Tracking and Measuring Nucleus Diameter per Frame



Hello everybody!
I am new to ImageJ and Macros, but I will need to use it everyday in the next months to quantify the live imaging data I am producing. At the moment, I have some videos of cells replicating and expressing two fluorescent proteins: one is fused to H2B so that to visualize the nucleus, the other is fused to a protein of interest in the cytosol.
What I want to do is to measure the diameter of the nucleus as a function of the presence of the other protein. Theoretically, I should come up with a dataset having this info for each frame of the video. Since I am producing a lot of videos, I want to automate the process by writing a script in Jython.

My answer is: which are the tools that would enable me to segment and measure the diameter of the nucleus? Can you suggest me a pipeline? I wanted to use TrackMate, but it actually doesn´t allow to measure the diameter for each nucleus, as it aproximates all of them with the same circle to track them.
Also, if you know of the existence of a Notebook where I can find functions doing more or less something similar to my task, please let me know! I have to learn a lot and I need a place to start :slight_smile:


Hey @Lucrezia

Welcome to ImageJ and the Forum!

Here are a few helpful links to get you started with ImageJ and Segmentation and Scripting:

Ok. I know this is a lot … so start with the workshops at least. :slight_smile: You can search the other links from there.

And - we can better help you come up with a workflow if you post some original datasets. Do you really need to track your cells? If you are only interested in nuclei diameter changes over time… might not need to track them. But you’ll need to look into connected-component tools (look at MorpholibJ)… I think? Let’s see what others think as well…

But for sure - this is enough to get you started!

eta :slight_smile:


Hi @Lucrezia

After learning how to segment your nuclei into binary images, you can try to measure the ‘feret’s diamater’ in each nuclear object. Just select it in ‘set measurements’ and include it in your eventual ‘Analyze particles’ workflow.

Good luck!


Thank you @etarena, I am going through all the tutorials you sent me, they´re great!
I tried some of them on my data and I am also trying to do what @Darren_Thomson kindly suggested me for the segmentation, but I have some issues since I apparently don´t have an even illumination. I tried to subtract the background, but it´s problematic, since the threshold should change for each frame of the stack.

How can I apply what I can see in the Segmentation tutorial, for instance, to each frame of the stack? Should I write a script where I apply the same function to each frame independently? Isn´t inefficient in terms of time?
Also, I tried the Weka Segmentation, which is supercool but seems not to work on my stack… and I couldn´t find any advice on the web.

I tried to upload an example, but I cannot see it when I publish the post