Sumanth mtc

Dockerfile

@@ -118,6 +118,7 @@ RUN apt-get update && \
 COPY . $mc_path/
 WORKDIR $mc_path
 
+RUN pip cache purge
 RUN pip install --no-cache-dir --upgrade --ignore-installed pip setuptools && \
     pip install --no-cache-dir .  && \
     pip install --no-cache-dir tensorboard cmake onnx && \
@@ -134,6 +135,7 @@ WORKDIR $mc_path/multic/cli
 RUN python -m slicer_cli_web.cli_list_entrypoint --list_cli
 RUN python -m slicer_cli_web.cli_list_entrypoint MultiCompartmentSegment --help
 RUN python -m slicer_cli_web.cli_list_entrypoint FeatureExtraction --help
+RUN python -m slicer_cli_web.cli_list_entrypoint FeatureExtraction-HG --help
 
 
 ENTRYPOINT ["/bin/bash", "docker-entrypoint.sh"]

multic/cli/FeatureExtraction-HG/FeatureExtraction-HG.py

@@ -0,0 +1,183 @@
+import sys
+import os, girder_client
+import numpy as np
+import pandas as pd
+from tiffslide import TiffSlide
+from ctk_cli import CLIArgumentParser
+import cv2
+import csv
+from skimage import measure
+from skimage.feature import graycomatrix, graycoprops
+from skimage.segmentation import find_boundaries
+from skimage.measure import regionprops
+from skimage.color import rgb2gray
+
+
+sys.path.append("..")
+
+
+def calculate_color_features(image, compartment_mask):
+
+    feature_values = {}
+    compartment_names = ['Luminal space compartment', 'PAS compartment', 'Nuclei compartment']
+
+    for compartment in range(3):  # As there are 3 compartments
+
+        compartment_pixels = image[compartment_mask == (compartment + 1)]
+
+        if len(compartment_pixels) > 0:
+
+            # Mean Color
+
+            mean_color = np.mean(compartment_pixels, axis=0)
+
+            for i, channel_value in enumerate(mean_color):
+
+                feature_values[f"Mean {['Red', 'Green', 'Blue'][i]} {compartment_names[compartment]}"] = channel_value
+
+            # Standard Deviation Color
+
+            std_dev_color = np.std(compartment_pixels, axis=0)
+
+            for i, channel_value in enumerate(std_dev_color):
+
+                feature_values[f"Standard Deviation {['Red', 'Green', 'Blue'][i]} {compartment_names[compartment]}"] = channel_value
+
+        else:
+
+            # If compartment has no pixels, set values to zero
+
+            for i in range(3):
+
+                feature_values[f"Mean {['Red', 'Green', 'Blue'][i]} {compartment_names[compartment]}"] = 0.0
+
+                feature_values[f"Standard Deviation {['Red', 'Green', 'Blue'][i]} {compartment_names[compartment]}"] = 0.0
+
+    return feature_values
+
+
+
+# Texture Features
+
+def calculate_texture_features(image, compartment_mask):
+
+    feature_values = {}
+
+    texture_feature_names = ['Contrast', 'Homogeneity', 'Correlation', 'Energy']
+
+    compartment_names = ['Luminal space compartment', 'PAS compartment', 'Nuclei compartment']
+
+    for compartment in range(3):  # As there are 3 compartments
+
+        compartment_pixels = (compartment_mask == compartment).astype(np.uint8)
+
+        compartment_image = cv2.bitwise_and(image, image, mask=compartment_pixels)
+
+        compartment_image_gray = rgb2gray(compartment_image)
+
+        compartment_image_gray_uint = (compartment_image_gray * 255).astype(np.uint8)
+
+        texture_matrix = graycomatrix(compartment_image_gray_uint, [1], [0], levels=256, symmetric=True, normed=True)
+
+        for i, texture_name in enumerate(texture_feature_names):
+
+            texture_feature_value = graycoprops(texture_matrix, texture_name.lower())
+
+            feature_values[f"{texture_name} {compartment_names[compartment]}"] = texture_feature_value[0][0]
+
+    return feature_values
+
+
+
+def calculate_features(image, compartment_mask):
+
+    color_features_dict = calculate_color_features(image, compartment_mask)
+
+    texture_features_dict = calculate_texture_features(image, compartment_mask)
+
+    all_features = {
+
+        #"Distance Transform Features": distance_transform_features_list,
+
+        "Color Features": color_features_dict,
+
+        "Texture Features": texture_features_dict,
+
+        #"Morphological Features": morphological_features_list
+
+    }
+
+    return all_features
+
+
+def main(args):
+    # Load the compartment mask and rgb image
+
+    compartment_mask = cv2.imread(args.sub_compartment_mask, cv2.IMREAD_GRAYSCALE)
+
+    image = cv2.imread(args.input_file, cv2.IMREAD_COLOR)
+
+    # Assigning custom labels to compartments. For easier calculations
+
+    # 29 - blue = nuclei,
+
+    # 76- red = PAS
+
+    # 149 - green = Luminal space
+
+    compartment_mask[compartment_mask == 29] = 3
+
+    compartment_mask[compartment_mask == 76] = 2
+
+    compartment_mask[compartment_mask == 149] = 1
+
+    compartment_mask[compartment_mask > 3] = 0
+
+    all_features = calculate_features(image, compartment_mask)
+
+    # Print the color features dictionary
+
+    print(all_features)
+
+    # Write the color features to a CSV file
+
+    csv_filename = args.basedir+"/tmp/features_output.csv"
+
+    os.makedirs(args.basedir+"/tmp/")
+
+    folder = args.basedir
+
+    gc_folder_id = folder.split('/')[-2]
+
+    gc = girder_client.GirderClient(apiUrl = args.girderApiUrl)
+
+    gc.setToken(args.girderToken)
+
+    f = open(csv_filename, 'w')
+    f.close()
+
+    with open(csv_filename, mode='w', newline='') as csv_file:
+
+        writer = csv.writer(csv_file)
+
+        for feature_name, value in all_features.items():
+
+            writer.writerow([feature_name, ""])  
+
+            if isinstance(value, dict):  
+
+                for sub_feature_name, sub_value in value.items():
+
+                    writer.writerow(["", f"{sub_feature_name}", sub_value])
+
+    file_name = csv_filename.split('/')[-1]
+
+    print("Uploading to DSA")
+    gc.uploadFileToFolder(gc_folder_id, csv_filename, name = file_name)
+    print("Done")
+
+
+
+
+
+

multic/cli/FeatureExtraction-HG/FeatureExtraction-HG.xml

@@ -0,0 +1,94 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<executable>
+  <category>HistomicsTK</category>
+  <title>Extract Features</title>
+  <description>Extracts a variety of features an annotated whole-slide image</description>
+  <version>0.1.0</version>
+  <documentation-url>https://github.com/SarderLab/deeplab-WSI</documentation-url>
+  <license>Apache 2.0</license>
+  <contributor>Sayat Mimar (UFL)</contributor>
+  <acknowledgements>This work is part of efforts in digital pathology by the Sarder Lab: University Of Florida.</acknowledgements>
+  <parameters>
+    <label>IO</label>
+    <description>Input/output parameters</description>
+    <directory>
+      <name>base_dir</name>
+      <label>WSI folder</label>
+      <description>Select the folder containing the annotated slide to be analyzed</description>
+      <channel>input</channel>
+      <index>0</index>
+    </directory>
+    <image>
+      <name>input_file</name>
+      <label>input_file</label>
+      <description>input file</description>
+      <channel>input</channel>
+      <index>1</index>
+    </image>
+    <double>
+      <name>downsample_factor</name>
+      <label>downsample_factor</label>
+      <description>downsample_factor</description>
+      <default>1.0</default>
+      <channel>input</channel>
+      <index>2</index>
+    </double>
+    <file fileExtensions=".xlsx" reference="output">
+      <name>output_filename</name>
+      <label>Output Excel Filename</label>
+      <description>Select the name and location of the Excel file produced. By default this will be saved in your Private folder.</description>
+      <channel>output</channel>
+      <index>3</index>
+    </file>
+    <file>
+      <name>sub_compartment_mask</name>
+      <label>sub_compartment_mask</label>
+      <description>sub_compartment mask</description>
+      <channel>input</channel>
+      <index>4</index>
+    </file>
+  </parameters>
+  <parameters advanced="true">
+    <label>Deconvolution Thresholds</label>
+    <description>Deconvolution Thresholds</description>
+    <integer>
+      <name>h_threshold</name>
+      <label>h-threhsold</label>
+      <description>h-threhshold for glomeruli and sclerotic glomeruli</description>
+      <longflag>h_threshold</longflag>
+      <default>160</default>
+    </integer>
+    <double>
+      <name>saturation_threshold</name>
+      <label>Saturation Threshold</label>
+      <description>Satruation Threshold for glomeruli and sclerotic glomeruli</description>
+      <longflag>saturation_threshold</longflag>
+      <default>0.3</default>
+    </double>
+    <double>
+      <name>whitespace_threshold</name>
+      <label>Whitespace Threshold</label>
+      <description>Whitespace Threshold for tubules</description>
+      <longflag>whitespace_threshold</longflag>
+      <default>0.88</default>
+    </double>
+  </parameters>
+  <parameters advanced="true">
+    <label>Girder API URL and Key</label>
+    <description>A Girder API URL and token for Girder client</description>
+    <string>
+      <name>girderApiUrl</name>
+      <longflag>api-url</longflag>
+      <label>Girder API URL</label>
+      <description>A Girder API URL (e.g., https://girder.example.com:443/api/v1)</description>
+      <default></default>
+    </string>
+    <string>
+      <name>girderToken</name>
+      <longflag>token</longflag>
+      <label>Girder API Token</label>
+      <description>A Girder token</description>
+      <default></default>
+    </string>
+  </parameters>
+</executable>

multic/cli/slicer_cli_list.json

@@ -4,5 +4,8 @@
   },
   "FeatureExtraction": {
     "type"    : "python"
+  },
+  "FeatureExtraction-HG": {
+    "type"    : "python"
   }
 }