Quantification of axon Twist

[Eric Denarier]

Hi All,

Is there a mean to Characterize/Quantify the behaviour of axons having a
tendency to twist in a clockwise orientation (see picture).
Thanks for your help


--

Eric Denarier
Grenoble Institut des Neurosciences
Inserm U1216
Chemin Fortuné Ferrini
38700 La Tronche
France

Tél :[phone]

http://neurosciences.ujf-grenoble.fr/


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

[Herbie]

Dear Eric,

what an impressive image!
(Cortical fiber layer?)

Attached please find the orientation salience function of the central
disc-shaped area (diameter 917pel) of your image. It shows that
horizontal structures (0deg) are dominant and the remaining orientations
(increasing clockwise) are about equally distributed.
(A (semi-)circular line would show as a straight line in the diagram.)

Please note that there is a grid-shaped shadow on your image that may
influence any analysis.

Hope it helps a bit

Herbie

:::::::::::::::::::::::::::::::::::::::::::
Am 23.03.17 um 08:57 schrieb Eric Denarier:
> Hi All,
>
> Is there a mean to Characterize/Quantify the behaviour of axons having a
> tendency to twist in a clockwise orientation (see picture).
> Thanks for your help
>
>


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

[Eric Denarier]

Hi Herbie,

Thanks for you help. Image is  actually Dorsal Root Ganglia neurons
grown on coverslip (part of the image).

How did you get the the orientation Salience ? With imageJ ?

Eric Denarier
Grenoble Institut des Neurosciences
Inserm U1216
Chemin Fortuné Ferrini
38700 La Tronche
France

Tél :[phone]

http://neurosciences.ujf-grenoble.fr/

Le 23/03/2017 à 10:12, Herbie a écrit :
> orientation salience function

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

[Herbie]

Dear Eric,

thanks for your explanation concerning the source of your image. So the
neurons are grown on glass...

re: Orientation Salience Function

Here are links for two papers dealing with theoretical and algorithmic
aspects. If you don't have much time, please try to understand at least
the first two pages of the report (2013).

<http://www.gluender.de/Writings/WritingsTexts/WritingsDownloads/1986_IntegralsCentralSlices.zip>
<http://www.gluender.de/Writings/WritingsTexts/WritingsDownloads/2013_Orientation.zip>

In fact Orientation Salience Functions are weighted histograms and could
be normalized as such. Weighted, because they don' reflect the pure
geometric property of orientation but the contrast/brightness of the
image structures as well.

With your image, it is interesting that the 90deg bending of the axons
shows up nicely by the decline of the function for angles greater 90deg.
The bending itself is reflected by the plateau between 20deg and 90deg.
Beyond 90deg the orientations are less regular.

The generation of the Orientation Salience Function is done with ImageJ
by use of a dedicated plugin (cf. the 2013 report) and a macro.

I shall write you off-list tomorrow for more details.

Best

Herbie

:::::::::::::::::::::::::::::::::::::::::::
Am 23.03.17 um 12:58 schrieb Eric Denarier:

> Hi Herbie,
>
> Thanks for you help. Image is  actually Dorsal Root Ganglia neurons
> grown on coverslip (part of the image).
>
> How did you get the the orientation Salience ? With imageJ ?
>
> Eric Denarier
> Grenoble Institut des Neurosciences
> Inserm U1216
> Chemin Fortuné Ferrini
> 38700 La Tronche
> France
>
> Tél :[phone]
>
> http://neurosciences.ujf-grenoble.fr/
>
> Le 23/03/2017 à 10:12, Herbie a écrit :
>> orientation salience function
>
>

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

[Herbie]

Dear Eric,
dear list,

there were two related errors in my previous posts.

1.
The orientation angle changes _counter_ clockwise.

2.
Consequently my interpretation

"With your image, it is interesting that the 90deg bending of the axons
shows up nicely [...]"

is wrong!
_________________________________________

re 2.

It turns out that the plateau between 25deg and 90deg is caused by a
very low frequency modulation in the image. It can best be seen in the
power spectrum. Maybe it is an illumination artifact.

Using a high-pass filter to remove the low frequency component leads to
the attached result which appears sound. (The local maximum at 90deg is
due to the grid-like shadow in the image.)

Best

Herbie

:::::::::::::::::::::::::::::::::::::::::::

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

[Straub, Volko A. (Dr.)]

I noticed previously a 'Directionality' plugin in ImageJ that may be of use (https://imagej.net/Directionality). I only looked at it briefly a while ago and don't have any real experience with it, but it looks like it might be useful for your purpose.

Volko

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Herbie
Sent: 23 March 2017 19:00
To: [hidden email]
Subject: Re: Quantification of axon Twist

Dear Eric,

thanks for your explanation concerning the source of your image. So the neurons are grown on glass...

re: Orientation Salience Function

Here are links for two papers dealing with theoretical and algorithmic aspects. If you don't have much time, please try to understand at least the first two pages of the report (2013).

<http://www.gluender.de/Writings/WritingsTexts/WritingsDownloads/1986_IntegralsCentralSlices.zip>
<http://www.gluender.de/Writings/WritingsTexts/WritingsDownloads/2013_Orientation.zip>

In fact Orientation Salience Functions are weighted histograms and could be normalized as such. Weighted, because they don' reflect the pure geometric property of orientation but the contrast/brightness of the image structures as well.

With your image, it is interesting that the 90deg bending of the axons shows up nicely by the decline of the function for angles greater 90deg.
The bending itself is reflected by the plateau between 20deg and 90deg.
Beyond 90deg the orientations are less regular.

The generation of the Orientation Salience Function is done with ImageJ by use of a dedicated plugin (cf. the 2013 report) and a macro.

I shall write you off-list tomorrow for more details.

Best

Herbie

:::::::::::::::::::::::::::::::::::::::::::
Am 23.03.17 um 12:58 schrieb Eric Denarier:

> Hi Herbie,
>
> Thanks for you help. Image is  actually Dorsal Root Ganglia neurons
> grown on coverslip (part of the image).
>
> How did you get the the orientation Salience ? With imageJ ?
>
> Eric Denarier
> Grenoble Institut des Neurosciences
> Inserm U1216
> Chemin Fortuné Ferrini
> 38700 La Tronche
> France
>
> Tél :[phone]
>
> http://neurosciences.ujf-grenoble.fr/
>
> Le 23/03/2017 à 10:12, Herbie a écrit :
>> orientation salience function
>
>

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

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

[Herbie]

Good day Volko,

the Directionality-plugin provides two approaches.

One that is strongly related to what I've suggested but, as mentioned
several times already on this list and on the forum, badly implemented
(sorry Jean-Yves), obviously due to lacking theoretical background.

The other is based on the (local) "structure tensor"-analysis as does
the much more refined plugin OrientationJ.

The decision, whether the second approach leads to satisfying results in
the case in question, is up to the original poster.

If however the first approach is of interest, I strongly recommend to
study the two papers mentioned earlier and then to decide whether the
Directionality-plugin should be used.

Regards

Herbie

:::::::::::::::::::::::::::::::::::::::::::::::::::
Am 24.03.17 um 14:34 schrieb Straub, Volko A. (Dr.):

> I noticed previously a 'Directionality' plugin in ImageJ that may be
> of use (https://imagej.net/Directionality). I only looked at it
> briefly a while ago and don't have any real experience with it, but
> it looks like it might be useful for your purpose.
>
> Volko
>
> -----Original Message----- From: ImageJ Interest Group
> [mailto:[hidden email]] On Behalf Of Herbie Sent: 23 March 2017
> 19:00 To: [hidden email] Subject: Re: Quantification of axon
> Twist
>
> Dear Eric,
>
> thanks for your explanation concerning the source of your image. So
> the neurons are grown on glass...
>
> re: Orientation Salience Function
>
> Here are links for two papers dealing with theoretical and
> algorithmic aspects. If you don't have much time, please try to
> understand at least the first two pages of the report (2013).
>
> <http://www.gluender.de/Writings/WritingsTexts/WritingsDownloads/1986_IntegralsCentralSlices.zip>
>
>

<http://www.gluender.de/Writings/WritingsTexts/WritingsDownloads/2013_Orientation.zip>

>
> In fact Orientation Salience Functions are weighted histograms and
> could be normalized as such. Weighted, because they don' reflect the
> pure geometric property of orientation but the contrast/brightness of
> the image structures as well.
>
> With your image, it is interesting that the 90deg bending of the
> axons shows up nicely by the decline of the function for angles
> greater 90deg. The bending itself is reflected by the plateau between
> 20deg and 90deg. Beyond 90deg the orientations are less regular.
>
> The generation of the Orientation Salience Function is done with
> ImageJ by use of a dedicated plugin (cf. the 2013 report) and a
> macro.
>
> I shall write you off-list tomorrow for more details.
>
> Best
>
> Herbie
>
> ::::::::::::::::::::::::::::::::::::::::::: Am 23.03.17 um 12:58
> schrieb Eric Denarier:
>> Hi Herbie,
>>
>> Thanks for you help. Image is  actually Dorsal Root Ganglia neurons
>>  grown on coverslip (part of the image).
>>
>> How did you get the the orientation Salience ? With imageJ ?
>>
>> Eric Denarier Grenoble Institut des Neurosciences Inserm U1216
>> Chemin Fortuné Ferrini 38700 La Tronche France
>>
>> Tél :[phone]
>>
>> http://neurosciences.ujf-grenoble.fr/
>>
>> Le 23/03/2017 à 10:12, Herbie a écrit :
>>> orientation salience function
>>
>>
>
> -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
> -- ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

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

[Eric Denarier]

Thanks Straub and Herbie,


Thanks for your suggestions. I will try to figure out how to handle the
porblem.

At the end I want to get a measure of the change in orientation and be
able to compare several images coming from different genotype.

As I have been suggested I may have to cut the images into sector and
get a value of the major axis of orientation in each sector and its
coherency.


Eric Denarier
Grenoble Institut des Neurosciences
Inserm U1216
Chemin Fortuné Ferrini
38700 La Tronche
France

Tél :[phone]

http://neurosciences.ujf-grenoble.fr/

Le 24/03/2017 à 14:34, Straub, Volko A. (Dr.) a écrit :

> I noticed previously a 'Directionality' plugin in ImageJ that may be of use (https://imagej.net/Directionality). I only looked at it briefly a while ago and don't have any real experience with it, but it looks like it might be useful for your purpose.
>
> Volko
>
> -----Original Message-----
> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Herbie
> Sent: 23 March 2017 19:00
> To: [hidden email]
> Subject: Re: Quantification of axon Twist
>
> Dear Eric,
>
> thanks for your explanation concerning the source of your image. So the neurons are grown on glass...
>
> re: Orientation Salience Function
>
> Here are links for two papers dealing with theoretical and algorithmic aspects. If you don't have much time, please try to understand at least the first two pages of the report (2013).
>
> <http://www.gluender.de/Writings/WritingsTexts/WritingsDownloads/1986_IntegralsCentralSlices.zip>
> <http://www.gluender.de/Writings/WritingsTexts/WritingsDownloads/2013_Orientation.zip>
>
> In fact Orientation Salience Functions are weighted histograms and could be normalized as such. Weighted, because they don' reflect the pure geometric property of orientation but the contrast/brightness of the image structures as well.
>
> With your image, it is interesting that the 90deg bending of the axons shows up nicely by the decline of the function for angles greater 90deg.
> The bending itself is reflected by the plateau between 20deg and 90deg.
> Beyond 90deg the orientations are less regular.
>
> The generation of the Orientation Salience Function is done with ImageJ by use of a dedicated plugin (cf. the 2013 report) and a macro.
>
> I shall write you off-list tomorrow for more details.
>
> Best
>
> Herbie
>
> :::::::::::::::::::::::::::::::::::::::::::
> Am 23.03.17 um 12:58 schrieb Eric Denarier:
>> Hi Herbie,
>>
>> Thanks for you help. Image is  actually Dorsal Root Ganglia neurons
>> grown on coverslip (part of the image).
>>
>> How did you get the the orientation Salience ? With imageJ ?
>>
>> Eric Denarier
>> Grenoble Institut des Neurosciences
>> Inserm U1216
>> Chemin Fortuné Ferrini
>> 38700 La Tronche
>> France
>>
>> Tél :[phone]
>>
>> http://neurosciences.ujf-grenoble.fr/
>>
>> Le 23/03/2017 à 10:12, Herbie a écrit :
>>> orientation salience function
>>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

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