如何在python中将图像写为dicom？

Question

我试图从python中以dicom格式保存一组图像。我只会提取PixelData字段，因此我尝试使用最少量的代码/标记来完成。我正在使用pydicom阅读文件：

ds = dicom.read_file(currentDir + file)
Image = ds.pixel_array 

不幸的是，似乎没有一个函数dicom.write_file（filename，header = MakeOneUp）。你能否提出一种最简单的方法来将numpy数组写成dicom图像，而不依赖于读取dicoms /复制头文件。我试图用这个例子来做：

from dicom.dataset import Dataset, FileDataset

file_meta = Dataset()
file_meta.MediaStorageSOPClassUID = '1.2.840.10008.5.1.4.1.1.2'
file_meta.MediaStorageSOPInstanceUID = "1.2.3"
file_meta.ImplementationClassUID = "1.2.3.4"

ds = FileDataset(FullPath, {}, file_meta=file_meta, preamble=b"\0" * 128)

ds.PatientName = "Test^Name"
ds.PatientID = "00-000-000"

# Set the transfer syntax
ds.is_little_endian = True
ds.is_implicit_VR = True

# Set creation date/time
dt = datetime.datetime.now()
ds.ContentDate = dt.strftime('%Y%m%d')
timeStr = dt.strftime('%H%M%S.%f')  # long format with micro seconds
ds.ContentTime = timeStr

但是，当我到达图像时，它会崩溃：

ds.PixelData = Image[:,:,l]
# Alternatively 
# ds.pixel_array = np.ascontiguousarray(Image[:,:,l]), but this isn't a standard field?
ds.save_as(FullPath)

这失败并出现错误：如果我尝试这样做

np.ascontiguousarray(Image[:,:,l])

这不是dicom.read_file作为人类可读的数组。

谢谢

P.S：被要求进行错误追溯。

ds.PixelData = np.ascontiguousarray(Image[:,:,l]) 
ds.save_as(FullPath)

结果是：

Traceback (most recent call last):
  File "/home/m047659/Programs/anaconda3/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2910, in run_code
    exec(code_obj, self.user_global_ns, self.user_ns)
  File "<ipython-input-2-078b166fa765>", line 1, in <module>
    ds.save_as(FullPath)
  File "/home/m047659/Programs/anaconda3/lib/python3.6/site-packages/dicom/dataset.py", line 491, in save_as
    dicom.write_file(filename, self, write_like_original)
  File "/home/m047659/Programs/anaconda3/lib/python3.6/site-packages/dicom/filewriter.py", line 356, in write_file
    write_dataset(fp, dataset)
  File "/home/m047659/Programs/anaconda3/lib/python3.6/site-packages/dicom/filewriter.py", line 200, in write_dataset
    write_data_element(fp, dataset[tag], dataset_encoding)
  File "/home/m047659/Programs/anaconda3/lib/python3.6/contextlib.py", line 99, in __exit__
    self.gen.throw(type, value, traceback)
  File "/home/m047659/Programs/anaconda3/lib/python3.6/site-packages/dicom/tagtools.py", line 21, in tag_in_exception
    raise type(e)(err)
ValueError: Invalid tag (7fe0, 0010): ndarray is not C-contiguous

Answer 1

您需要为PixelData分配Image的字节字符串表示。 Image是一个numpy数组，需要转换为字节字符串。

ds.PixelData = Image.tobytes()

您还需要为pixel_array设置其他与图像相关的属性，如行，列，PixelRepresentation等，以便正确填充。见http://dicom.nema.org/medical/dicom/2014c/output/chtml/part03/sect_C.7.6.3.html#sect_C.7.6.3.1.3。