Skip to content

Latest commit

 

History

History
152 lines (102 loc) · 4.14 KB

README.md

File metadata and controls

152 lines (102 loc) · 4.14 KB
Raw

DICraft

Convert DICOM files into voxel/pixelcloud, edit Minecraft style and 3D-print them.

This project tries to fill the gap between X-ray imaging (slices, single images, ultrasound images) and 3D-printing.
For medical and scientific use (you will find application cases).

Visit projekt page

DISCLAIMER

This "program" is under heavy development and it is NOT guaranteed that these instructions always fit the actual behavior!

note

Current workflow:

  • convert DICOM files into PNG files with an external software (my MRT-scans came with a software that could export series as PNG)
  • convert PNG files into a voxel file with png2save.py

in my example:

python png2save.py tmp/Seq_9_EP2D_DIFF_3_5MM_pnd savefile=bison_brain.sav minVal=50 maxVal=256 materialSwitch=4 cutRectangle=358,36,940,913 resize=256,256
python DICraft.py savefile=bison_brain.sav maxVoxels=1000000

Installation

Create a venv and install the requirements:

python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

If you want/have to convert DICOM files with convert.py that we can use them,
you need dcmj2pnm what is part of the dcmtk

    sudo apt-get install dcmtk

USE

At first, we have to convert some DCM files into an readable format.
This will convert ALL .dcm files under the given directory into the tmp directory.

python convert.py multiImageTest/ tmp/

With this command we create a "voxel" file our engine can work with.
This command will automatically overwrite "saves/quicksave.sav" or with the command switch "savefile="
you can name the destination file.

python dcm2save.py tmp/ savefile=roflcopter.sav

Finally we can start the main program (if you named your file, you have to tell the program)

python DICraft.py savefile=roflcopter.sav

From here you can save your work with F5 or export it for 3D printing with F6.

example

Placed my MRT scans in tmp/my/ and want to save the result in tmp/

python convert.py tmp/my/ tmp/
python dcm2save.py tmp/ savefile=bison_brain.sav minVal=10 maxVal=255 materialSwitch=10
python DICraft.py savefile=bison_brain.sav

Configuration

png2save

minVal = 130
maxVal = 134
materialSwitch = 15
maxVoxels = -1
resize = 256,256
cutRectangle = 398,69,878,851

cutRectangle = x,y,width,height

dcm2save

In dcm2save.py are 3 special configuration variables:

minVal = 130
maxVal = 134
materialSwitch = 15
maxVoxels = -1

minVal and maxVal are the raw gray values from the image, every value in between would be displayed later.
You HAVE to play with those values to get an accurate result (later there will be a GUI for that....).

materialSwitch is in which intervals another texture is picked.
You may have to play with it too.
The maximum value is 99.

maxVoxels is the maximum number of voxels to be processed. If you have a large dataset, you can limit the number of voxels to be processed.

All have command line switches too.

python dcm2save.py tmp/ savefile=bison_brain.sav minVal=10 maxVal=255 materialSwitch=10

Controls

Exit: ESC
Release mouse: F1

Save your work: F5
Export to OpenScad: F6

Movement: W A S D, UP(SPACE), DOWN(LEFT CTRL)
Remove block: left mouse button
Place block: right mouse button
Select "material": number 1 to 0
Reset position (in case of getting "lost"): R

Remove group of blocks:
focus a block, press "DEL" and all blocks that stick together are removed
WARNING
This may take "forever", be careful! USE WITH CARE

LEGACY

old stuff/for reference only

OLD installation

Debian based:

sudo apt-get install python-pyglet python-dicom python-qt4

Others "maybe":

sudo pip install pyglet python-dicom python-qt4

In some distributions python-dicom is named pydicom
The conversion parts should work without GUI, so you will need Qt4 ONLY when strating the GUI-Tools.

If you want/have to convert DICOM files with convert.py that we can use them,
you need dcmj2pnm what is part of the dcmtk

sudo apt-get install dcmtk