Recover intensity of cells and not a disk with "Track Processor Intensity Profile"

[louise-laure]

Hi Nicolas & Fabrice,

Thanks for these plugins, that can save us a lot of time,

We've tried to use it to track some intensity of cells over time but could you help in a difficulty encountered?

1: First I detect the cells through Spot detector wich finds them with there shape in 3D,
=> could the plugin spot detector be updated so that each spot could have a label (in order to find back each spot that is being tracked) just as the ROIs?
=> can an extra spot (or ROI) be added manually (for instance a rectangle) so that it could be added in track manager with the other spots to be also measured over time?

2: Then I use your plugin "Track Processor Intensity Profile" to track their intensity over time but the problem is you only ask for a diameter around the barycenter,
when it would be great to only have the intensity measured in the shape of the spot detected (which, in my case, is a neuron with the shape of a "blob" more than a sphere)

I've seen the same kind of question in another part of the forum so I think it would be very usefull for many users :-)
Thanks in advance for your much needed help here ++,
this could let me analyze all my data this summer !

Best,
LL

[Fab]

Dear LL,

I added a label for each detection in the Spot detector. This is an option that you can activate in the Display tab. You can also display the font size of the text.

=> can an extra spot (or ROI) be added manually (for instance a rectangle) so that it could be added in track manager with the other spots to be also measured over time?

This point is also discussed in other threads. I will see what I can do about this.

2: Then I use your plugin "Track Processor Intensity Profile" to track their intensity over time but the problem is you only ask for a diameter around the barycenter,
when it would be great to only have the intensity measured in the shape of the spot detected (which, in my case, is a neuron with the shape of a "blob" more than a sphere)

This plugin ( http://icy.bioimageanalysis.org/plugin/Track_Processor_Intensity_Profile ) is managed by Nicolas who is very busy at the moment. So please be patient for this feature ;)

Best,

Fabrice

[louise-laure]

Hi,
thanks for adding this labeling option on spot detector, it's really helpful!
and also for looking into the manually added extra ROI because it's very much need when you do FRET imaging.

I hope Nicolas will have some spare time at some point.

Now I have a little extra problem,
could it be possible to match the number/label of the tracks to the number/label of the spots found in the first frame
(because now, I have to look manually at each track to know to which spot it belongs...)

Thanks again for you help,
LL

[louise-laure]

Hello Nicolas & Fabrice,
I hope you enjoyed your summer,
I'm asking again for some help,

Fabrice,
would you have time to look into the spot detector plugin adding the option for the manually added extra ROI for each plane/frame because it's very much needed when you do FRET imaging to calculate the (spot intensity - background intensity).

Nicolas,
did you ever have my post about changing the option for the plugin intensity profile?
Indeed, the cells we study are in 3D and they aren't all with the same diameter,
so, when we select a pre-defined diameter for the disk we don't measure the real intensity of our cells because they're all different in size and shape.
Could it be possible to import the shape of the object detected by spot detector (as on the binary image) and then calculate the intensity of each cell instead of a disk around the center of mass?

Thank you again for your very needed help guys,

If you need me to bring some images to help you change your plugins,
please let me know and I can drop by on any wednesday to the Icy club,

LL

[Fab]

On Tuesday, September 3, 2013 5:18:09 PM UTC+2, louise-laure wrote:

Hello Nicolas & Fabrice,
I hope you enjoyed your summer,
I'm asking again for some help,

Fabrice,
would you have time to look into the spot detector plugin adding the option for the manually added extra ROI for each plane/frame because it's very much needed when you do FRET imaging to calculate the (spot intensity - background intensity).

Hi Louise Laure,

Things are not very clear to me:

- If you wish to create an ROI somewhere in the image, and get its mean/max/min intensity over time, you can use several plugins to do so, I think the best is the one of Nicolas (I rated it 5 stars):

http://icy.bioimageanalysis.org/plugin/ROI_Intensity_Evolution

To use it, just type "intensity evolution" in the search bar of icy and pick the result.

I don't see why you would like to integrate this into the tracking. Do you wish to add manually some detection ?

Best,

Fabrice
 

[louise-laure]

Thanks Fabrice,
I was aiming at following the ROI intensity over time in each plane of the stack but in fact I don't need to
(that's why it took me a little time to answer)
I will deal with the other plugin.

Do you know if Nicolas has had a chance to look into the monitoring of the intensity measured in the shape of the spot detected (which, in my case, is a neuron with the shape of a "blob" not a predefined a disk) or at least the shape of the spot in one plane and not a predefined diameter?

Thanks, I'm almost able to do my image analysis now :-)
LL

[louise-laure]

Hello,
just wondering,

I saw in the intensity profile plugin in track manager,
there is a new option "Gaussian fit" instead of disk,
but no explanation...
and it doesn't measure mean intensity but "amplitude" or "background"

Could you tell us exactly what it does?
does it measure the amplitude intensity (max - min) but where?

Thanks in advance for the explanation,
LL

[nicolas]

Dear Louise Laure,

I just uploaded a new version of the processor. You can find a 'spot mask' analysis. It computes intensity indicators over the segmentation mask provided by the spot detector plugin, and should work in 3D, even though I don't have good images to test it.
This will only work if you use the combination spot detector + probabilistic particle tracker.

Please try it and come back to me if you have an issue.

Regarding the Gaussing fit: it can fit the amplitude of a Gaussian shape at the location of the detection. The computed value of the background is also provided. The standard deviation (size) of the Gaussian shape is user-defined. Unfortunately I have also implemented this solution in 2D. Please don't use 3D data.

Sorry for the delayed answer.

Nicolas

[louise-laure]

So cooooooollll!!!!!

This is exactly what I needed,

thanks again Nicolas for modifiying the intensity track manager plugin
but I haven't been able to test it yet though
because there seems to be a big bug in the multiple motion model of probabilistic particle tracker.

I'll try it ASAP and let you know how it goes,
btwy,
Iif you need some more images to test your plugin, I can provide you some, let me know.

Thanks again !
LL

[louise-laure]

Hello again,
actually I've tried,
even with using the Probabilistic particle tracker with simple motion => which seems to work ok

but then there is a bug any time I try to launch the Track Intensity plugin
(it's when it's starts, so wether I try the DISK option or the SPOTMASK option, it won't run)

java.lang.NullPointerException
    at plugins.nchenouard.trackprocessorintensityprofile.TrackProcessorIntensityProfile.fillSeries(TrackProcessorIntensityProfile.java:1280)
    at plugins.nchenouard.trackprocessorintensityprofile.TrackProcessorIntensityProfile$FillSeriesThread.run(TrackProcessorIntensityProfile.java:995)
java.lang.NullPointerException

here is the message I get,
Did I do smthg wrong or is it the plugin that crashes ?

LL

[nicolas]

Hi Louise Laure,

I just uploaded a bug fix for IMM in 3D. I will look at the pb with the intensity profile in 3D.

Best,

nicolas

[louise-laure]

Thanks,

Well,
maybe I'm not doing it properly,
but I still get the "submatrix indices" bug message...

So I guess the bug fix hasn't worked

LL (just trying to give some feedback)

[nicolas]

My bad. Please update to the latest version which includes the bug fix. It should work now.

Best,

Nicolas

[louise-laure]

Cool the probabilistic particle tracker works now :-)
(though you seem to have change the default settings of covariance for instance,
the trajectories on the same sequences with the same settings seem to have changed a bit...but don't seem wrong either.
Is there a reason for changing those default settings?)

Thanks for looking at the intensity profile problem.
If I just may add, the plugin doesn't launch at all so it's not a problem with a specific option

Good luck!
LL