Hi 
I'm a new user to the MRIcroGL tool. it is awesome. However, I have some issues about use some functions of the tool in the python programming. Can I use the embedded scripting function in my python code such as the gl package (import gl)
I found the functions of special interest the tool bar -> the view -> "remove haze" and "remove haze and smooth edges" functions. how can I use them in the script of the tool? I have not found any information about that part in the help document.
Thanks very much

I am assuming you are using the latest version (1.2.20200331). The extract function will remove haze. A minimal script might look like this: 

The help for this function describes the arguments:
You can see all the in-built functions and their usage by running the Scripting/Templates/Help menu item. The other scripts in the Scripting/Templates menu describe the functions in more detail. The NITRC wiki also provides a brief introduction to scripting. Feel free to contribute to the Wiki if you want to improve it for others.

Hi Chris
Thanks very much for the reply. I also want to know how can I use the gl package in common python coding? (more specifically in the command line rather than the graphical interface (the View/Scripting menu item)) Can I pip install the package from Python Package Manager? or it that only the embedded package in the MRIcroGL tool. Thanks

MRIcroGL is a natively compiled executable using the native widgetset of your operating system (Windows API for Windows, QT or GTK2 for Linux and Cocoa for MacOS), with Harare accelerated graphics and compute using OpenGL (or optionally Metal for MacOS). Unlike FSLeyes, it is not written in Python, rather Python commands are interpreted (using PyArg_ParseTuple) as native commands. The benefit is that it is very fast, the disadvantage is that you can not run it directly from the Python command line. In this respect, it is like Blender, where you can launch it from the Python command line and have it load a Python script. If you want to handle NIfTI images from the Python command line, I would think FSLeyes would be a good choice. On the other hand, MRIcroGL is open source, so if you want to extend it, you could always add in hooks for IPC. If you want to develop this, create a fork of the Github repository, and when you are happy with your solution generate a pull request to share your solution with the community.

By the way, you can always use the Scripting Editor in MRIcroGL as a kind of command line, entering one or more commands at a time. Consider paste this script into the scripting editor and choose Control-R to run the command: You can now add one or more commands to change the image (e.g. new instructions are applied to the exisiting state machine). So you could run the command: And it will change image intensity without any other effects. You can add effects one at a time. The one special command is gl.resetdefaults(), which sets many settings back to their default settings. You only want to use this at the very start of a script to ensure that all scripts start with in a common state.

Originally posted by Chris Rorden:
Thanks so much for your reply Chris, 
I have tried this, it is very useful. Threrefore I want to use some tools inside the MRIcrogl. The "remove haze and smooth" is of special insterest to me and I want to incorporate the function in my current experiment pipeline coded with python (bash is also good). I tried with the "remove haze and smooth". It will help me to remove the bracing structures in the CT scans of head (the head fixer in CT machine appears in the CT scan).
Could you tell how do you implement that? Did you just use image process tool to keep the largest connected region in the image to achieve that?
Also it will be great if the FSLeye has that function, so I can directely use that tool to remove those unnecessary parts. 
Thanks so much //



window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());

gtag('config', 'UA-172659890-2');




// ]]>

Consider the command "Remove Haze with Options"

1. Segment the image with a multi-level Otsu's Method
    https://en.wikipedia.org/wiki/Otsu%27s_m...
    https://scikit-image.org/docs/dev/auto_e...
 The user selects a threshold from 1..5
  a. Level 1: Compute Otsu 4 level, choose darkest as "air"
  b. Level 2: Compute Otsu 3 level, choose darkest as "air"
  c. Level 3: Compute Otsu 2 level, choose darkest as "air" 
  d. Level 4: Compute Otsu 3 level, choose all but brightest as "air"
  e. Level 5: Compute Otsu 4 level, choose all but brightest as "air"
Air voxels are set to darkest value in volume.


2. (Optional) If the user selects "Only largest object", we compute connectivity based on 6 neighbors (e.g. to be part of a cluster, a neighbor must share a face, sharing an edge or corner does not count). Voxels that are not part of the largest cluster are set to the darkest value in the volume.

3. (Optional) If the user chooses to "Smooth Edges" a mask is generated that contains surviving voxels that are near air (as defined by 3 dilations). Voxels in this region are blurred with the neighbors. The goal here is to ensure that voxels near the air-brain surface do not appear jagged, while interior voxels retain their details. 

4. The size of the surviving clusters is dilated by two voxels, and the "air" voxels that are near the object get their original values. The goal here is to feather the edges and preserve partial volume effects.

The MRIcroGL scripting function "extract" can perform this. 
  extract (built-in function):
    extract(|b,s,t) -> Remove haze from background image. Blur edges (b: 0=no, 1=yes, default), single object (s: 0=no, 1=yes, default), threshold (t: 1..5=high threshold, 5 is default, higher values yield larger objects)

And it is easy to call MRIcroGL scripts from Python or BASH scripts. However, the current scripting language only has commands to save bitmaps, not modified NIfTI images (e.g. you can not invoke the File/SaveNIfTI menu item from a script). I will consider adding this to a future release. However, you can easily emulate this function using a Python script using the recipe I provide above.

Hi Chris 
Thanks, so the step 1 to step 4 are the pipeline how you implement the "remove haze and smooth edges" function. right?
So level means the threshold for the image. 
follow the https://scikit-image.org/docs/dev/auto_examples/segmentation/plot_multiotsu.html, I should first use the 1.5 as the threshold. the do the following computation? 
sorry I did not get  the  abcde part. do you mean use Level 1 step 1?  So in a, I compute the Otsu 4 levels (use 4 thresholds and set the darkest as air) b. step 2: use Otsu 3 levels (3 thresholds and set the darkest as air etc.) could you explain more detailed?

I think calling MRIcroGL scripts from Python or BASH scripts will also help me in some sense.
//





window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());

gtag('config', 'UA-172659890-2');






// ]]>

I assume you are using a recent version of MRIcroGL, e.g. v1.2.20200707. The "View" menu has two options, "Remove Haze" and "Remove Haze with Options". The latter shows an options window where you set the threshold level, whether to smooth edges, and whether to only extract the largest object. The former runs the routine with the default options: threshold level 5, YES for smooth edges and YES for only extract largest object.

 Step 1, abcde refer to the 5 possible threshold levels (1..5). Note that "Level 3" uses the classical Otsu method for segmenting the image into two classes (e.g. black and white, as described in the wiki page). The other options separate the image into more classes (the multi-Otsu described on the SciKit page). For example, consider a T1-weigthed scan where looking at a histogram reveals roughly three class of brightness: dark (hard bone, air, csf), medium (gray matter, muscle, veins, soft bone) and bright (fat, white matter). For such an image, a two-level Otsu method may not reliably find a satisfactory threshold for dark versus other tissue, while a three-level classification would be a better fit.

While setting air to have a uniform dark value is useful for visualization and leads to smaller compressed file sizes, you need to think carefully regarding whether this is a method that you want to use earlier in an image processing pipeline. Specifically, the algorithm may not distinguish between different classes of dark tissue (e.g. air, bone, css for T1 scan). Further, mixture of gaussian methods for segmentation may assume an unrealistically narrow variance for dark tissues. Additionally, homogeneity bias correction methods may be deprived of useful variations in intensity. In general, I would most processing tools have numerous implicit assumptions for how an MRI scan appears. Therefore, unless you carefully inspect your pipeline, I would restrict the "remove haze" step to the final visualization step.