

• use WSI_Scanning.readWSI() to read an WSI Image

• Input: WSI path or directory

• Output: functioning numpy array of WSI Image with dtype int32 Function

• readWSI(WSI_path, magnification_level, annotation_file, annonated_level) Arguments

  • WSI_path: Directory of WSI
• • magnification_level: level of zoom, example (40x,20x,10x,5x). Default magnification is “20x” - Note if magnification 40x for max zoom level of 20x image an error will be raised.
  • annotation_file: Default annotation = None. if annotation are available in xml formats. use annotation = inputxml file path.
  • annonated_level= if annotation is not giving no need to consider this variable. if annotation is given then mention z-axis of annotations. Default annotatedlevel =0

     ###Reading image example
     from WSI_Preprocessing.Preprocessing import WSI_Scanning 
     import cv2 
     img,slide_dim = WSI_Scanning.readWSI("example.svs") 
     cv2.imwrite("example.png",img)
  

Denoising WSI

Description

use Denoising.denoising() to remove stains, folds and other background noise in WSI

• input: WSI Path or directory
• Output: functioning numpy array of WSI Image (After denoising) with dtype int32.

Function

denoising(inputsvs, magnification, filtering, patch_size, upperlimit, lowerlimit, red_value, green_value, blue_value, Annotation, Annotatedlevel, Requiredlevel, meannumber) Arguments

• inputsvs: path or location of WSI.
• magnification: level of zoom, example (40x,20x,10x,5x). Default magnification is “20x”
  • Note if magnification 40x for max zoom level of 20x image an error will be raised.
• filtering: GuassianBlur, RGBThersholding, None
  • GuassianBlur: Homogeneity calculations based on image smoothing and Gaussian blur equations. We compute sum of square differences between two consecutive Gaussian blurred images as score for homogeneity
    - Upper limit: upper threshold of homogeneity score. default value is 9500 with kernel size of 1111 - lower limit: lower threshold of homogeneity score. default value is 1500 with kernel size of 1111 - Patch size: Not significant parameters for GuassianBlur filtering - meannumber: Theshold of mean area to remove small patches default number is 8
  • RGBThersholding: Validated patches based on RGB values of patches - red_value, green-value, blue_values are threshold for RGB
• • None: Only removes Background
• • Note that our default is GuassianBlur technique. GuassianBlur is highly effective and requires more computational power (RAM). RGBThersholding is less effective which needs less computational power

import cv2 
# Here mandatory options are example.svs and magnification 
img = Denoising.denoising("example.svs", "20x" ) 
cv2.imwrite("example.png",img)

Extracting Patches

Description

use Extractingpatches.extractingPatches() to extract patches from WSI.

• input: WSI Path or directory
• output: patches from WSI.

Function: extractingPatches(inputsvs, outputpath, magnification, patch_extraction_creatia, number_of_patches, filtering, patch_size, upperlimit, lowerlimit, red_value, green_value, blue_value, Annotation, Annotationlevel, Requiredlevel, reconstructionimagepath)

Arguments - inputsvs, magnification, patch_extraction_creatia, filtering, patch_size, upperlimit, lowerlimit, red_value, green_value, blue_value, Annotation, Annotationlevel, Requiredlevel, arguments is same as denosing module. - patch_extraction_creatia: random, None - - Default is None. For extracting a fixed number of patches for WSI we can use random. - Default number of patches is 2000 - outputpath: folder to store the extracted patches - reconstructionimagepath: we you want to compare the patches with WSI we can mention the reconstructionimagepath. - Default is None - - Note: it only works with patch_extraction_creatia = None.

• • Note: For WSI number of patches can exceed 20k.

from WSI_Preprocessing.Preprocessing import Extarctingpatches
import cv2 
img = Extarctingpatches.extractingPatches("example.svs","temp" ,"20x" ) 
cv2.imwrite("exampler.png",img)
     # Here mandatory options are example.svs and magnification, and outputpath 
Extractingpatches.extractingPatches(example.svs, outputpath, magnification)