Cleaning up

L3_finder.py

@@ -267,6 +267,10 @@ def pixel_data(self):
             self.prediction_result.prediction.predicted_y_in_px,
             sagittal_z_pos_pair=self.sagittal_series.z_range_pair
         )
+
+    def pixel_data_manualL3(self, manualL3):
+        # Provide manual L3 location starting with 1 (not 0).
+        return self.axial_series.image_at_manualL3(manualL3)
 
     @property
     def height_of_sagittal_image(self):

MRI_notebooks/.neptune/async/0f6e5179-f99b-47f9-a34f-55dfc87c092d/exec-0-2021-12-17_01.32.58.064466/container_type

@@ -0,0 +1 @@
+run

MRI_notebooks/.neptune/async/0f6e5179-f99b-47f9-a34f-55dfc87c092d/exec-0-2021-12-17_01.32.58.064466/last_ack_version

@@ -0,0 +1 @@
+728

MRI_notebooks/.neptune/async/0f6e5179-f99b-47f9-a34f-55dfc87c092d/exec-0-2021-12-17_01.32.58.064466/last_put_version

@@ -0,0 +1 @@
+728

MRI_notebooks/.neptune/async/49a0e306-49ab-44f4-8e1e-0e3a05f071a3/exec-0-2021-12-16_01.43.52.539758/container_type

@@ -0,0 +1 @@
+run

MRI_notebooks/.neptune/async/49a0e306-49ab-44f4-8e1e-0e3a05f071a3/exec-0-2021-12-16_01.43.52.539758/last_ack_version

@@ -0,0 +1 @@
+21883

MRI_notebooks/.neptune/async/49a0e306-49ab-44f4-8e1e-0e3a05f071a3/exec-0-2021-12-16_01.43.52.539758/last_put_version

@@ -0,0 +1 @@
+21883

MRI_notebooks/.neptune/async/50259e50-463b-4feb-b36a-0cec876deefe/exec-0-2021-12-17_01.41.13.058866/container_type

@@ -0,0 +1 @@
+run

MRI_notebooks/.neptune/async/50259e50-463b-4feb-b36a-0cec876deefe/exec-0-2021-12-17_01.41.13.058866/last_ack_version

@@ -0,0 +1 @@
+648

MRI_notebooks/.neptune/async/50259e50-463b-4feb-b36a-0cec876deefe/exec-0-2021-12-17_01.41.13.058866/last_put_version

@@ -0,0 +1 @@
+648

MRI_notebooks/.neptune/async/7046ff50-5a30-4e41-961f-20db02f54dc6/exec-0-2021-12-17_02.10.32.382671/container_type

@@ -0,0 +1 @@
+run

MRI_notebooks/.neptune/async/7046ff50-5a30-4e41-961f-20db02f54dc6/exec-0-2021-12-17_02.10.32.382671/last_ack_version

@@ -0,0 +1 @@
+21814

MRI_notebooks/.neptune/async/7046ff50-5a30-4e41-961f-20db02f54dc6/exec-0-2021-12-17_02.10.32.382671/last_put_version

@@ -0,0 +1 @@
+21814

MRI_notebooks/.neptune/async/b560997f-6eb6-4061-99f1-bdea18e28da7/exec-0-2021-12-17_01.42.03.224406/container_type

@@ -0,0 +1 @@
+run

MRI_notebooks/.neptune/async/b560997f-6eb6-4061-99f1-bdea18e28da7/exec-0-2021-12-17_01.42.03.224406/last_ack_version

@@ -0,0 +1 @@
+21962

MRI_notebooks/.neptune/async/b560997f-6eb6-4061-99f1-bdea18e28da7/exec-0-2021-12-17_01.42.03.224406/last_put_version

@@ -0,0 +1 @@
+21962

MRI_notebooks/TIF_loader.py

@@ -0,0 +1,211 @@
+# import the necessary packages
+import os
+import tables
+import glob
+import numpy as np
+import SimpleITK as sitk
+from pathlib import Path
+import tables
+
+def normalize_image(image):
+    image = sitk.GetArrayFromImage(image)
+    mean = image.mean()
+    std = image.std()
+    #print("orig im mean: ", mean,"orig im std: ", std)
+    norm_im = (image - mean) / std
+    #print("norm im mean: ", norm_im.mean(),"norm im std: ", norm_im.std())
+    return sitk.GetImageFromArray(norm_im)
+
+def _load_l3_tif(imagepath,normalize=True):
+    img = sitk.GetArrayFromImage(sitk.ReadImage(imagepath))
+    if normalize:
+        img = normalie_image(img)
+    return img
+
+def invertimage_excluding_background(MR_nda,l_thresh_p=5,h_thresh_p=85,seed=[128, 128]): #v3
+    MR = normalize_image(MR_nda)    
+    #** Only for V2 and above
+    MR_range = max(MR) - min(MR)
+    #     print('MR min', min(MR))
+    l_thresh = MR_range*l_thresh_p/100 + min(MR)
+    h_thresh = MR_range*h_thresh_p/100 + min(MR)
+    #    print("new h_thresh: ", h_thresh)
+    # Blur using CurvatureFlowImageFilter
+    #
+    blurFilter = sitk.CurvatureFlowImageFilter()
+    blurFilter.SetNumberOfIterations(5)
+    blurFilter.SetTimeStep(0.125)
+    image = blurFilter.Execute(MR)
+    #
+    # Set up ConneMRedThresholdImageFilter for segmentation
+    #
+    segmentationFilter = sitk.ConnectedThresholdImageFilter()
+    segmentationFilter.SetLower(float(l_thresh))
+    segmentationFilter.SetUpper(float(h_thresh))
+    segmentationFilter.SetReplaceValue(1)
+    segmentationFilter.AddSeed(seed)
+    segmentationFilter.AddSeed([64,128]) # Adding additional seeds for safety
+    segmentationFilter.AddSeed([128,64])
+    # Run the segmentation filter
+    image = segmentationFilter.Execute(image)
+    image[seed] = 1
+     # Fill holes
+    image = sitk.BinaryFillhole(image);
+    # Intensity Invert FIlter
+    invertfilter = sitk.InvertIntensityImageFilter()
+    MR_invert = normalize_image(invertfilter.Execute(MR))
+    #     print("MR_invert min: ", min(MR_invert))
+    #     print("MR_invert max: ", max(MR_invert))
+    # Masking FIlter
+    maskingFilter = sitk.MaskImageFilter()
+    background_value = min(MR_invert)-0.5
+    maskingFilter.SetOutsideValue(background_value)
+    #     print("Name: ", maskingFilter.GetName())
+    #     print("Masking Value: ", maskingFilter.GetMaskingValue())
+    #     print("Outside Value: ", maskingFilter.GetOutsideValue())
+    MR_noTable = maskingFilter.Execute(MR_invert,image)
+    #     print("MR_noTable min: ", min(MR_noTable))
+    return MR_noTable, image
+
+def _load_l3_tif_invert(imagepath):
+    img = sitk.ReadImage(imagepath)
+    desired_img, mask = invertimage_excluding_background(img)
+    return sitk.GetArrayFromImage(desired_img) #, mask, img
+
+def reslice_image_set_TIF(in_files, image_shape, out_files=None, label_indices=None,normalize=True,invert_image=False):
+    images = list()
+    for f, file in enumerate(in_files):
+         # Check if LABEL
+        label = False
+        image = None
+        if f == label_indices:
+            label = True
+            image = _load_l3_tif(file,False)
+            image[image > 0] = 1
+        else:
+            if invert_image:
+                image = _load_l3_tif_invert(file)
+            else:
+                image = _load_l3_tif(file,normalize)
+        # Resize
+        if (image_shape[0] != image.shape[0]) or (image_shape[0] != image.shape[1]):
+            desired_size = image_shape[0]
+            image = resize_pad(image,desired_size,label)
+        print('After Resizing image max: ', np.max(image))
+        print('After Resizing image min: ', np.min(image))
+        images.append(image)
+
+    if out_files:
+        for image, out_file in zip(images, out_files):
+            image.to_filename(out_file)
+        return [os.path.abspath(out_file) for out_file in out_files]
+    else:
+        return images
+
+class mri_TIF_loader:
+    def __init__(self,problem_type='Classification',input_images=None,input_shape=(128,128),image_modalities=['T2'],mask=None,slice_number=0,invert_image=False,normalize_on_load=True):
+        self.input_images = input_images
+        self.input_shape = input_shape
+        self.problem_type = problem_type
+        self.image_modalities = image_modalities
+        self.mask = mask
+        self.slice_number = slice_number
+        self.invert_image = invert_image # Flag to invert or not invert MRI images
+        self.normalize_on_load = normalize_on_load # Flag to normalize input images, works for MR. For CT do not use if HU needs to be preserved! Will not apply for mask/label/truth files
+        if self.problem_type is 'Segmentation':
+            training_data_files = list()
+            subject_ids = list()
+            for subject_dir in glob.glob(os.path.join(self.input_images, "*")):
+                subject_ids.append(os.path.basename(subject_dir))
+                subject_files = list()
+                for modality in self.image_modalities + self.mask:
+                    subject_files.append(os.path.join(subject_dir, modality + ".tif"))
+                training_data_files.append(tuple(subject_files))
+            self.data_files = training_data_files
+            self.ids = subject_ids
+            self.image_data_shape = tuple([0, len(self.image_modalities)] + list(self.input_shape))
+            self.truth_data_shape = tuple([0, len(self.mask)] + list(self.input_shape))
+            self.n_channels = len(self.image_modalities)
+
+
+        elif problem_type is 'Classification':
+            training_data_files = list()
+            subject_ids = list()
+            for classes in glob.glob(os.path.join(self.input_images, "*")):
+                for subject_dir in glob.glob(os.path.join(classes, "*")):
+                    subject_ids.append(os.path.basename(subject_dir))
+                    subject_files = list()
+                    for modality in self.image_modalities + self.mask:
+                        subject_files.append(os.path.join(subject_dir, modality + ".tif"))
+                    training_data_files.append(tuple(subject_files))
+            self.data_files = training_data_files
+            self.ids = subject_ids
+            self.image_data_shape = tuple([0, len(self.image_modalities)+len(self.mask)] + list(self.input_shape))
+            self.truth_data_shape = tuple([0,])
+            self.n_channels = len(self.image_modalities)+len(self.mask)
+
+        #elif problem_type is 'Regression':
+        #    training_data_files = list()
+        #    subject_ids = list()
+        #    for subject_dir in glob.glob(os.path.join(self.input_images, "*")):
+        #        subject_ids.append(os.path.basename(subject_dir))
+        #        subject_files = list()
+        #        for modality in self.image_modalities:
+        #            subject_files.append(os.path.join(subject_dir, modality + ".TIF"))
+        #        training_data_files.append(tuple(subject_files))
+        #    self.data_files = training_data_files
+        #    self.ids = subject_ids
+
+    def get_sample_ids(self):
+        return self.ids
+
+    def set_sample_ids(self,new_ids):
+        self.ids = new_ids
+
+    def load_toHDF5(self,hdf5_file=None,verbose=-1):
+		    # initialize the list of features and labels
+            n_samples = len(self.ids)
+            filters = tables.Filters(complevel=5, complib='blosc')
+            image_storage = hdf5_file.create_earray(hdf5_file.root, 'imdata', tables.Float32Atom(), shape=self.image_data_shape, filters=filters, expectedrows=n_samples)
+
+            if self.problem_type is "Classification":
+                truth_storage =  hdf5_file.create_earray(hdf5_file.root, 'truth', tables.StringAtom(itemsize=15), shape=self.truth_data_shape, filters=filters, expectedrows=n_samples)
+            elif self.problem_type is "Segmentation":
+                truth_storage =  hdf5_file.create_earray(hdf5_file.root, 'truth', tables.UInt8Atom(), shape=self.truth_data_shape, filters=filters, expectedrows=n_samples)
+
+            # loop over the input images
+            for (i, imagePath) in enumerate(self.data_files):
+			    # load the image and extract the class label assuming
+			    # that our path has the following format:
+			    # /path/to/dataset/{class}/{image}.jpg
+                if self.problem_type is "Classification":
+                    subject_name = imagePath[0].split(os.path.sep)[-2]
+                    if subject_name in self.ids:
+                        images = self.get_images(in_files=imagePath, image_shape=self.input_shape, label_indices=len(imagePath)-1,invert_image=self.invert_image,normalize=self.normalize_on_load)
+                        label = imagePath[0].split(os.path.sep)[-3]
+                        subject_data = [image for image in images]
+                        image_storage.append(np.asarray(subject_data)[np.newaxis])
+                        truth_storage.append(np.asarray(label)[np.newaxis])
+
+                elif self.problem_type is "Segmentation":
+                    subject_name = imagePath[0].split(os.path.sep)[-2]
+
+                    if subject_name in self.ids:
+                        images = self.get_images(in_files=imagePath, image_shape=self.input_shape, label_indices=len(imagePath)-1,invert_image=self.invert_image,normalize=self.normalize_on_load)
+                        subject_data = [image for image in images]
+                        image_storage.append(np.asarray(subject_data[:self.n_channels])[np.newaxis])
+                        #DEBUG 
+                        #image = np.asarray(subject_data[:self.n_channels])
+                        truth_storage.append(np.asarray(subject_data[self.n_channels],dtype=np.uint8)[np.newaxis][np.newaxis])
+
+                #elif self.problem_type is "Regression":
+                #    image = cv2.imread(imagePath)
+
+			    # show an update every `verbose` images
+                if verbose > 0 and i > 0 and (i + 1) % verbose == 0:
+                    print("[INFO] processed {}/{}".format(i + 1,len(self.ids)))
+            return(image_storage)
+
+    def get_images(self, **loader_args):
+        return reslice_image_set_TIF(**loader_args)
+

MRI_notebooks/Untitled.ipynb

@@ -0,0 +1,6 @@
+{
+ "cells": [],
+ "metadata": {},
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

MRI_notebooks/Untitled1.ipynb

@@ -0,0 +1,32 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.6.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}