Overview
Read IMOD model files as pandas dataframes in Python.
Usage
import imodmodel
df = imodmodel.read('my_model_file.mod')
In [3]: df.head()
Out[3]:
object_id contour_id x y z
0 0 0 64.333336 64.666664 80.0
1 0 0 47.000000 77.333336 80.0
2 0 0 51.333332 45.666668 80.0
3 0 0 87.333336 49.666668 80.0
4 0 0 76.000000 82.000000 80.0
Slicer angles saved in the slicer window are stored in the IMOD binary file with both centerpoints and angles.
These annotations can be read in by setting annotation='slicer_angle' when calling imodmodel.read()
import imodmodel
df = imodmodel.read('file_with_slicer_angles.mod', annotation='slicer_angles')
In [3]: df.head()
Out[3]:
object_id slicer_angle_id time x_rot y_rot z_rot center_x center_y center_z label
0 0 0 1 13.100000 0.0 -30.200001 235.519577 682.744141 302.0
0 0 1 1 -41.400002 0.0 -47.700001 221.942444 661.193237 327.0
0 0 2 1 -41.400002 0.0 -41.799999 232.790726 671.332031 327.0
0 0 3 1 -35.500000 0.0 -36.000000 240.129181 679.927795 324.0
ImodModel
The resulting dataframe from imodmodel.read() contains only information about the contours or slicer angles.
The full set of information from the imod model file can be parsed using ImodModel
from imodmodel import ImodModel
my_model = ImodModel.from_file("my_model_file.mod")
in [3]: my_model.model_field_set
out[3]:
{'id', 'extra', 'objects', 'slicer_angles', 'header'}
my_model.id
my_model.id contains the IMOD file id and the version id
in [4]: my_model.id
out[4]:
ID(IMOD_file_id='IMOD', version_id='V1.2')
my_model.header
my_model.header is contains the model structure data mainly used by IMOD.
in [5]: my_model.header
out[5]:
ModelHeader(name='IMOD-NewModel', xmax=956, ymax=924, zmax=300, objsize=3, flags=62976, drawmode=1,
mousemode=1, blacklevel=145, whitelevel=173, xoffset=0.0, yoffset=0.0, zoffset=0.0, xscale=1.0, yscale=10,
zscale=1.0, object=2, contour=-1, point=-1, res=3, thresh=128, pixelsize=1.9733333587646484, units=-9,
csum=704518946, alpha=0.0, beta=0.0, gamma=0.0)
my_model.objects
my_model.objects is a list IMOD objects.
in [6]: my_model.objects[0].header
out[6]:
ObjectHeader(name='', extra_data=[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], contsize=0,
flags=402653184, axis=0, drawmode=1, red=0.0, green=1.0, blue=0.0, pdrawsize=0, symbol=1, symsize=3,
linewidth2=1, linewidth=1, linesty=0, symflags=0, sympad=0, trans=0, meshsize=0, surfsize=0)
This is where object values like contours, meshes, and IMAT information are located.
in [7]: my_model.objects[1].meshes[0].indices
out[7]:
array([[38, 40, 52],
[38, 52, 50],
[50, 52, 64],
[50, 64, 60],
...,
[ 4, 10, 26],
[ 4, 26, 20],
[20, 26, 38],
[20, 38, 32]])
in [8]: my_model.objects[1].imat
out[8]:
IMAT(ambient=102, diffuse=255, specular=127, shininess=4, fillred=0, fillgreen=0, fillblue=0,
quality=0, mat2=0, valblack=0, valwhite=255, matflags2=0, mat3b3=0)
in [9]: my_model.objects[1].contours[0].points
out[9]:
array([[367.00006104, 661.83343506, 134. ],
[415.66674805, 667.83343506, 134. ],
[474.33340454, 662.50012207, 134. ]])
my_model.slicer_angles
my_model.slicer_angles is a list of slicer angles.
in [10]: my_model.slicer_angles[0]
out[10]:
SLAN(time=1, angles=(0.0, 0.0, 0.0), center=(533.5, 717.0, 126.0), label='\x00')
That's it!
Installation
imodmodel can be installed from the Python Package Index (PyPI)
pip install imodmodel
We recommend installing into a clean virtual environment.