Hi Aura,
> These protrusions belong to 4 different geometrical categories
> (mutually exclusive), and can be classified given the combination of
> three geometrical parameters (diameter of the spine neck, diameter of
> the spine head and total length of the spine).
This sounds like a good possible fit with the Trainable Weka Segmentation
plugin:
http://fiji.sc/Trainable_Weka_SegmentationOut of the box, it would not use your criteria specifically. But it would
likely be able to automatically classify areas of your images into the four
desired categories, after some manual training.
If having the exact criteria you mention is very important, it would also
be possible to extend the Trainable Weka Segmentation plugin with new
features corresponding to your calculations, although it would be some
programming effort to do so.
Regards,
Curtis
On Thu, Jul 31, 2014 at 12:49 PM, auraf85 <
[hidden email]> wrote:
> Hi, I am very new to ImageJ so sorry if my question is not written
> properly.
>
> I am doing a study which requires me to analyse the shape of dendritic
> spines of neurons. I have tried to have a look on internet if anyone else
> has tried to semi-automatize this process with ImageJ but I haven't found
> anything yet.
>
> Trying to simplify the problem, I need a way for classifying protrusions
> arising from cell membrane.
> These protrusions belong to 4 different geometrical categories (mutually
> exclusive), and can be classified given the combination of three
> geometrical
> parameters (diameter of the spine neck, diameter of the spine head and
> total
> length of the spine).
> I have defined quantitative rules for this classification (something like
> "IF spine length > neck diameter AND head diameter < 2*neck diameter THEN
> the spine belongs to CATEGORY 1, ELSE ...the other rules"
> I was wondering if it is possible to write a program to make ImageJ perform
> the classification automatically (acquiring manually the input measures I
> need), and take counts of the number of different spines in each category.
>
> Is it possible at all? Could you give me some links / ideas I can take
> inspiration from? I am very happy to learn how to this if possible but I
> understand probably I need to start from the basic, so just a re-assurance
> whether this can be done or is impossible and some indications how to
> proceed would be really really appreciated.
>
>
>
>
> --
> View this message in context:
>
http://imagej.1557.x6.nabble.com/geometrical-shapes-categorization-tp5009001.html> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list:
http://imagej.nih.gov/ij/list.html>
--
ImageJ mailing list:
http://imagej.nih.gov/ij/list.html
Locked
