Using ROI MANAGER

bonej
thresholding
segmentation
Tags: #<Tag:0x00007fd69c342c98> #<Tag:0x00007fd69c342b58> #<Tag:0x00007fd69c3429c8>

#1

Hi everyone, I’ve got a simple question about ROI. Using a specific thresholding for binarizing my images is the ROI automatically chosen ? And if yes, how could I see the value so that I can simulate it manually? Thank you in advanced.


#2

Good day!

And if yes, how could I see the value so that I can simulate it manually?

Which value?

ROI = “region of interest”

Regards

Herbie


#3

I need to know what is the Region of interest when I am doing a binarization of some images. I guess the ROI changes for each kind of thresholding (isodata, otsu, etc…). Am I wrong?


#4

@Student95

Welcome to the Forum!

To save your selections to the ROI Manager - you can Add to Manager (Analyze▷Tools▷ROI Manager…) directly or use the Analyze Particles interface by selecting the “Add to Manager” option.

I’m assuming you want to Measure different parameters of your selections? To look at the full list of measurements possible, you can see Set Measurements.

As you can see - I’ve inserted a few links to the ImageJ User Guide - this a good place to get started. There are also a host of great resources on the ImageJ wiki - specifically in the Learn section.

If you have more specific questions - feel free to reply or post again.

eta :slight_smile:


#5

Am I wrong?

No, but a ROI has no value, except an index in the ROI-manager.

To make the ROI visible after thresholdiing, you may use “Edit >> Selection >> Create Selection”.

But perhaps I still don’t understand what you are looking for.

Regards

Herbie


#6

Perhaps you can post an example image and let us know what you aim to measure? What is the current workflow that you are applying? Tell us the steps you are using at the moment. That will better assist us in helping you…

For thresholding … here are some helpful links to get you up-to-speed:

eta


#7

I definitely expressed myself in the wrong way. I am working with some images and I am trying to find some parameters with BoneJ but my first step is to binarize the images. According to the ROI I choose I get a value of these parameters near to the literature. So I was wondering if the only way to get these parameters correctly is to choose a ROI manually, or better to make a selection and then add it to the ROI MANAGER.


#8

Yes exactly. I am working with some images and need to find some parameters using BoneJ. The problem is that I need to know if the only way to get these parameters’ value near to the literature is by adding a selection to the ROI Manager manually. Thank you so much, by the way, for the earlier answers.


#9

Well, that sounds indeed differently.

I have no experience with BoneJ, so you need to ask BoneJ-experts and perhaps someone can add the BoneJ-topic to this thread.

Regards

Herbie


#10

@mdoube or @rimadoma - you can better assist this user regarding their analysis using BoneJ… :slight_smile:

Saying that… @Student95 - it would be really beneficial for them helping you if you can clearly state exactly what you wish to measure. Start with posting an original image as an example and explain clearly what you aim to segment, measure, etc. Typically - using manual segmentation/thresholding can introduce bias and it is recommended to use “auto” methods when possible to keep your analysis impartial and robust. (The Principles page of the ImageJ wiki is worth a look-see - especially this section.)

:slight_smile:

eta


#11

BoneJ understands different kinds of ROI depending on the command you are using.

Slice Geometry can be restricted to a rectangular ‘box’ ROI (the button on far left of the ImageJ window).

Thickness can be restricted to 3D freehand ROIs, which are actually just a series of 2D ROIs added to the ROI manager.

I think you need to clarify in your mind the difference between segmented phases in the image (i.e. the black and white pixels in a binarised image), and ROIs (pixels contained within an outline), and how these are represented in ImageJ. The other responders in this thread have pointed you towards some good resources to study these ideas.


#12

Thanks for your replying. I’m trying to get more into BoneJ and ImageJ.