总结一下目前遇到的一些医学图像预处理步骤，但才开始自学两周，所以很多东西或许不对或者不全面，以后争取每两周回顾以前的文章，将内容完善，并提高质量。（2019-3-5）
由于篇幅原因，将预处理的步骤分开来，这一篇将总结读取图片的方法。
读取DICOM格式
用到的包：pydicom
例如，以公共数据集3Dircadb为例，其文件组织如下：
其中3Dircadb1.1代表病人1.1号，PATIENT_DICOM是病人腹部切面，LABELED_DICOM 是标记了不同部位的切片，MASKS_DICOM下面又有子文件夹，分别是不同部位的mask。所有的图片都是灰度图像。
读取dicom文件
label_path = '~/Downloads/3Dircadb/3Dircadb1.1/LABELLED_DICOM'
# 若不遍历读出，则获得的是单张切片，而非三维图片
slices = [pydicom.dcmread(label_path + '/' + s) for s in os.listdir(label_path)]
slices.sort(key = lambda x: int(x.InstanceNumber))
# slice[i]为pydicom.dataset.FileDataset，储存了切片相关的信息
# .pixel_array用来获取像素信息,每张切片都是灰度图
image = np.stack([s.pixel_array for s in slices])
image = image.astype(np.int16)
# 目前还不知道这些label值对应的器官分别是什么？
np.unique(image)
#array([ 0, 1, 17, 33, 65, 97, 129, 193, 257, 321, 385, 449, 451,
# 453, 465, 481, 513], dtype=int16)
label_path = '~/Downloads/3Dircadb/3Dircadb1.1/MASKS_DICOM/artery'
slices = [pydicom.dcmread(label_path + '/' + s) for s in os.listdir(label_path)]
slices.sort(key = lambda x: int(x.InstanceNumber))
image = np.stack([s.pixel_array for s in slices])
image = image.astype(np.int16)
np.unique(image)
# array([ 0, 255], dtype=int16)
读取普通的图片
用到的包 PIL，cv2, skimage
注意它们读取出的格式有细微的区别
| 包 | 图片通道 | 格式 | |
|---|---|---|---|
| PIL | RGB | Image | channel_last |
| cv2 | BGR | array | channel_last |
| skimage | RGB | array | channel_last |
还需要注意，对于opencv来说，无论读取灰度图还是彩图都是(H,W,3)的shape，灰度图的读取会把单通道复制三遍。因此，读取灰度图时得显示声明img = cv2.imread('gray.jpg', cv2.CV_LOAD_IMAGE_GRAYSCALE)或者img = cv2.imread('gray.jpg',0)其中0代表灰度图，1代表彩图，读出来的shape为（H,W)
案例一：skimage读取图片
下面使kaggle竞赛的一个数据集，数据集为超声波灰度图像。
import os
import numpy as np
from skimage.io import imsave, imread
data_path = '～/kaggle_Ultrasound'
image_rows = 420
image_cols = 580
train_data_path = os.path.join(data_path, 'train')
images = os.listdir(train_data_path)
# 原图为灰度图
imgs = np.ndarray((total, image_rows, image_cols), dtype=np.uint8)
for image_name in images:
img = imread(os.path.join(train_data_path, image_name), as_grey=True)
# 相当于扩维处理 由(420,580) --> (1,420,580)
img = np.array([img])
imgs[i] = img
# 可选项
np.save('imgs_train.npy', imgs)
"""
当然，在输入网络前还得进行处理，包括增加通道使格式与网络匹配、归一化处理等等。
"""
案例二，PIL读取图片
公共数据集DRIVE，分割眼球血管
文件组织格式：
下面的代码没有进行精简，可以只看重点
import os
import h5py
import numpy as np
from PIL import Image
def write_hdf5(arr,outfile):
with h5py.File(outfile,"w") as f:
f.create_dataset("image", data=arr, dtype=arr.dtype)
#------------Path of the images --------------------------------------------------------------
#train
original_imgs_train = "./DRIVE/training/images/"
groundTruth_imgs_train = "./DRIVE/training/1st_manual/"
borderMasks_imgs_train = "./DRIVE/training/mask/"
#test
original_imgs_test = "./DRIVE/test/images/"
groundTruth_imgs_test = "./DRIVE/test/1st_manual/"
borderMasks_imgs_test = "./DRIVE/test/mask/"
#---------------------------------------------------------------------------------------------
Nimgs = 20
channels = 3
height = 584
width = 565
dataset_path = "./DRIVE_datasets_training_testing/"
def get_datasets(imgs_dir,groundTruth_dir,borderMasks_dir,train_test="null"):
imgs = np.empty((Nimgs,height,width,channels))
groundTruth = np.empty((Nimgs,height,width))
border_masks = np.empty((Nimgs,height,width))
for path, subdirs, files in os.walk(imgs_dir): #list all files, directories in the path
for i in range(len(files)):
#original
print("original image: ",files[i])
img = Image.open(imgs_dir+files[i])
imgs[i] = np.asarray(img)
#corresponding ground truth
groundTruth_name = files[i][0:2] + "_manual1.gif"
print("ground truth name: ", groundTruth_name)
g_truth = Image.open(groundTruth_dir + groundTruth_name)
groundTruth[i] = np.asarray(g_truth)
#corresponding border masks
border_masks_name = ""
if train_test=="train":
border_masks_name = files[i][0:2] + "_training_mask.gif"
elif train_test=="test":
border_masks_name = files[i][0:2] + "_test_mask.gif"
else:
print("specify if train or test!!")
exit()
print("border masks name: ", border_masks_name)
b_mask = Image.open(borderMasks_dir + border_masks_name)
border_masks[i] = np.asarray(b_mask)
print("imgs max: ", str(np.max(imgs)))
print("imgs min: ", str(np.min(imgs)))
assert(np.max(groundTruth)==255 and np.max(border_masks)==255)
assert(np.min(groundTruth)==0 and np.min(border_masks)==0)
print("ground truth and border masks are correctly withih pixel value range 0-255 (black-white)")
assert(imgs.shape == (Nimgs,height,width,channels))
groundTruth = np.reshape(groundTruth,(Nimgs,height,width,1))
border_masks = np.reshape(border_masks,(Nimgs,height,width,1))
assert(groundTruth.shape == (Nimgs,height,width,1))
assert(border_masks.shape == (Nimgs,height,width,1))
return imgs, groundTruth, border_masks
if not os.path.exists(dataset_path):
os.makedirs(dataset_path)
#getting the training datasets
imgs_train, groundTruth_train, border_masks_train = get_datasets(original_imgs_train,groundTruth_imgs_train,borderMasks_imgs_train,"train")
print("saving train datasets")
write_hdf5(imgs_train, dataset_path + "DRIVE_dataset_imgs_train.hdf5")
write_hdf5(groundTruth_train, dataset_path + "DRIVE_dataset_groundTruth_train.hdf5")
write_hdf5(border_masks_train,dataset_path + "DRIVE_dataset_borderMasks_train.hdf5")
读取nii格式
用到的包：SimpleITK
需要注意，一个nii文件就是一个三维图片，与上面的dcm格式不同，而且SimpleITK读取出来的是( z, y ,x) 即第一维是轴向，然后是高、宽。 seg = itk.ReadImage()读取出的顺序为(x,y,z)，即(width, height, depth)，可用seg.GetSize() seg.GetWidth() seg.GetHeight() image.GetDepth()验证，注意！！当将其转化为数组后
segimg = sitk.GetArrayFromImage(seg)，通道顺序会反向，变成(z,y,x) ，即(depth, height, width)
用到的是公共数据集LITS17，原始文件组织：
为了方便处理，对其进行了分类处理：
unzip ./*.zip
mkdir ct
mkdir seg
mv ./volume-*.nii ./ct/
mv ./segmentation-*.nii ./seg/
这样将ct与分割图像分开来，也就有了下面示例代码的路径写法
ct_dir = '～/LITS17/ct/'
seg_dir = '～/LITS17/seg/'
for ct_file in os.listdir(ct_dir):
# 将CT和金标准入读内存
ct = sitk.ReadImage(os.path.join(ct_dir, ct_file), sitk.sitkInt16)
# ct_array:(629, 512, 512)
# 注意读取出来是z y x，即切片数量在最前
# 而 origin和position读取出来的是 x y z
ct_array = sitk.GetArrayFromImage(ct)
# vol_values=np.unique(ct_array) 有2708个值
seg = sitk.ReadImage(os.path.join(seg_dir, ct_file.replace('volume', 'segmentation')), sitk.sitkInt8)
seg_array = sitk.GetArrayFromImage(seg)