ImageJ

Getting stats from Trainable Weka Segmentation plugin?

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

7 messages Options

Mr Patrick Fortuna

Getting stats from Trainable Weka Segmentation plugin?

Hi ImageJ users/developers,

I'm quite new with using ImageJ (Fiji) and was wondering if anyone could help out with getting some stats?
I'm using the Trainable Weka Segmentation plugin through Fiji to train and segment an H/E stained section of a teratoma (benign tumor formed from undifferentiated stem cells).
Once I've trained the classifiers for the various classes I'm interested in, I'd be really interested in getting stats on what percentage of each classifier is present within the image I'm working with.
I've tried using the Measurement of Percent Area (as described here: http://www.nist.gov/lispix/imlab/measure/pctarea.html) although I couldn't seem to set the threshold cause I have more than one color in the segmented image (due to 5 classifiers being run at the same time).
Could anyone recommend how to get the stats of my interested classes?
Any help would be much appreciated!

Thanks and regards,
Patrick

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

Ignacio Arganda-Carreras

Re: Getting stats from Trainable Weka Segmentation plugin?

Dear Patrick,

Here you are a macro that does (I believe) what you want. Let me know if
you have questions.

(i copy/paste it here because the server doesn't seem to like me attaching
it)

<code>

// id values of each class in the classified image
classLabels = newArray( 0, 64, 128, 191, 255 );

setBatchMode( true );
run("Set Measurements...", "area_fraction redirect=None decimal=3");

for( i=0; i<classLabels.length; i ++ )
{
minTh = classLabels[ i ];
maxTh = classLabels[ i ];
if( i < classLabels.length - 1 )
maxTh = classLabels[ i+1 ] -1;
else // last value is assumed to be 255
{
minTh = classLabels[ i-1 ] +1;
maxTh = classLabels[ i ];
}
selectWindow("Classified image");
run("Duplicate...", "title=[to-be-thresholded]");
selectWindow("to-be-thresholded");
setThreshold( minTh, maxTh );
run("Convert to Mask");
run("Measure");
close();
}

</code>

Cheers!

ignacio

On Mon, Sep 9, 2013 at 7:47 AM, Mr Patrick Fortuna <
[hidden email]> wrote:

> Hi ImageJ users/developers,
>
> I'm quite new with using ImageJ (Fiji) and was wondering if anyone could
> help out with getting some stats?
> I'm using the Trainable Weka Segmentation plugin through Fiji to train and
> segment an H/E stained section of a teratoma (benign tumor formed from
> undifferentiated stem cells).
> Once I've trained the classifiers for the various classes I'm interested
> in, I'd be really interested in getting stats on what percentage of each
> classifier is present within the image I'm working with.
> I've tried using the Measurement of Percent Area (as described here:
> http://www.nist.gov/lispix/imlab/measure/pctarea.html) although I
> couldn't seem to set the threshold cause I have more than one color in the
> segmented image (due to 5 classifiers being run at the same time).
> Could anyone recommend how to get the stats of my interested classes?
> Any help would be much appreciated!
>
> Thanks and regards,
> Patrick
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
Ignacio Arganda-Carreras, Ph.D.
Seung's lab, 46-5065
Department of Brain and Cognitive Sciences
Massachusetts Institute of Technology
43 Vassar St.
Cambridge, MA 02139
USA

Phone: (001) 617-324-3747
Website: http://bioweb.cnb.csic.es/~iarganda/index_EN.html

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

Mr Patrick Fortuna

Re: Getting stats from Trainable Weka Segmentation plugin?

Hi Ignacio,

Thanks for the reply!
I tried running the macro, but it comes back with this message:
'No window with the title "Classified image" found'.
Followed by giving the same area measurement for the image 5 times.
I tried changing the "selectwindow" to Trainable Weka Segmentation thinking that this might locate the image, but no luck with that on getting different area measurements as the output.
Any ideas on how to fix this?

Thanks and regards,
Patrick
________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Ignacio Arganda-Carreras [[hidden email]]
Sent: 09 September 2013 07:30 PM
To: [hidden email]
Subject: Re: Getting stats from Trainable Weka Segmentation plugin?

Dear Patrick,

Here you are a macro that does (I believe) what you want. Let me know if
you have questions.

(i copy/paste it here because the server doesn't seem to like me attaching
it)

<code>

// id values of each class in the classified image
classLabels = newArray( 0, 64, 128, 191, 255 );

setBatchMode( true );
run("Set Measurements...", "area_fraction redirect=None decimal=3");

for( i=0; i<classLabels.length; i ++ )
{
minTh = classLabels[ i ];
maxTh = classLabels[ i ];
if( i < classLabels.length - 1 )
maxTh = classLabels[ i+1 ] -1;
else // last value is assumed to be 255
{
minTh = classLabels[ i-1 ] +1;
maxTh = classLabels[ i ];
}
selectWindow("Classified image");
run("Duplicate...", "title=[to-be-thresholded]");
selectWindow("to-be-thresholded");
setThreshold( minTh, maxTh );
run("Convert to Mask");
run("Measure");
close();
}

</code>

Cheers!

ignacio

On Mon, Sep 9, 2013 at 7:47 AM, Mr Patrick Fortuna <
[hidden email]> wrote:

Ignacio Arganda-Carreras

Re: Getting stats from Trainable Weka Segmentation plugin?

My bad, the macro assumes you have clicked on "Create result" and therefore
the "Classified image" window is there. I should have told you.

ignacio

On Tue, Sep 10, 2013 at 11:49 PM, Mr Patrick Fortuna <
[hidden email]> wrote:

> Hi Ignacio,
>
> Thanks for the reply!
> I tried running the macro, but it comes back with this message:
> 'No window with the title "Classified image" found'.
> Followed by giving the same area measurement for the image 5 times.
> I tried changing the "selectwindow" to Trainable Weka Segmentation
> thinking that this might locate the image, but no luck with that on getting
> different area measurements as the output.
> Any ideas on how to fix this?
>
> Thanks and regards,
> Patrick
> ________________________________________
> From: ImageJ Interest Group [[hidden email]] on behalf of Ignacio
> Arganda-Carreras [[hidden email]]
> Sent: 09 September 2013 07:30 PM
> To: [hidden email]
> Subject: Re: Getting stats from Trainable Weka Segmentation plugin?
>
> Dear Patrick,
>
> Here you are a macro that does (I believe) what you want. Let me know if
> you have questions.
>
> (i copy/paste it here because the server doesn't seem to like me attaching
> it)
>
> <code>
>
>
> // id values of each class in the classified image
> classLabels = newArray( 0, 64, 128, 191, 255 );
>
> setBatchMode( true );
> run("Set Measurements...", "area_fraction redirect=None decimal=3");
>
> for( i=0; i<classLabels.length; i ++ )
> {
> minTh = classLabels[ i ];
> maxTh = classLabels[ i ];
> if( i < classLabels.length - 1 )
> maxTh = classLabels[ i+1 ] -1;
> else // last value is assumed to be 255
> {
> minTh = classLabels[ i-1 ] +1;
> maxTh = classLabels[ i ];
> }
> selectWindow("Classified image");
> run("Duplicate...", "title=[to-be-thresholded]");
> selectWindow("to-be-thresholded");
> setThreshold( minTh, maxTh );
> run("Convert to Mask");
> run("Measure");
> close();
> }
>
>
>
>
> </code>
>
>
>
> Cheers!
>
> ignacio
>
>
>
> On Mon, Sep 9, 2013 at 7:47 AM, Mr Patrick Fortuna <
> [hidden email]> wrote:
>
> > Hi ImageJ users/developers,
> >
> > I'm quite new with using ImageJ (Fiji) and was wondering if anyone could
> > help out with getting some stats?
> > I'm using the Trainable Weka Segmentation plugin through Fiji to train
> and
> > segment an H/E stained section of a teratoma (benign tumor formed from
> > undifferentiated stem cells).
> > Once I've trained the classifiers for the various classes I'm interested
> > in, I'd be really interested in getting stats on what percentage of each
> > classifier is present within the image I'm working with.
> > I've tried using the Measurement of Percent Area (as described here:
> > http://www.nist.gov/lispix/imlab/measure/pctarea.html) although I
> > couldn't seem to set the threshold cause I have more than one color in
> the
> > segmented image (due to 5 classifiers being run at the same time).
> > Could anyone recommend how to get the stats of my interested classes?
> > Any help would be much appreciated!
> >
> > Thanks and regards,
> > Patrick
> >
> > --
> > ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> >
>
>
>
> --
> Ignacio Arganda-Carreras, Ph.D.
> Seung's lab, 46-5065
> Department of Brain and Cognitive Sciences
> Massachusetts Institute of Technology
> 43 Vassar St.
> Cambridge, MA 02139
> USA
>
> Phone: (001) 617-324-3747
> Website: http://bioweb.cnb.csic.es/~iarganda/index_EN.html
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

Mr Patrick Fortuna

Re: Getting stats from Trainable Weka Segmentation plugin?

Thanks for that, it worked perfectly!
Would there be any way to increase the amount of classes allowed within one classifier set?
Or do I just need to make several classifier sets to run on the same image?

Cheers,
Patrick
________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Ignacio Arganda-Carreras [[hidden email]]
Sent: 11 September 2013 05:43 PM
To: [hidden email]
Subject: Re: Getting stats from Trainable Weka Segmentation plugin?

My bad, the macro assumes you have clicked on "Create result" and therefore
the "Classified image" window is there. I should have told you.

ignacio

On Tue, Sep 10, 2013 at 11:49 PM, Mr Patrick Fortuna <
[hidden email]> wrote:

Ignacio Arganda-Carreras

Re: Getting stats from Trainable Weka Segmentation plugin?

Hello Patrick,

As it is now, the plugin GUI only accepts 5 classes. If more people are
interested, I can change that value. The only reason I set it to 5 is not
to have a huge interface with too many buttons. How many classes would you
need?

ignacio

On Tue, Sep 17, 2013 at 12:41 AM, Mr Patrick Fortuna <
[hidden email]> wrote:

> Thanks for that, it worked perfectly!
> Would there be any way to increase the amount of classes allowed within
> one classifier set?
> Or do I just need to make several classifier sets to run on the same image?
>
> Cheers,
> Patrick
> ________________________________________
> From: ImageJ Interest Group [[hidden email]] on behalf of Ignacio
> Arganda-Carreras [[hidden email]]
> Sent: 11 September 2013 05:43 PM
> To: [hidden email]
> Subject: Re: Getting stats from Trainable Weka Segmentation plugin?
>
> My bad, the macro assumes you have clicked on "Create result" and therefore
> the "Classified image" window is there. I should have told you.
>
> ignacio
>
>
> On Tue, Sep 10, 2013 at 11:49 PM, Mr Patrick Fortuna <
> [hidden email]> wrote:
>
> > Hi Ignacio,
> >
> > Thanks for the reply!
> > I tried running the macro, but it comes back with this message:
> > 'No window with the title "Classified image" found'.
> > Followed by giving the same area measurement for the image 5 times.
> > I tried changing the "selectwindow" to Trainable Weka Segmentation
> > thinking that this might locate the image, but no luck with that on
> getting
> > different area measurements as the output.
> > Any ideas on how to fix this?
> >
> > Thanks and regards,
> > Patrick
> > ________________________________________
> > From: ImageJ Interest Group [[hidden email]] on behalf of Ignacio
> > Arganda-Carreras [[hidden email]]
> > Sent: 09 September 2013 07:30 PM
> > To: [hidden email]
> > Subject: Re: Getting stats from Trainable Weka Segmentation plugin?
> >
> > Dear Patrick,
> >
> > Here you are a macro that does (I believe) what you want. Let me know if
> > you have questions.
> >
> > (i copy/paste it here because the server doesn't seem to like me
> attaching
> > it)
> >
> > <code>
> >
> >
> > // id values of each class in the classified image
> > classLabels = newArray( 0, 64, 128, 191, 255 );
> >
> > setBatchMode( true );
> > run("Set Measurements...", "area_fraction redirect=None decimal=3");
> >
> > for( i=0; i<classLabels.length; i ++ )
> > {
> > minTh = classLabels[ i ];
> > maxTh = classLabels[ i ];
> > if( i < classLabels.length - 1 )
> > maxTh = classLabels[ i+1 ] -1;
> > else // last value is assumed to be 255
> > {
> > minTh = classLabels[ i-1 ] +1;
> > maxTh = classLabels[ i ];
> > }
> > selectWindow("Classified image");
> > run("Duplicate...", "title=[to-be-thresholded]");
> > selectWindow("to-be-thresholded");
> > setThreshold( minTh, maxTh );
> > run("Convert to Mask");
> > run("Measure");
> > close();
> > }
> >
> >
> >
> >
> > </code>
> >
> >
> >
> > Cheers!
> >
> > ignacio
> >
> >
> >
> > On Mon, Sep 9, 2013 at 7:47 AM, Mr Patrick Fortuna <
> > [hidden email]> wrote:
> >
> > > Hi ImageJ users/developers,
> > >
> > > I'm quite new with using ImageJ (Fiji) and was wondering if anyone
> could
> > > help out with getting some stats?
> > > I'm using the Trainable Weka Segmentation plugin through Fiji to train
> > and
> > > segment an H/E stained section of a teratoma (benign tumor formed from
> > > undifferentiated stem cells).
> > > Once I've trained the classifiers for the various classes I'm
> interested
> > > in, I'd be really interested in getting stats on what percentage of
> each
> > > classifier is present within the image I'm working with.
> > > I've tried using the Measurement of Percent Area (as described here:
> > > http://www.nist.gov/lispix/imlab/measure/pctarea.html) although I
> > > couldn't seem to set the threshold cause I have more than one color in
> > the
> > > segmented image (due to 5 classifiers being run at the same time).
> > > Could anyone recommend how to get the stats of my interested classes?
> > > Any help would be much appreciated!
> > >
> > > Thanks and regards,
> > > Patrick
> > >
> > > --
> > > ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> > >
> >
> >
> >
> > --
> > Ignacio Arganda-Carreras, Ph.D.
> > Seung's lab, 46-5065
> > Department of Brain and Cognitive Sciences
> > Massachusetts Institute of Technology
> > 43 Vassar St.
> > Cambridge, MA 02139
> > USA
> >
> > Phone: (001) 617-324-3747
> > Website: http://bioweb.cnb.csic.es/~iarganda/index_EN.html
> >
> > --
> > ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> >
> >
> > --
> > ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> >
>
>
>
> --
> Ignacio Arganda-Carreras, Ph.D.
> Seung's lab, 46-5065
> Department of Brain and Cognitive Sciences
> Massachusetts Institute of Technology
> 43 Vassar St.
> Cambridge, MA 02139
> USA
>
> Phone: (001) 617-324-3747
> Website: http://bioweb.cnb.csic.es/~iarganda/index_EN.html
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

Mr Patrick Fortuna

Re: Getting stats from Trainable Weka Segmentation plugin?

Hi Ignacio,

Sorry for the late reply.
I'd be interested in possibly having up to 10 classes, between 8-10 would be great.
If it's possible of course.

Thanks and regards,
Patrick
________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Ignacio Arganda-Carreras [[hidden email]]
Sent: 19 September 2013 01:42 AM
To: [hidden email]
Subject: Re: Getting stats from Trainable Weka Segmentation plugin?

Hello Patrick,

As it is now, the plugin GUI only accepts 5 classes. If more people are
interested, I can change that value. The only reason I set it to 5 is not
to have a huge interface with too many buttons. How many classes would you
need?

ignacio

On Tue, Sep 17, 2013 at 12:41 AM, Mr Patrick Fortuna <
[hidden email]> wrote: