使用 numpy/vtk 實現 3d 圖像可視化

Question

我正在嘗試使用 numpy/vtk 顯示更多圖像（ct-scan），如本示例代碼（ http://www.vtk.org/Wiki/VTK/Examples/Python/vtkWithNumpy ）中所述，但我不明白它，不知道為什么。 如果有人可以幫助我，那就太好了。

這是我的代碼：

import vtk
import numpy as np
import os
import cv, cv2
import matplotlib.pyplot as plt
import PIL
import Image

DEBUG =True
directory="splitted_mri/"
w = 226
h = 186
d = 27
stack = np.zeros((w,d,h))

k=-1 #add the next picture in a differente level of depth/z-positions
for file in os.listdir(directory):
    k+=1
    img = directory + file
    im = Image.open(img)
    temp = np.asarray(im, dtype=int)
    stack[:,k,:]= temp
print stack.shape

#~ plt.imshow(test)
#~ plt.show()

print type(stack[10,10,15])

res = np.amax(stack)
res1 = np.amin(stack)
print res, type(res)
print res1, type(res1)

#~ for (x,y,z), value in np.ndenumerate(stack):
    #~ stack[x,y,z]=np.require(stack[x,y,z],dtype=np.int16)
    #~ print type(stack[x,y,z])

stack = np.require(stack,dtype=np.uint16)
print stack.dtype


if DEBUG : print stack.shape
dataImporter = vtk.vtkImageImport()
data_string = stack.tostring()

dataImporter.CopyImportVoidPointer(data_string, len(data_string))
dataImporter.SetDataScalarTypeToUnsignedChar()
dataImporter.SetNumberOfScalarComponents(1)
dataImporter.SetDataExtent(0, w-1, 0, 1, 0, h-1)
dataImporter.SetWholeExtent(0, w-1, 0, 1, 0, h-1)
essai = raw_input()
alphaChannelFunc = vtk.vtkPiecewiseFunction()
colorFunc = vtk.vtkColorTransferFunction()
for i in range (0,255):
    alphaChannelFunc.AddPoint(i, 0.9)
    colorFunc.AddRGBPoint(i,i,i,i)

volumeProperty = vtk.vtkVolumeProperty()
volumeProperty.SetColor(colorFunc)
#volumeProperty.ShadeOn()
volumeProperty.SetScalarOpacity(alphaChannelFunc)

# This class describes how the volume is rendered (through ray tracing).
compositeFunction = vtk.vtkVolumeRayCastCompositeFunction()
# We can finally create our volume. We also have to specify the data for it, as well as how the data will be rendered.
volumeMapper = vtk.vtkVolumeRayCastMapper()
volumeMapper.SetVolumeRayCastFunction(compositeFunction)
volumeMapper.SetInputConnection(dataImporter.GetOutputPort())

# The class vtkVolume is used to pair the preaviusly declared volume as well as the properties to be used when rendering that volume.
volume = vtk.vtkVolume()
volume.SetMapper(volumeMapper)
volume.SetProperty(volumeProperty)

# With almost everything else ready, its time to initialize the renderer and window, as well as creating a method for exiting the application
renderer = vtk.vtkRenderer()
renderWin = vtk.vtkRenderWindow()
renderWin.AddRenderer(renderer)
renderInteractor = vtk.vtkRenderWindowInteractor()
renderInteractor.SetRenderWindow(renderWin)

# We add the volume to the renderer ...
renderer.AddVolume(volume)
# ... set background color to white ...
renderer.SetBackground(1, 1, 1)
# ... and set window size.
renderWin.SetSize(400, 400)

# A simple function to be called when the user decides to quit the application.
def exitCheck(obj, event):
    if obj.GetEventPending() != 0:
        obj.SetAbortRender(1)

# Tell the application to use the function as an exit check.
renderWin.AddObserver("AbortCheckEvent", exitCheck)

#to quit, press q
renderInteractor.Initialize()
# Because nothing will be rendered without any input, we order the first render manually before control is handed over to the main-loop.
renderWin.Render()
renderInteractor.Start()

Answer 1

我終於發現出了什么問題，這是我的新代碼

import vtk
import numpy as np
import os
import matplotlib.pyplot as plt
import PIL
import Image

DEBUG =False
directory="splitted_mri/"

l = []

k=0 #add the next picture in a differente level of depth/z-positions
for file in os.listdir(directory):
    img = directory + file
    if DEBUG : print img
    l.append(img)
# the os.listdir function do not give the files in the right order 
#so we need to sort them
l=sorted(l)

temp = Image.open(l[0])
h, w = temp.size
d = len(l)*5 #with our sample each images will be displayed 5times to get a better view
if DEBUG : print 'width, height, depth : ',w,h,d

stack = np.zeros((w,d,h),dtype=np.uint8)

for i in l:
    im = Image.open(i)
    temp = np.asarray(im, dtype=int)
    for i in range(5):
        stack[:,k+i,:]= temp
    k+=5
    #~ stack[:,k,:]= temp
    #~ k+=1

if DEBUG :
    res = np.amax(stack)
    print 'max value',res
    res1 = np.amin(stack)
print 'min value',res1

#convert the stack in the right dtype
stack = np.require(stack,dtype=np.uint8)

if DEBUG :#check if the image have not been modified
test = stack [:,0,:]
plt.imshow(test,cmap='gray')
plt.show()

if DEBUG : print 'stack shape & dtype' ,stack.shape,',',stack.dtype

dataImporter = vtk.vtkImageImport()
data_string = stack.tostring()

dataImporter.CopyImportVoidPointer(data_string, len(data_string))
dataImporter.SetDataScalarTypeToUnsignedChar()
dataImporter.SetNumberOfScalarComponents(1)

#vtk uses an array in the order : height, depth, width which is 
#different of numpy (w,h,d) 
w, d, h = stack.shape
dataImporter.SetDataExtent(0, h-1, 0, d-1, 0, w-1)
dataImporter.SetWholeExtent(0, h-1, 0, d-1, 0, w-1)

alphaChannelFunc = vtk.vtkPiecewiseFunction()
colorFunc = vtk.vtkColorTransferFunction()
for i in range(256):
    alphaChannelFunc.AddPoint(i, 0.2)
    colorFunc.AddRGBPoint(i,i/255.0,i/255.0,i/255.0)
# for our test sample, we set the black opacity to 0 (transparent) so as
#to see the sample  
alphaChannelFunc.AddPoint(0, 0.0)
colorFunc.AddRGBPoint(0,0,0,0)

volumeProperty = vtk.vtkVolumeProperty()
volumeProperty.SetColor(colorFunc)
#volumeProperty.ShadeOn()
volumeProperty.SetScalarOpacity(alphaChannelFunc)

# This class describes how the volume is rendered (through ray tracing).
compositeFunction = vtk.vtkVolumeRayCastCompositeFunction()
# We can finally create our volume. We also have to specify the data for
# it, as well as how the data will be rendered.
volumeMapper = vtk.vtkVolumeRayCastMapper()
# function to reduce the spacing between each image
volumeMapper.SetMaximumImageSampleDistance(0.01)

volumeMapper.SetVolumeRayCastFunction(compositeFunction)
volumeMapper.SetInputConnection(dataImporter.GetOutputPort())

# The class vtkVolume is used to pair the preaviusly declared volume as 
#well as the properties to be used when rendering that volume.
volume = vtk.vtkVolume()
volume.SetMapper(volumeMapper)
volume.SetProperty(volumeProperty)

# With almost everything else ready, its time to initialize the renderer and window,
# as well as creating a method for exiting the application
renderer = vtk.vtkRenderer()
renderWin = vtk.vtkRenderWindow()
renderWin.AddRenderer(renderer)
renderInteractor = vtk.vtkRenderWindowInteractor()
renderInteractor.SetRenderWindow(renderWin)

# We add the volume to the renderer ...
renderer.AddVolume(volume)
# ... set background color to white ...
renderer.SetBackground(1, 1, 1)
# ... and set window size.
renderWin.SetSize(550, 550)
renderWin.SetMultiSamples(4)
# A simple function to be called when the user decides to quit the application.
def exitCheck(obj, event):
    if obj.GetEventPending() != 0:
        obj.SetAbortRender(1)

# Tell the application to use the function as an exit check.
renderWin.AddObserver("AbortCheckEvent", exitCheck)

#to auit, press q
renderInteractor.Initialize()
# Because nothing will be rendered without any input, we order the first
# render manually before control is handed over to the main-loop.
renderWin.Render()
renderInteractor.Start()

Answer 2

如果您對不使用 VTK 的解決方案沒有意見，您可以使用 Matplotlib imshow 和帶鍵的交互式導航。

本教程展示了如何： https : //www.datacamp.com/community/tutorials/matplotlib-3d-volumetric-data https://github.com/jni/mpl-volume-viewer

這里是查看 RTdose 文件的實現： https : //github.com/pydicom/contrib-pydicom/pull/19

另見： https : //github.com/napari/napari