3D Filaments

[NicoPiC]

Hi everyone,

I 'd like to know if there is something in the icy development about filaments for cilia, tubules or neurones 3D studies ?

We actually used the expensive-bugged-Imaris software (filament tracer) and I looking for more solution i'd tried to use the imageJ fiji solution but its not compatible with our needs :

Is there a way to do some manual parts of filaments (by moving Z section or orthogonal sections, link them automatically (as it have been done chronologically) or manually and measure it, export it as ROI etc ?

For example its possible to link some points for the tracking is there a way to do it for Z instead of T (I have tried it with manual tracking but its not possible to "go back" in  T (only Doc or Marty can do it ;)) )

Best,
Nicolas

[Stephane]

Hi Nicolas,

There is different segmentations and filtering tools available in Icy and that might be a good start to detect the filaments structures (this will give you some 3D ROI for the filaments). You can also try to use the tracking tools by changing your sequence to a time sequence temporary (you can do it easily in the "Sequence operation" tab) then exploit tracks as filament instead (just need to consider the T dimension as Z). That's just some ideas... hope that helps.

Best,

- Stephane

[NicoPiC]

Thanks Stephane,

I have done the tracking (with manual tracking) and it work till your path only "goes into deep"  (as your z become t you can only add some time)
For the segmentation, in my case its a tube that's why I can't use the segmentation.
I have tried the segmented line for each z section but it not possible to link them together

In my opinion the way it link all the "clic point" together like a tracking could be an interesting way to do it

Maybe what I'll say is stupid but for example just clic the path you want to follow (it add X Y Z and T coordinate) and then try to link them with the shortest way like that even if you don't move in Z or go up or down it doesn't change anything the only things to do is to keep a "chronological" way to clic. And why not at least treat it like the tracking segment (for linking them for example etc.) and use a plug-in like track painter to show it ... I'm not enough good to write something like that but maybe it can give some ideas.


Best,
-Nicolas

[Stephane]

I think i see the idea, we don't exactly have that but something close :
http://icy.bioimageanalysis.org/plugin/ROI_Mover

It allow you to create 3D ROI slice by slice, by default it duplicates ROI from previous slice then you can modify it if needed. Is this something which can help you ?

Best,

- Stephane

[NicoPiC]

I tried it but for an accumulation of 2D area it could be very interesting (but I don't see how to merge each Z sections en a entire 3D volume I'll try it) but for filaments it didn't work with me ...

I think I'll come on Wednesday morning to an Icy club (I'm a neighbour, I work at Necker institute ;)) and show you my problem maybe it could be easy to solve with real examples, is there a specific week to come or next week is possible ?

Best,
- Nicolas

[Stephane]

I think the plugin allows you to create 3D Roi this way but it won't really help for filament indeed. I won't be there for some weeks but i guess you will always find someone ready to help you in the labs so you can come next Wednesday. For your problem i believe Alexandre could help you the best ;)

- Stephane

[NicoPiC]

Thanks a lot Stephane,

I will harass Alexandre I think an evolution of its manual tracking could help a lot ;)

Thanks again for your help and your "tolerance" ;) Stephane

Best,
- Nicolas

[Stephane]

You're welcome ;)

- Stephane