yuta-hi / pycuda_drr Goto Github PK

when I specific mhd_filename with a path of dicom, an error occured like this:

Traceback (most recent call last):
  File "rendering.py", line 129, in <module>
    main()
  File "rendering.py", line 31, in main
    volume, spacing = load_image(mhd_filename)
  File "rendering.py", line 11, in load_image
    itkimage = sitk.ReadImage(filename)
  File "/home/huangtianbo/anaconda3/lib/python3.7/site-packages/SimpleITK/SimpleITK.py", line 8876, in ReadImage
    return _SimpleITK.ReadImage(*args)
RuntimeError: Exception thrown in SimpleITK ReadImage: /tmp/SimpleITK/Code/IO/src/sitkImageReaderBase.cxx:107:
sitk::ERROR: Unable to determine ImageIO reader for "/home/huangtianbo/zp/spine_direction_classification/template_expretiment/s1"

can you give more detail about the parameter...

Hi Yuta,
May I learn which algorithm you used while producing a DRR? Did you implement a algorithm from a research paper or book?
Thanks in advance

Hi,
I have the following questions to consult。

1. I tried my own CT data，but the resulting drr.mhd file cannot be converted into a drr.png image.
2. How many photos are the generated drr.mhd files?Is the three-dimensional image information? (1280x1280x3)
3. How to directly generate a DRR image from the source code, Just like in your Example.png.
   Looking forward to your reply. Many thanks！

Hi Yuta,
I implemented on a sample dataset and I got images as the attached picture. I have two additional questions:

1. I don't need the middle image, how can I stop producing the middle image?
2. My rotations are not what I wanted, How can I control the image rotations? For example I want first upside down, the third image as vertical position.

Thanks in advance
Best

Hi,
I tried rendering.py for a ct image (converted to mhd) and got an error like:
T_Nx4x4.shape: (1, 4, 4)
Result image shape: (1280, 1280, 3)
Traceback (most recent call last):
File "rendering.py", line 118, in
main()
File "rendering.py", line 60, in main
from chainer_bcnn.data import save_image
ModuleNotFoundError: No module named 'chainer_bcnn'

Could you help me regarding this issue?
Thanks

I need to get an x-ray as close to real size as possible. Does making SDD and SOD values the same is the right way or should I do something different?