search

AIM-Harvard / pyradiomics

Open-source python package for the extraction of Radiomics features from 2D and 3D images and binary masks. Support: https://discourse.slicer.org/c/community/radiomics
http://pyradiomics.readthedocs.io/
BSD 3-Clause "New" or "Revised" License
1.18k stars 497 forks source link

GLCM - Information Measures of Correlation? #400

Closed ejulrich closed 6 years ago

ejulrich commented 6 years ago

I wanted to find the background of the Gray Level Co-occurrence Matrix "Informal Measures of Correlation", and I had a lot of trouble finding the term. Were these supposed to be called the "Information Measures of Correlation"? [1]

[1] Haralick, et al. Textural features for image classification. (1973).

fedorov commented 6 years ago

@ejulrich I think you are right. IMC1 and IMC2 seem to correspond to First and Second measures of information correlation, as defined in 3.6.24 and 3.6.25 of the IBSI v6 document. I think this is a bug. Let's wait for @JoostJM to get back to address this.

fedorov commented 6 years ago

And I can't figure out why the equations are not rendered correctly for those in readthedocs!

https://pyradiomics.readthedocs.io/en/latest/features.html?highlight=imc#radiomics.glcm.RadiomicsGLCM.getImc1FeatureValue

JoostJM commented 6 years ago

The name is indeed a bug, it should be "informational measure of correlation". I'll also take a look at why the docs do not render correctly...

JoostJM commented 6 years ago

And I can't figure out why the equations are not rendered correctly for those in readthedocs!

There is no error in the formula, but a rather strange behaviour caused by the fact that "IMC" is highlighted...

JoostJM commented 6 years ago

Typo fixed in 4ba909a.