Automatic counting of nuclear envelope invaginations

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
LP
Reply | Threaded
Open this post in threaded view
|

Automatic counting of nuclear envelope invaginations

LP
Dear ImageJ comunity

I am attempting to write a macro that would allow me to count automatically
tubular invaginations of the nuclear envelope. However, I have no success. I
tried to use the 3D object counter plugin and the Trainable Weka
Segmentation, but I could't make anything useful. The issue being that the
invaginations are continuous with the rest of the nuclear envelope.

Below are 3 montages of 3 cells with invaginations. Original Z-stacks
available in the link below.
https://drive.google.com/open?id=0B2u4cC3n7OtFRjRFTTVYTG9sNlU
<https://drive.google.com/open?id=0B2u4cC3n7OtFRjRFTTVYTG9sNlU>  

Does anyone have an idea of how I could count these invaginations
automatically ?
The criteria being that the invagination needs to be visible on at least one
plane or fragment that is not a cresting section, or if it is visible in
cresting sections only it should be a much brighter spot than the
surroundings (such as the bottom invagination of the second montage) and
visible on multiple planes.

Thank you very much for your help,
<http://imagej.1557.x6.nabble.com/file/t381377/montage_2_long_invaginations_1_short.jpg>
<http://imagej.1557.x6.nabble.com/file/t381377/montage_3_long_invaginations_1_short.jpg>
<http://imagej.1557.x6.nabble.com/file/t381377/montage_4_short_invaginations.jpg>




--
Sent from: http://imagej.1557.x6.nabble.com/

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: Automatic counting of nuclear envelope invaginations

Thomas Boudier-4
Dear LP,

I think the idea is to first detect the nucleus and fill the holes
inside. Then you may need to erode the nucleus to remove the outer mb
staining.

Then you can detect signal inside this eroded nucleus and count the
objects. I used this approach and could detect 2 tubes in first image, 3
in second but third image was too difficult because the tubes are too
close to the mb. I used TANGO as a framework to do this (see result in
attached image).

Hope this helps,

Best,

Thomas


On 22/10/2017 01:10, LP wrote:

> Dear ImageJ comunity
>
> I am attempting to write a macro that would allow me to count automatically
> tubular invaginations of the nuclear envelope. However, I have no success. I
> tried to use the 3D object counter plugin and the Trainable Weka
> Segmentation, but I could't make anything useful. The issue being that the
> invaginations are continuous with the rest of the nuclear envelope.
>
> Below are 3 montages of 3 cells with invaginations. Original Z-stacks
> available in the link below.
> https://drive.google.com/open?id=0B2u4cC3n7OtFRjRFTTVYTG9sNlU
> <https://drive.google.com/open?id=0B2u4cC3n7OtFRjRFTTVYTG9sNlU>
>
> Does anyone have an idea of how I could count these invaginations
> automatically ?
> The criteria being that the invagination needs to be visible on at least one
> plane or fragment that is not a cresting section, or if it is visible in
> cresting sections only it should be a much brighter spot than the
> surroundings (such as the bottom invagination of the second montage) and
> visible on multiple planes.
>
> Thank you very much for your help,
> <http://imagej.1557.x6.nabble.com/file/t381377/montage_2_long_invaginations_1_short.jpg>
> <http://imagej.1557.x6.nabble.com/file/t381377/montage_3_long_invaginations_1_short.jpg>
> <http://imagej.1557.x6.nabble.com/file/t381377/montage_4_short_invaginations.jpg>
>
>
>
>
> --
> Sent from: http://imagej.1557.x6.nabble.com/
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
>
--
  /***************************************************************/
       Thomas Boudier, Associate Professor, UPMC,
       Universit√© Pierre et Marie Curie, Paris, France.
       BioInformatics Institute (BII)/IPAL, Singapore.
/**************************************************************/


--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

Snapshot.png (71K) Download Attachment