Trying to measure nuclear volume using IMAGEJ

[Derek Ingram]

Hi there

 I 'm not sure if this is the right forum to post this...

 I 'm a new user, trying to measure nuclear area to determine ploidy for
salmonids. I'm not sure how to isolate the nucleus. I would like to upload
numerous images and compare nuclear area of different samples. Any help
would be appreciated.

Thanks
Derek

[Fabrice Senger]

Derek Ingram a écrit :

> Hi there
>
>  I 'm not sure if this is the right forum to post this...
>
>  I 'm a new user, trying to measure nuclear area to determine ploidy for
> salmonids. I'm not sure how to isolate the nucleus. I would like to upload
> numerous images and compare nuclear area of different samples. Any help
> would be appreciated.
>
> Thanks
> Derek
>  

Hi there,

it is possible to calculate nuclear volumes with the FISH3D plugin from
Thomas Boudier. For this you need to acquire stacks of your samples
where the nuclueus is fluorescently stained. If you use Hoechst or DAPI
you will stain the DNA, it is possible aswell to stain the nuclear
envelope...

Hope this helps,

Fabrice.

--
Senger Fabrice

[Thomas Boudier]

Hi,

The simplest tool for 3D measurements is the 3D Object Counter :

http://imagejdocu.tudor.lu/doku.php?id=plugin:analysis:3d_object_counter:start

It works like analyse particles, choose you measurements and then
segment your 3D objects, it will do the measurements.

My plugin is more specific to fish analysis.

cheers,

Thomas


Senger Fabrice a écrit :

> Derek Ingram a écrit :
>> Hi there
>>
>>  I 'm not sure if this is the right forum to post this...
>>
>>  I 'm a new user, trying to measure nuclear area to determine ploidy
>> for salmonids. I'm not sure how to isolate the nucleus. I would like
>> to upload numerous images and compare nuclear area of different
>> samples. Any help would be appreciated.
>>
>> Thanks
>> Derek
>>  
> Hi there,
>
> it is possible to calculate nuclear volumes with the FISH3D plugin from
> Thomas Boudier. For this you need to acquire stacks of your samples
> where the nuclueus is fluorescently stained. If you use Hoechst or DAPI
> you will stain the DNA, it is possible aswell to stain the nuclear
> envelope...
>
> Hope this helps,
>
> Fabrice.
>

--
   /**********************************************************/
      Thomas Boudier, MCU Université Pierre et Marie Curie,
      IFR 83. Bat B 7ème étage, porte 709, Jussieu.
      Tel : [phone]  Fax : [phone]
/*******************************************************/

[Derek Ingram]

Fabrice

 Thanks for the suggestion, I 'am trying to find nuclear area with the use of a general stain (wright giemsa, I'm having trouble isolating the actual nucleus and not the complete cell.

Thanks Again for your reply

Derek

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Senger Fabrice
Sent: Tuesday, February 17, 2009 12:06 AM
To: [hidden email]
Subject: Re: Trying to measure nuclear volume using IMAGEJ

Derek Ingram a écrit :

> Hi there
>
>  I 'm not sure if this is the right forum to post this...
>
>  I 'm a new user, trying to measure nuclear area to determine ploidy
> for salmonids. I'm not sure how to isolate the nucleus. I would like
> to upload numerous images and compare nuclear area of different
> samples. Any help would be appreciated.
>
> Thanks
> Derek
>  

Hi there,

it is possible to calculate nuclear volumes with the FISH3D plugin from Thomas Boudier. For this you need to acquire stacks of your samples where the nuclueus is fluorescently stained. If you use Hoechst or DAPI you will stain the DNA, it is possible aswell to stain the nuclear envelope...

Hope this helps,

Fabrice.

--
Senger Fabrice

[karo03]

I assume you have normal pictures, possibly in color with bright  
background and blue nuclei. Try first to detect the nuclei with simple  
threshold:
Image->Type->8bit
Image->Adjust->Threshold...

If there are enough isolated nuclei, possible fill them (Process-
 >Binary->Fill Holes), select them with Wandtool and add them into the  
ROI Manager (cmd T). Then click Measure in ROI Manager. You will get  
beside other things the area of your selected nuclei provided you have  
the image scales adjusted. An estimate for the volume is the ellipsoid  
volume based on the smallest and largest diameter.

Hope that helps
Karsten


Am 17.02.2009 um 17:27 schrieb Ingram, Derek:

> Fabrice
>
> Thanks for the suggestion, I 'am trying to find nuclear area with  
> the use of a general stain (wright giemsa, I'm having trouble  
> isolating the actual nucleus and not the complete cell.
>
> Thanks Again for your reply
>
> Derek
>
> -----Original Message-----
> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf  
> Of Senger Fabrice
> Sent: Tuesday, February 17, 2009 12:06 AM
> To: [hidden email]
> Subject: Re: Trying to measure nuclear volume using IMAGEJ
>
> Derek Ingram a écrit :
>> Hi there
>>
>> I 'm not sure if this is the right forum to post this...
>>
>> I 'm a new user, trying to measure nuclear area to determine ploidy
>> for salmonids. I'm not sure how to isolate the nucleus. I would like
>> to upload numerous images and compare nuclear area of different
>> samples. Any help would be appreciated.
>>
>> Thanks
>> Derek
>>
> Hi there,
>
> it is possible to calculate nuclear volumes with the FISH3D plugin  
> from Thomas Boudier. For this you need to acquire stacks of your  
> samples where the nuclueus is fluorescently stained. If you use  
> Hoechst or DAPI you will stain the DNA, it is possible aswell to  
> stain the nuclear envelope...
>
> Hope this helps,
>
> Fabrice.
>
> --
> Senger Fabrice

Karsten
[hidden email]

[Joris FA Meys]

You can first try to use RGB-split and select the channel that gives the
best contrast between the nuclei and the cytoplasm. I tested some pics on
the internet, and the red channel gave me the best result for that. Your
mile may vary. Then the Treshold (try also the one in
Process->Binary->Treshold) might give you already the result you want.

Kind regards
Joris

On Tue, Feb 17, 2009 at 6:04 PM, Karsten Rodenacker <[hidden email]>wrote:

> I assume you have normal pictures, possibly in color with bright background
> and blue nuclei. Try first to detect the nuclei with simple threshold:
> Image->Type->8bit
> Image->Adjust->Threshold...
>
> If there are enough isolated nuclei, possible fill them
> (Process->Binary->Fill Holes), select them with Wandtool and add them into
> the ROI Manager (cmd T). Then click Measure in ROI Manager. You will get
> beside other things the area of your selected nuclei provided you have the
> image scales adjusted. An estimate for the volume is the ellipsoid volume
> based on the smallest and largest diameter.
>
> Hope that helps
> Karsten
>
>
> Am 17.02.2009 um 17:27 schrieb Ingram, Derek:
>
>  Fabrice
>>
>> Thanks for the suggestion, I 'am trying to find nuclear area with the use
>> of a general stain (wright giemsa, I'm having trouble isolating the actual
>> nucleus and not the complete cell.
>>
>> Thanks Again for your reply
>>
>> Derek
>>
>> -----Original Message-----
>> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
>> Senger Fabrice
>> Sent: Tuesday, February 17, 2009 12:06 AM
>> To: [hidden email]
>> Subject: Re: Trying to measure nuclear volume using IMAGEJ
>>
>> Derek Ingram a écrit :
>>
>>> Hi there
>>>
>>> I 'm not sure if this is the right forum to post this...
>>>
>>> I 'm a new user, trying to measure nuclear area to determine ploidy
>>> for salmonids. I'm not sure how to isolate the nucleus. I would like
>>> to upload numerous images and compare nuclear area of different
>>> samples. Any help would be appreciated.
>>>
>>> Thanks
>>> Derek
>>>
>>>  Hi there,
>>
>> it is possible to calculate nuclear volumes with the FISH3D plugin from
>> Thomas Boudier. For this you need to acquire stacks of your samples where
>> the nuclueus is fluorescently stained. If you use Hoechst or DAPI you will
>> stain the DNA, it is possible aswell to stain the nuclear envelope...
>>
>> Hope this helps,
>>
>> Fabrice.
>>
>> --
>> Senger Fabrice
>>
>
> Karsten
> [hidden email]
>

[Dale Callaham]

Derek's original stated goal was about determining ploidy level. I
replied to Derek off list that nuclear volume may not be an accurate
representation of ploidy level and that measuring the integrated signal
of fluorescently labeled DNA is how it used to be done and might be
better. I saw a similar suggestion sent to the list. While ImageJ is
capable of marvelous image manipulations, perhaps others could comment
on the suitability of measuring volume for determining ploidy before we
go too far down the road of helping with the wrong thing. Is this how it
is done, properly?

Dale


joris meys wrote:

> ---------------------- Information from the mail header -----------------------
> Sender:       ImageJ Interest Group <[hidden email]>
> Poster:       joris meys <[hidden email]>
> Subject:      Re: Trying to measure nuclear volume using IMAGEJ
> -------------------------------------------------------------------------------
>
> You can first try to use RGB-split and select the channel that gives the
> best contrast between the nuclei and the cytoplasm. I tested some pics on
> the internet, and the red channel gave me the best result for that. Your
> mile may vary. Then the Treshold (try also the one in
> Process->Binary->Treshold) might give you already the result you want.
>
> Kind regards
> Joris
>
> On Tue, Feb 17, 2009 at 6:04 PM, Karsten Rodenacker <[hidden email]>wro=
> te:
>
>> I assume you have normal pictures, possibly in color with bright backgrou=
> nd
>> and blue nuclei. Try first to detect the nuclei with simple threshold:
>> Image->Type->8bit
>> Image->Adjust->Threshold...
>>
>> If there are enough isolated nuclei, possible fill them
>> (Process->Binary->Fill Holes), select them with Wandtool and add them int=
> o
>> the ROI Manager (cmd T). Then click Measure in ROI Manager. You will get
>> beside other things the area of your selected nuclei provided you have th=
> e
>> image scales adjusted. An estimate for the volume is the ellipsoid volume
>> based on the smallest and largest diameter.
>>
>> Hope that helps
>> Karsten
>>
>>
>> Am 17.02.2009 um 17:27 schrieb Ingram, Derek:
>>
>>  Fabrice
>>> Thanks for the suggestion, I 'am trying to find nuclear area with the us=
> e
>>> of a general stain (wright giemsa, I'm having trouble isolating the actu=
> al
>>> nucleus and not the complete cell.
>>>
>>> Thanks Again for your reply
>>>
>>> Derek
>>>
>>> -----Original Message-----
>>> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
>>> Senger Fabrice
>>> Sent: Tuesday, February 17, 2009 12:06 AM
>>> To: [hidden email]
>>> Subject: Re: Trying to measure nuclear volume using IMAGEJ
>>>
>>> Derek Ingram a =E9crit :
>>>
>>>> Hi there
>>>>
>>>> I 'm not sure if this is the right forum to post this...
>>>>
>>>> I 'm a new user, trying to measure nuclear area to determine ploidy
>>>> for salmonids. I'm not sure how to isolate the nucleus. I would like
>>>> to upload numerous images and compare nuclear area of different
>>>> samples. Any help would be appreciated.
>>>>
>>>> Thanks
>>>> Derek
>>>>
>>>>  Hi there,
>>> it is possible to calculate nuclear volumes with the FISH3D plugin from
>>> Thomas Boudier. For this you need to acquire stacks of your samples wher=
> e
>>> the nuclueus is fluorescently stained. If you use Hoechst or DAPI you wi=
> ll
>>> stain the DNA, it is possible aswell to stain the nuclear envelope...
>>>
>>> Hope this helps,
>>>
>>> Fabrice.
>>>
>>> --
>>> Senger Fabrice
>>>
>> Karsten
>> [hidden email]
>>

[Bill Christens-Barry]

Dale,

I'm not up on current methodology, but when I was doing this kind of thing there was always a
concern that many colorimetric (e.g. H&E) and fluorometric dyes did not bind stoichiometrically.
We confined ourselves to Feulgen dyes, which were better behaved. It would make sense to verify
the stoichiometry of the regime being used.

Bill

On Tue, 17 Feb 2009 13:14:40 -0500, Dale Callaham <[hidden email]> wrote:

>Derek's original stated goal was about determining ploidy level. I
>replied to Derek off list that nuclear volume may not be an accurate
>representation of ploidy level and that measuring the integrated signal
>of fluorescently labeled DNA is how it used to be done and might be
>better. I saw a similar suggestion sent to the list. While ImageJ is
>capable of marvelous image manipulations, perhaps others could comment
>on the suitability of measuring volume for determining ploidy before we
>go too far down the road of helping with the wrong thing. Is this how it
>is done, properly?
>
>Dale
>
>
>joris meys wrote:
>> ---------------------- Information from the mail header -----------------------
>> Sender:       ImageJ Interest Group <[hidden email]>
>> Poster:       joris meys <[hidden email]>
>> Subject:      Re: Trying to measure nuclear volume using IMAGEJ
>> ------------------------------------------------------------------------

-------

>>
>> You can first try to use RGB-split and select the channel that gives the
>> best contrast between the nuclei and the cytoplasm. I tested some pics on
>> the internet, and the red channel gave me the best result for that. Your
>> mile may vary. Then the Treshold (try also the one in
>> Process->Binary->Treshold) might give you already the result you want.
>>
>> Kind regards
>> Joris
>>
>> On Tue, Feb 17, 2009 at 6:04 PM, Karsten Rodenacker <[hidden email]>wro=
>> te:
>>
>>> I assume you have normal pictures, possibly in color with bright backgrou=
>> nd
>>> and blue nuclei. Try first to detect the nuclei with simple threshold:
>>> Image->Type->8bit
>>> Image->Adjust->Threshold...
>>>
>>> If there are enough isolated nuclei, possible fill them
>>> (Process->Binary->Fill Holes), select them with Wandtool and add them int=
>> o
>>> the ROI Manager (cmd T). Then click Measure in ROI Manager. You will get
>>> beside other things the area of your selected nuclei provided you have th=
>> e
>>> image scales adjusted. An estimate for the volume is the ellipsoid volume
>>> based on the smallest and largest diameter.
>>>
>>> Hope that helps
>>> Karsten
>>>
>>>
>>> Am 17.02.2009 um 17:27 schrieb Ingram, Derek:
>>>
>>>  Fabrice
>>>> Thanks for the suggestion, I 'am trying to find nuclear area with the us=
>> e
>>>> of a general stain (wright giemsa, I'm having trouble isolating the actu=
>> al
>>>> nucleus and not the complete cell.
>>>>
>>>> Thanks Again for your reply
>>>>
>>>> Derek
>>>>
>>>> -----Original Message-----
>>>> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
>>>> Senger Fabrice
>>>> Sent: Tuesday, February 17, 2009 12:06 AM
>>>> To: [hidden email]
>>>> Subject: Re: Trying to measure nuclear volume using IMAGEJ
>>>>
>>>> Derek Ingram a =E9crit :
>>>>
>>>>> Hi there
>>>>>
>>>>> I 'm not sure if this is the right forum to post this...
>>>>>
>>>>> I 'm a new user, trying to measure nuclear area to determine ploidy
>>>>> for salmonids. I'm not sure how to isolate the nucleus. I would like
>>>>> to upload numerous images and compare nuclear area of different
>>>>> samples. Any help would be appreciated.
>>>>>
>>>>> Thanks
>>>>> Derek
>>>>>
>>>>>  Hi there,
>>>> it is possible to calculate nuclear volumes with the FISH3D plugin from
>>>> Thomas Boudier. For this you need to acquire stacks of your samples wher=
>> e
>>>> the nuclueus is fluorescently stained. If you use Hoechst or DAPI you wi=
>> ll
>>>> stain the DNA, it is possible aswell to stain the nuclear envelope...
>>>>
>>>> Hope this helps,
>>>>
>>>> Fabrice.
>>>>
>>>> --
>>>> Senger Fabrice
>>>>
>>> Karsten
>>> [hidden email]
>>>

[Hugo A. M. Torres]

On Tue, 2009-02-17 at 14:48 -0500, Bill Christens-Barry wrote:

> Dale,
>
> I'm not up on current methodology, but when I was doing this kind of thing there was always a
> concern that many colorimetric (e.g. H&E) and fluorometric dyes did not bind stoichiometrically.
> We confined ourselves to Feulgen dyes, which were better behaved. It would make sense to verify
> the stoichiometry of the regime being used.
>
> Bill


Hello, may I start a discussion on this matter of stoichiometry?

I once discussed this issue with Gabriel Landini and he expressed
concerns on the same issue.  If I can recall correctly when the
antigen-antibody interaction is not stoichiometric it's not possible in
principle to quantitate histochemical staining.

On the other hand there seems to exist the problem of sensitivity, that
while a reaction is stoichiometric it might not be detectable by imaging
systems so one is forced to resort to signal amplification systems which
often are non-stoichiometric. See for instance O. Schmitt et al. / Brain
Research Protocols 12 (2004) 157–171

So what is the state of the acceptance of optical density measurements
in histochemistry methods that employ molecular signal amplification
systems?

I ask because I found different opinions published in the literature, so
I would like to know more what this community thinks of the matter.

I understand this may be a little off-topic, but imageJ is frequently
cited in densitometry assays, so I thought it would be appropriate.

--
Hugo Arruda de Moura Torres
==================================
Departamento de Biofísica
Universidade Federal de São Paulo
Rua Botucatu 862 7o. andar
CEP 04023-062 Vila Clementino
São Paulo - SP - Brasil
Tel:[phone] r.220
Fax: [phone]

[Michael Doube]

Bill & Dale

I tried Feulgen staining on tendon some years ago and it was pretty harsh on the tissue; intercalating dyes like propidium iodide, Hoechst or DAPI may be kinder.  I can't speak for the quantitative, fluorimetric value of these stains though as I only used PI for nuclear morphology.

http://en.wikipedia.org/wiki/Hoechst_stain

Mike

________________________________________
From: ImageJ Interest Group [[hidden email]] On Behalf Of Bill Christens-Barry [[hidden email]]
Sent: Tuesday, February 17, 2009 9:48 PM
To: [hidden email]
Subject: Re: Trying to measure nuclear volume using IMAGEJ

Dale,

I'm not up on current methodology, but when I was doing this kind of thing there was always a
concern that many colorimetric (e.g. H&E) and fluorometric dyes did not bind stoichiometrically.
We confined ourselves to Feulgen dyes, which were better behaved. It would make sense to verify
the stoichiometry of the regime being used.

Bill

On Tue, 17 Feb 2009 13:14:40 -0500, Dale Callaham <[hidden email]> wrote:

>Derek's original stated goal was about determining ploidy level. I
>replied to Derek off list that nuclear volume may not be an accurate
>representation of ploidy level and that measuring the integrated signal
>of fluorescently labeled DNA is how it used to be done and might be
>better. I saw a similar suggestion sent to the list. While ImageJ is
>capable of marvelous image manipulations, perhaps others could comment
>on the suitability of measuring volume for determining ploidy before we
>go too far down the road of helping with the wrong thing. Is this how it
>is done, properly?
>
>Dale
>
>
>joris meys wrote:
>> ---------------------- Information from the mail header -----------------------
>> Sender:       ImageJ Interest Group <[hidden email]>
>> Poster:       joris meys <[hidden email]>
>> Subject:      Re: Trying to measure nuclear volume using IMAGEJ
>> ------------------------------------------------------------------------

-------

>>
>> You can first try to use RGB-split and select the channel that gives the
>> best contrast between the nuclei and the cytoplasm. I tested some pics on
>> the internet, and the red channel gave me the best result for that. Your
>> mile may vary. Then the Treshold (try also the one in
>> Process->Binary->Treshold) might give you already the result you want.
>>
>> Kind regards
>> Joris
>>
>> On Tue, Feb 17, 2009 at 6:04 PM, Karsten Rodenacker <[hidden email]>wro=
>> te:
>>
>>> I assume you have normal pictures, possibly in color with bright backgrou=
>> nd
>>> and blue nuclei. Try first to detect the nuclei with simple threshold:
>>> Image->Type->8bit
>>> Image->Adjust->Threshold...
>>>
>>> If there are enough isolated nuclei, possible fill them
>>> (Process->Binary->Fill Holes), select them with Wandtool and add them int=
>> o
>>> the ROI Manager (cmd T). Then click Measure in ROI Manager. You will get
>>> beside other things the area of your selected nuclei provided you have th=
>> e
>>> image scales adjusted. An estimate for the volume is the ellipsoid volume
>>> based on the smallest and largest diameter.
>>>
>>> Hope that helps
>>> Karsten
>>>
>>>
>>> Am 17.02.2009 um 17:27 schrieb Ingram, Derek:
>>>
>>>  Fabrice
>>>> Thanks for the suggestion, I 'am trying to find nuclear area with the us=
>> e
>>>> of a general stain (wright giemsa, I'm having trouble isolating the actu=
>> al
>>>> nucleus and not the complete cell.
>>>>
>>>> Thanks Again for your reply
>>>>
>>>> Derek
>>>>
>>>> -----Original Message-----
>>>> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
>>>> Senger Fabrice
>>>> Sent: Tuesday, February 17, 2009 12:06 AM
>>>> To: [hidden email]
>>>> Subject: Re: Trying to measure nuclear volume using IMAGEJ
>>>>
>>>> Derek Ingram a =E9crit :
>>>>
>>>>> Hi there
>>>>>
>>>>> I 'm not sure if this is the right forum to post this...
>>>>>
>>>>> I 'm a new user, trying to measure nuclear area to determine ploidy
>>>>> for salmonids. I'm not sure how to isolate the nucleus. I would like
>>>>> to upload numerous images and compare nuclear area of different
>>>>> samples. Any help would be appreciated.
>>>>>
>>>>> Thanks
>>>>> Derek
>>>>>
>>>>>  Hi there,
>>>> it is possible to calculate nuclear volumes with the FISH3D plugin from
>>>> Thomas Boudier. For this you need to acquire stacks of your samples wher=
>> e
>>>> the nuclueus is fluorescently stained. If you use Hoechst or DAPI you wi=
>> ll
>>>> stain the DNA, it is possible aswell to stain the nuclear envelope...
>>>>
>>>> Hope this helps,
>>>>
>>>> Fabrice.
>>>>
>>>> --
>>>> Senger Fabrice
>>>>
>>> Karsten
>>> [hidden email]
>>>

[Gabriel Landini]

On Tuesday 17 February 2009 21:07:14 Doube, Michael wrote:
> I tried Feulgen staining on tendon some years ago and it was pretty harsh
> on the tissue; intercalating dyes like propidium iodide, Hoechst or DAPI
> may be kinder.  I can't speak for the quantitative, fluorimetric value of
> these stains though as I only used PI for nuclear morphology.

Some time ago Darek Kedra posted this reference, which is very good:
David C. Hardie, T. Ryan Gregory, and Paul D.N. Hebert, “From Pixels to
Picograms: A Beginners' Guide to Genome Quantification by Feulgen Image
Analysis Densitometry,” J. Histochem. Cytochem. 50, no. 6 (June 1, 2002):
735-750.

I however fail to understand why the authors suggest not to use background
correction (they do not give an explanation either). After all, if background
correction is not done, one cannot be sure of measuring the darkness of the
Feulgen stain or a dark spot on the illumination field.

Cheers

G.