-
Notifications
You must be signed in to change notification settings - Fork 1
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
3 changed files
with
272 additions
and
56 deletions.
There are no files selected for viewing
Binary file not shown.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,217 @@ | ||
| from __future__ import print_function, unicode_literals, absolute_import, division | ||
| import argparse | ||
|
|
||
| import sys | ||
| import numpy as np | ||
| import matplotlib | ||
| matplotlib.rcParams["image.interpolation"] = 'none' | ||
| import matplotlib.pyplot as plt | ||
|
|
||
| from glob import glob | ||
| from tifffile import imread | ||
| from csbdeep.utils import Path, normalize | ||
| from csbdeep.io import save_tiff_imagej_compatible | ||
|
|
||
| from stardist import random_label_cmap, _draw_polygons, export_imagej_rois | ||
| from stardist.models import StarDist2D | ||
|
|
||
| #Load previously saved model | ||
| from keras.models import load_model | ||
| from tensorflow.keras.metrics import MeanIoU | ||
|
|
||
| import tensorflow as tf | ||
| import datetime | ||
|
|
||
| from tensorflow.keras.utils import normalize | ||
| import os | ||
| import cv2 | ||
| from PIL import Image | ||
| import numpy as np | ||
| from matplotlib import pyplot as plt | ||
| from sklearn.preprocessing import MinMaxScaler | ||
| from keras.optimizers import Adam | ||
|
|
||
| from glob import glob | ||
| from tqdm import tqdm | ||
| from tifffile import imread | ||
| from csbdeep.utils import Path, normalize | ||
|
|
||
| import pandas as pd | ||
|
|
||
| def calculate_padded_dimensions(original_height, original_width, tile_size=256): | ||
| padded_height = (original_height + tile_size - 1) // tile_size * tile_size | ||
| padded_width = (original_width + tile_size - 1) // tile_size * tile_size | ||
| return padded_height, padded_width | ||
|
|
||
| #path = '/Users/danielfurth/Documents/GitHub/FluorPLA/figure01/230913/MAX_CMYC_MAX_1_MMStack_Pos0.ome.tif' | ||
|
|
||
| def main(image_path): | ||
| model = load_model("cytosol_unet_400epochs.hdf5", compile=False) | ||
|
|
||
| # Get the basename of the image file | ||
| image_basename = os.path.basename(image_path) | ||
|
|
||
| # Remove the file extension and add '.csv' as the new extension | ||
| csv_file = os.path.splitext(image_basename)[0] + '.csv' | ||
|
|
||
| # List to store the loaded image | ||
| multiP = [] | ||
|
|
||
| ret, multiP = cv2.imreadmulti(image_path, flags=cv2.IMREAD_ANYDEPTH) | ||
|
|
||
| # Check if images were loaded successfully | ||
| if ret: | ||
| # Separate the channels | ||
| antibody = multiP[0] # multiP[2] # First channel | ||
| nuclei = multiP[1] # Second channel | ||
| cytosol = multiP[2] # multiP[0] # Third channel | ||
|
|
||
| # Create a mask for values below the threshold | ||
| maskTmp = (antibody < 150) # 400 | ||
|
|
||
| # Set values below the threshold to 0 | ||
| antibody[maskTmp] = 0 | ||
|
|
||
|
|
||
| image_dataset = np.array(cytosol) | ||
|
|
||
| image_dataset = image_dataset /image_dataset.max() #Can also normalize or scale using MinMax scaler | ||
|
|
||
|
|
||
| # Assuming you have a NumPy array called 'image_dataset' with shape (2808, 2688) | ||
| original_height, original_width = image_dataset.shape | ||
| tile_size = 256 | ||
|
|
||
| padded_height, padded_width = calculate_padded_dimensions(original_height, original_width) | ||
| image_dataset_padded = np.pad(image_dataset, ((0, padded_height - original_height), (0, padded_width - original_width)), mode='constant') | ||
|
|
||
| # Calculate the number of tiles in both dimensions | ||
| num_tiles_height = padded_height // tile_size | ||
| num_tiles_width = padded_width // tile_size | ||
|
|
||
| # Initialize the tiled image array | ||
| tiledImage = np.zeros((num_tiles_height * num_tiles_width, tile_size, tile_size, 1), dtype=np.float32) | ||
|
|
||
| # Iterate over the original image and extract tiles | ||
| tile_index = 0 | ||
| for i in range(num_tiles_height): | ||
| for j in range(num_tiles_width): | ||
| # Extract a 256x256 tile from the original image | ||
| tile = image_dataset_padded[i * tile_size: (i + 1) * tile_size, j * tile_size: (j + 1) * tile_size] | ||
| tile = np.expand_dims(tile, axis=-1) # Add a channel dimension | ||
| tiledImage[tile_index] = tile | ||
| tile_index += 1 | ||
|
|
||
|
|
||
| y_pred=model.predict(tiledImage) | ||
|
|
||
| # Assuming you have a NumPy array called 'tiledImage' with shape (x, 256, 256, 1) | ||
| num_tiles, tile_height, tile_width, num_channels = y_pred.shape | ||
| tile_size = 256 | ||
|
|
||
| # Calculate the dimensions of the original image | ||
| new_height = num_tiles_height * tile_height | ||
| new_width = num_tiles_width * tile_width | ||
|
|
||
| # Initialize the original image array | ||
| reconstructed_image = np.zeros((new_height, new_width, num_channels), dtype=np.float32) | ||
|
|
||
| # Iterate over the tiles and reconstruct the original image | ||
| tile_index = 0 | ||
| for i in range(num_tiles_height): | ||
| for j in range(num_tiles_width): | ||
| # Get the tile from 'tiledImage' | ||
| tile = y_pred[tile_index] | ||
|
|
||
| # Place the tile in the reconstructed image | ||
| reconstructed_image[i * tile_size: (i + 1) * tile_size, j * tile_size: (j + 1) * tile_size] = tile | ||
| tile_index += 1 | ||
|
|
||
|
|
||
| float32_array = reconstructed_image | ||
|
|
||
| # Assuming you have a NumPy array called 'float32_array' | ||
| # Scale the float32 array to the range [0, 255] | ||
| scaled_array = ((float32_array - float32_array.min()) / (float32_array.max() - float32_array.min()) * 255).astype(np.uint8) | ||
|
|
||
| # Create a monochrome OpenCV Mat | ||
| monochrome_mat = cv2.cvtColor(scaled_array, cv2.COLOR_GRAY2BGR) | ||
|
|
||
| # Convert the BGR image to grayscale | ||
| monochrome_mat = cv2.cvtColor(monochrome_mat, cv2.COLOR_BGR2GRAY) | ||
|
|
||
| monochrome_mat = monochrome_mat[:original_height, :original_width] | ||
|
|
||
| # Apply a Gaussian filter with a 3-pixel sigma | ||
| sigma = 3 | ||
| filtered_image = cv2.GaussianBlur(monochrome_mat, (0, 0), sigma) | ||
|
|
||
| # Threshold the filtered image to create a binary image | ||
| _, binary_image = cv2.threshold(filtered_image, 0, 255, cv2.THRESH_BINARY) | ||
|
|
||
| # Perform image segmentation to create a labeled image | ||
| num_labels, labeled_image = cv2.connectedComponents(binary_image) | ||
|
|
||
| # Note: 'labeled_image' now contains the labeled regions. Each region has a unique integer label. | ||
| # 'num_labels' represents the total number of labeled regions. | ||
|
|
||
| # You can display the labeled image using a colormap for visualization: | ||
| #colored_labeled_image = cv2.applyColorMap((labeled_image * 255 / num_labels).astype(np.uint8), cv2.COLORMAP_JET) | ||
|
|
||
|
|
||
| axis_norm = (0,1) # normalize channels independently | ||
| # axis_norm = (0,1,2) # normalize channels jointly | ||
|
|
||
| model = StarDist2D(None, name='dapi', basedir='models') | ||
|
|
||
| img = normalize(nuclei, 1,99.8, axis=axis_norm) | ||
| labels, details = model.predict_instances(img) | ||
|
|
||
| # Get unique labels in the "labels" image | ||
| unique_labels = np.unique(labels) | ||
|
|
||
| # Create a DataFrame to store label intensities | ||
| data = {"Label": [], "Parent": [], "Cyto": [], "Background": [], "Nuclei": []} | ||
|
|
||
| # Iterate through unique labels and calculate mean intensity | ||
| for label in unique_labels: | ||
| if label == 0: | ||
| continue # Skip background label (0) | ||
|
|
||
| # Create a mask for the current label | ||
| label_mask = (labels == label).astype(np.int8) | ||
|
|
||
|
|
||
| # Calculate the mean intensity within the mask in the "antibody" image | ||
| label_intensity = np.mean(antibody[label_mask == 1]) | ||
| parent = int(np.mean(labeled_image[label_mask == 1])) | ||
|
|
||
| cytomask = (labeled_image == parent).astype(np.int8) | ||
| label_mask2 = cytomask - label_mask | ||
| background = np.mean(antibody[labeled_image == 0]) | ||
|
|
||
| cyto_intensity = np.mean(antibody[label_mask2 == 1]) | ||
|
|
||
| # Append the data to the DataFrame | ||
| data["Label"].append(label) | ||
| data["Parent"].append(parent) | ||
| data["Background"].append(background) | ||
| data["Cyto"].append(cyto_intensity) | ||
| data["Nuclei"].append(label_intensity) | ||
|
|
||
|
|
||
| # Create a DataFrame | ||
| df = pd.DataFrame(data) | ||
|
|
||
| # Save the results to a CSV file | ||
| #csv_file = "nuclei_intensities.csv" | ||
| df.to_csv(csv_file, index=False) | ||
|
|
||
| print(f"Results saved to {csv_file}") | ||
|
|
||
| if __name__ == "__main__": | ||
| parser = argparse.ArgumentParser(description="Image processing script") | ||
| parser.add_argument("image_path", type=str, help="Path to the image file") | ||
|
|
||
| args = parser.parse_args() | ||
| main(args.image_path) |