本笔记来源于B站Up主: 有Li 的影像组学系列教学视频
本节(35)主要介绍: 2D超声影像组学的特征提取
视频中李博士情境再现了小白研究者可能碰到的各种技术难题，并演示了解决这些问题的思路。
-
将
dicom格式的2D 超声图像转为压缩的nifti格式，将其命名为test.nii.gz; 勾画ROI后同样保存为压缩的nifti格式，命名为mask.nii.gz
Fig1.jpg 尝试提取影像学特征：
import radiomics
from radiomics import featureextractor
imageFile = "/Users/Mac/Documents/JianShuNotes/2D_Ultrasound/test.nii.gz"
maskFile = "/Users/Mac/Documents/JianShuNotes/2D_Ultrasound/mask.nii.gz"
extractor = featureextractor.RadiomicsFeatureExtractor()
featureVector = extractor.execute(imageFile, maskFile)
#print(featureVector.items())
for featureName in featureVector.keys():
print("%s: %s" % (featureName, featureVector[featureName]))
出现了如下报错信息：
sitk::ERROR: Pixel type: vector of 8-bit unsigned integer is not supported in 3D byN3itk6simple26LabelStatisticsImageFilterE
- 按照提示，有可能是数据格式的问题。回到源代码，修改文件相应的格式，将其命名为
test1.nii.gz
import SimpleITK as sitk
import numpy as np
folderPath = "/Users/Mac/Documents/JianShuNotes/2D_Ultrasound/"
reader = sitk.ImageSeriesReader()
dicom_names = reader.GetGDCMSeriesFileNames(folderPath)
reader.SetFileNames(dicom_names)
image = reader.Execute()
image_arr = sitk.GetArrayFromImage(image) # Note: order:z, y, x !!
size = image.GetSize()
origin = image.GetOrigin() #order: x, y, z
spacing = image.GetSpacing() #order:x, y, z
direction = image.GetDirection()
#print(spacing)
pixelType = sitk.sitkInt8 #注意这里是Int8
image_new = sitk.Image(size,pixelType)
#image_arr_new = image_arr[:,:,::-1] #镜像翻转操作
image_arr_new = image_arr
#print(image_arr_new.shape)
image_new = sitk.GetImageFromArray(image_arr_new)
image_new.SetDirection(direction)
image_new.SetSpacing(spacing)
image_new.SetOrigin(origin)
sitk.WriteImage(image_new,folderPath + "test1.nii.gz")
- 将步骤[2]中的
test.nii.gz替换为test1.nii.gz,再次执行该步骤
此时又出现了报错信息：
sitk::ERROR: Input "labelImage" for "LabelStatisticsImageFilter" has dimension of 3 which does not match the primary input's dimension of 2!
- 执行
print(image_arr_new.shape)后显示为：
(1, 900, 1600, 3)
Notes: 这里涉及python读入nifti文件的格式问题，其顺序为 z,y,x(对应这里的1,900,1600); 这里的3为不同的slice
- 探索这个“3”代表的意义：将每一个slice导出来，通过对比发现其中的端倪。执行代码：
image_arr_new = image_arr[:,:,:,0]
#image_arr_new = image_arr[:,:,:,1]
#image_arr_new = image_arr[:,:,:,2]
#print(image_arr_new.shape)
image_new = sitk.GetImageFromArray(image_arr_new)
image_new.SetDirection(direction)
image_new.SetSpacing(spacing)
image_new.SetOrigin(origin)
sitk.WriteImage(image_new,folderPath + "test-slice0.nii.gz")
#sitk.WriteImage(image_new,folderPath + "test-slice1.nii.gz")
#sitk.WriteImage(image_new,folderPath + "test-slice2.nii.gz")
- 将
test-slice0.nii.gz，test-slice1.nii.gz，test-slice2.nii.gz在软件中打开，查看它们之间的差异(肉眼似乎看不出啥区别) - 使用写代码的方法来探索
test-slice0.nii.gz，test-slice1.nii.gz，test-slice2.nii.gz之间的差异(视频里李博士演示了对比前二者)
import SimpleITK as sitk
import numpy as np
slicer0 = sitk.ReadImage("test-slice0.nii.gz")
slicer1 = sitk.ReadImage("test-slice1.nii.gz")
slicer0_arr = sitk.GetArrayFromImage(slicer0)
slicer1_arr = sitk.GetArrayFromImage(slicer1)
comp = slicer0_arr == slicer1_arr
print(comp)
Fig2.jpg
- 将差异的部分保存为影像导出
image_arr = slicer0_arr - slicer1_arr
image_arr[image_arr != 0] = 1
size = slicer0.GetSize()
origin = slicer0.GetOrigin() #order: x, y, z
spacing = slicer0.GetSpacing() #order:x, y, z
direction = slicer0.GetDirection()
image_new = sitk.GetImageFromArray(image_arr)
image_new.SetDirection(direction)
image_new.SetSpacing(spacing)
image_new.SetOrigin(origin)
sitk.WriteImage(image_new,"comp.nii.gz")
- 此时在软件中打开
comp.nii.gz文件，查看差异(在于右下角的水印部分)(视频中是左下角的水印)。换句话说，这3张test-slice0.nii.gz，test-slice1.nii.gz，test-slice2.nii.gz是一样的，差别仅在于右下角的水印。所以，后续的工作只需选择其中的一个即可。 - 再次查看
test-slice0.nii.gz和mask文件，发现mask文件和源文件不对应。mask.nii.gz是基于test.nii.gz文件画出并保存的，但是在test.nii.gz保存为test1.nii.gz时存在一个翻转(我没有搞明白这一点)，所以应重新保存一下mask文件
Fig3.jpg
import SimpleITK as sitk
import numpy as np
folderPath = "/Users/Mac/Documents/JianShuNotes/2D_Ultrasound/"
mask = sitk.ReadImage('mask.nii.gz')
mask_arr = sitk.GetArrayFromImage(mask)
reader = sitk.ImageSeriesReader()
dicom_names = reader.GetGDCMSeriesFileNames(folderPath)
reader.SetFileNames(dicom_names)
image = reader.Execute()
image_arr = sitk.GetArrayFromImage(image) # Note: order:z, y, x !!
size = image.GetSize()
origin = image.GetOrigin() #order: x, y, z
spacing = image.GetSpacing() #order:x, y, z
direction = image.GetDirection()
pixelType = sitk.sitkInt8 #注意这里是Int8
image_new = sitk.Image(size,pixelType)
mask_new = sitk.Image(size,pixelType)
#image_arr_new = image_arr[:,:,::-1] #镜像翻转操作
image_arr_new = image_arr[:,:,:,0]
print(image_arr.shape)
image_new = sitk.GetImageFromArray(image_arr_new)
image_new.SetDirection(direction)
image_new.SetSpacing(spacing)
image_new.SetOrigin(origin)
mask_new = sitk.GetImageFromArray(mask_arr) # 视频里勘误为mask_new,其实就应该是mask_arr, print二者可以看出区别来
mask_new.SetDirection(direction)
mask_new.SetSpacing(spacing)
mask_new.SetOrigin(origin)
sitk.WriteImage(image_new,"test0.nii.gz")
sitk.WriteImage(mask_new,"mask0.nii.gz")
Fig4.jpg
-
再次尝试提取影像组学特征(代码[2])，发现代码可以运行。但是这里有一个隐藏的bug，就是mask文件虽然重新保存了，但是并不能和原来的位置匹配。通过软件查看图像可以发现其中的区别
Fig5.jpg 再次回到原始代码里查找问题来源
import SimpleITK as sitk
import numpy as np
folderPath = "/Users/Mac/Documents/JianShuNotes/2D_Ultrasound/"
mask = sitk.ReadImage('mask.nii.gz')
mask_arr = sitk.GetArrayFromImage(mask)
#reader = sitk.ImageSeriesReader()
#dicom_names = reader.GetGDCMSeriesFileNames(folderPath)
#reader.SetFileNames(dicom_names)
#image = reader.Execute()
image = sitk.ReadImage('test.nii.gz')
image_arr = sitk.GetArrayFromImage(image) # Note: order:z, y, x !!
size = image.GetSize()
origin = image.GetOrigin() #order: x, y, z
spacing = image.GetSpacing() #order:x, y, z
direction = image.GetDirection()
pixelType = sitk.sitkInt8 #注意这里是Int8
image_new = sitk.Image(size,pixelType)
mask_new = sitk.Image(size,pixelType)
#image_arr_new = image_arr[:,:,::-1] #镜像翻转操作
image_arr_new = image_arr[:,:,:,0]
print(image_arr.shape)
image_new = sitk.GetImageFromArray(image_arr_new)
image_new.SetDirection(direction)
image_new.SetSpacing(spacing)
image_new.SetOrigin(origin)
mask_new = sitk.GetImageFromArray(mask_arr)
mask_new.SetDirection(direction)
mask_new.SetSpacing(spacing)
mask_new.SetOrigin(origin)
sitk.WriteImage(image_new,"test0.nii.gz")
sitk.WriteImage(mask_new,"mask0.nii.gz")
- 历尽周折，现在test0.nii.gz文件和mask0.nii.gz文件相匹配了！可以愉快地提取影像组学特征了
Fig6.jpg
import radiomics
from radiomics import featureextractor
imageFile = "/Users/Mac/Documents/JianShuNotes/2D_Ultrasound/test0.nii.gz"
maskFile = "/Users/Mac/Documents/JianShuNotes/2D_Ultrasound/mask0.nii.gz"
extractor = featureextractor.RadiomicsFeatureExtractor()
featureVector = extractor.execute(imageFile, maskFile)
#print(featureVector.items())
for featureName in featureVector.keys():
print("%s: %s" % (featureName, featureVector[featureName]))
Fig7.jpg