def get_ct_cropped():
paths_to_ct = relative_paths('./../../data_source/images/ct_nrrd', target_format='nrrd')
paths_to_mask = relative_paths('./../../data_source/images/masks_nrrd', target_format='nrrd')
ct_cropped = {}
for num, path_to_ct in enumerate(paths_to_ct):
ct_image, _ = nrrd.read(path_to_ct)
mask_image, _ = nrrd.read(paths_to_mask[num])
ct_cropped[path_to_ct] = ct_image * mask_image
return ct_cropped
def pet_cropped_img_value_stats():
path_to_fig = '../image_value_stats/cropped_pet_statistics.pdf'
path_to_images = relative_paths('./../../data_source/images/pet_nrrd',
target_format='nrrd')
path_to_masks = relative_paths('./../../data_source/images/masks_nrrd',
target_format='nrrd')
plt.figure()
plt.xlabel('Patient ID')
plt.ylabel('PET Intensity Statistic')
plot_img_value_stats(path_to_fig,
path_to_images,
path_to_masks,
include_legend=True)
def plot_clustering_ct_max_stats():
path_kmeans_distort = '../image_value_stats/kmeans_elbow.pdf'
path_to_images = relative_paths('./../../data_source/images/ct_nrrd',
target_format='nrrd')
images = load_images(path_to_images)
patient_id = CONFIG.patient_axis_ticks()
clustering_ct_max_stats(images, patient_id, path_kmeans_distort)
def cropping_illustration_imgs():
paths_to_ct = relative_paths('./../../data_source/images/ct_nrrd', target_format='nrrd')
paths_to_pet = relative_paths('./../../data_source/images/pet_nrrd', target_format='nrrd')
paths_to_mask = relative_paths('./../../data_source/images/masks_nrrd', target_format='nrrd')
path_to_ct = './../../data_source/images/ct_nrrd/P222CT.nrrd'
path_to_pet = './../../data_source/images/pet_nrrd/P222PET.nrrd'
path_to_mask = './../../data_source/images/masks_nrrd/P222mask.nrrd'
path_to_ct_img = '../illustrations/ct_img.pdf'
path_to_pet_img = '../illustrations/pet_img.pdf'
path_to_ct_cropped_img = '../illustrations/ct_cropped_img.pdf'
path_to_pet_cropped_img = '../illustrations/pet_cropped_img.pdf'
show = False
ct_image, _ = nrrd.read(path_to_ct)
pet_image, _ = nrrd.read(path_to_pet)
mask, _ = nrrd.read(path_to_mask)
n = 30
ct_image = ct_image[n, :, :]
pet_image = pet_image[n, :, :]
mask = mask[n, :, :]
cropped_ct = ct_image * mask
cropped_pet = pet_image * mask
plt.figure()
plt.axis('off')
plt.imshow(ct_image, vmin=np.min(ct_image), vmax=np.max(ct_image))
plt.savefig(
path_to_ct_img, bbox_inches='tight', transparent=True, dpi=DPI
)
if show:
plt.show()
plt.figure()
plt.axis('off')
plt.imshow(cropped_ct, vmin=np.min(ct_image), vmax=np.max(ct_image))
plt.savefig(
path_to_ct_cropped_img, bbox_inches='tight', transparent=True, dpi=DPI
)
if show:
plt.show()
plt.figure()
plt.axis('off')
plt.imshow(pet_image, vmin=np.min(pet_image), vmax=np.max(pet_image))
plt.savefig(
path_to_pet_img, bbox_inches='tight', transparent=True, dpi=DPI
)
if show:
plt.show()
plt.figure()
plt.axis('off')
plt.imshow(cropped_pet, vmin=np.min(pet_image), vmax=np.max(pet_image))
plt.savefig(
path_to_pet_cropped_img, bbox_inches='tight', transparent=True, dpi=DPI
)
if show:
plt.show()