ImageJ

Cell counting with imageJ

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

13 messages Options

Antonis Giakountis

Cell counting with imageJ

Dear all,

I am a plant Biologist and I would like to use ImageJ in order to count cell numbers (diameter 10-20 microns on average) and to estimate percentage of cells that have fluorescence.

Is it possible to to that with ImageJ and if yes do I have to install a certain plugin? If not can you recomment another software?

Thank you in advance for your time,

Antonis

--------------------------------------------
Antonis Giakountis, PhD
Prof. David Baulcombe group
Department of Plant Sciences
University of Cambridge
Downing Street
Cambridge CB2 3EA
UK

dpoburko

Re: Cell counting with imageJ

Dear Antonis,

Image J is definitely suited to counting cells. How you go about doing
so depends entirely on what your images look like and how you plan to
identify cells that are fluorescent and non-fluorescent. You'll have to
provide more details of what your images look like. Also have a read
through the Image J documentation if you haven't done so already. I
think there are some good hints to getting started in there.

Cheers,
Damon

Antonis Giakountis wrote:

> Dear all,
>
> I am a plant Biologist and I would like to use ImageJ in order to count cell numbers (diameter 10-20 microns on average) and to estimate percentage of cells that have fluorescence.
>
> Is it possible to to that with ImageJ and if yes do I have to install a certain plugin? If not can you recomment another software?
>
> Thank you in advance for your time,
>
> Antonis
>
> --------------------------------------------
> Antonis Giakountis, PhD
> Prof. David Baulcombe group
> Department of Plant Sciences
> University of Cambridge
> Downing Street
> Cambridge CB2 3EA
> UK
>

G. Esteban Fernandez

Re: Cell counting with imageJ

In reply to this post by Antonis Giakountis

How many images/cells do you need to count? Is this something that
you can do by hand in a matter of a couple of days or do you want an
automated procedure? Often it takes a few days to get an automated
procedure to work correctly, if it ever works at all, during which
time you could've done the counting by hand. There are probably
ImageJ plugins that keep track of counts as you click on cells in the
images.

Do you have images that match the fluorescence images to get total
cell numbers? For example transmitted light (brightfield/phase/DIC)
images?

-Esteban

On Fri, Nov 13, 2009 at 7:43 AM, Antonis Giakountis <[hidden email]> wrote:

--
G. Esteban Fernandez, Ph.D.

Associate Director
Molecular Cytology Core Facility
University of Missouri
120 Bond Life Sciences Center
Columbia, MO 65211

http://www.biotech.missouri.edu/mcc

(573)882-4895
(573)884-9676 fax

Antonis Giakountis

Re: Cell counting with imageJ

In reply to this post by dpoburko

Thanks Damon, I am doing an immunofluorescence as we speak. If I get it to
work nicely I will attach an image and send it over. I expect my
fluorescence to be in the nucleus (approx. 1-2 microns) of a rather small
cell size population (10-15 microns). I must admit that I haven't played
with imageJ before so I will have a look at the documentation as well.

Cheers,

Antonis
----- Original Message -----
From: "Damon Poburko" <[hidden email]>
To: <[hidden email]>
Sent: Friday, November 13, 2009 2:04 PM
Subject: Re: Cell counting with imageJ

> Dear Antonis,
>
> Image J is definitely suited to counting cells. How you go about doing so
> depends entirely on what your images look like and how you plan to
> identify cells that are fluorescent and non-fluorescent. You'll have to
> provide more details of what your images look like. Also have a read
> through the Image J documentation if you haven't done so already. I think
> there are some good hints to getting started in there.
>
> Cheers,
> Damon
>
>
> Antonis Giakountis wrote:
>> Dear all,
>>
>> I am a plant Biologist and I would like to use ImageJ in order to count
>> cell numbers (diameter 10-20 microns on average) and to estimate
>> percentage of cells that have fluorescence.
>>
>> Is it possible to to that with ImageJ and if yes do I have to install a
>> certain plugin? If not can you recomment another software?
>>
>> Thank you in advance for your time,
>>
>> Antonis
>>
>> --------------------------------------------
>> Antonis Giakountis, PhD
>> Prof. David Baulcombe group
>> Department of Plant Sciences
>> University of Cambridge
>> Downing Street
>> Cambridge CB2 3EA
>> UK
>>
>

Antonis Giakountis

Re: Cell counting with imageJ

In reply to this post by G. Esteban Fernandez

Hi Esteban,

I really would prefer not to do it by hand. A typical experiment of mine
involves several timepoints (10-16) and at least five reps per line. Each
one of those has hundreds of cells and I was hoping for some stitistics out
of this. To make a long story sort I expect thousants of cells and I would
like to get an estimation of the percentage of them that has fluorescence
out of the total.

In the worst case I could deal with user define-clinking as long as I could
mark the cells that are counted. As for the images I am aiming for
triple-stain nuclear immunofluorescence, but since this is done with the
confocal I can easily scan each channel separately, including the
transmitted-light one (BF).

Antonis
----- Original Message -----
From: "G. Esteban Fernandez" <[hidden email]>
To: <[hidden email]>
Sent: Friday, November 13, 2009 2:10 PM
Subject: Re: Cell counting with imageJ

> How many images/cells do you need to count? Is this something that
> you can do by hand in a matter of a couple of days or do you want an
> automated procedure? Often it takes a few days to get an automated
> procedure to work correctly, if it ever works at all, during which
> time you could've done the counting by hand. There are probably
> ImageJ plugins that keep track of counts as you click on cells in the
> images.
>
> Do you have images that match the fluorescence images to get total
> cell numbers? For example transmitted light (brightfield/phase/DIC)
> images?
>
> -Esteban
>
>
> On Fri, Nov 13, 2009 at 7:43 AM, Antonis Giakountis <[hidden email]>
> wrote:
>> Dear all,
>>
>> I am a plant Biologist and I would like to use ImageJ in order to count
>> cell numbers (diameter 10-20 microns on average) and to estimate
>> percentage of cells that have fluorescence.
>>
>> Is it possible to to that with ImageJ and if yes do I have to install a
>> certain plugin? If not can you recomment another software?
>>
>> Thank you in advance for your time,
>>
>> Antonis
>>
>> --------------------------------------------
>> Antonis Giakountis, PhD
>> Prof. David Baulcombe group
>> Department of Plant Sciences
>> University of Cambridge
>> Downing Street
>> Cambridge CB2 3EA
>> UK
>>
>
>
>
> --
> G. Esteban Fernandez, Ph.D.
>
> Associate Director
> Molecular Cytology Core Facility
> University of Missouri
> 120 Bond Life Sciences Center
> Columbia, MO 65211
>
> http://www.biotech.missouri.edu/mcc
>
> (573)882-4895
> (573)884-9676 fax
>

Jason Bailey-2

Re: Cell counting with imageJ

Antonis,

I have had much luck with the Image-Based Tool for Counting Nuclei (ITCN)
with mammalian cells. I imagine it could work well with yours as well. I
typically do two counts--first count all nuclei labeled with Hoechst 33342,
then count the nuclei labeled with a different dye (either ethidium or FITC,
usually). Although this might require some changes to how you label your
cells.

If you must resort to manual counting, the "Cell Counter" plugin marks
cells/nuclei as you click and lets you track multiple cell types as you go.

Both are available at the ImageJ Plugins page:
http://rsbweb.nih.gov/ij/plugins/index.html

Good luck!

Jason

On Fri, Nov 13, 2009 at 9:25 AM, Antonis Giakountis <[hidden email]> wrote:

> Hi Esteban,
>
> I really would prefer not to do it by hand. A typical experiment of mine
> involves several timepoints (10-16) and at least five reps per line. Each
> one of those has hundreds of cells and I was hoping for some stitistics out
> of this. To make a long story sort I expect thousants of cells and I would
> like to get an estimation of the percentage of them that has fluorescence
> out of the total.
>
> In the worst case I could deal with user define-clinking as long as I could
> mark the cells that are counted. As for the images I am aiming for
> triple-stain nuclear immunofluorescence, but since this is done with the
> confocal I can easily scan each channel separately, including the
> transmitted-light one (BF).
>
> Antonis
> ----- Original Message ----- From: "G. Esteban Fernandez" <
> [hidden email]>
>
> To: <[hidden email]>
> Sent: Friday, November 13, 2009 2:10 PM
>
> Subject: Re: Cell counting with imageJ
>
>
> How many images/cells do you need to count? Is this something that
>> you can do by hand in a matter of a couple of days or do you want an
>> automated procedure? Often it takes a few days to get an automated
>> procedure to work correctly, if it ever works at all, during which
>> time you could've done the counting by hand. There are probably
>> ImageJ plugins that keep track of counts as you click on cells in the
>> images.
>>
>> Do you have images that match the fluorescence images to get total
>> cell numbers? For example transmitted light (brightfield/phase/DIC)
>> images?
>>
>> -Esteban
>>
>>
>> On Fri, Nov 13, 2009 at 7:43 AM, Antonis Giakountis <[hidden email]>
>> wrote:
>>
>>> Dear all,
>>>
>>> I am a plant Biologist and I would like to use ImageJ in order to count
>>> cell numbers (diameter 10-20 microns on average) and to estimate percentage
>>> of cells that have fluorescence.
>>>
>>> Is it possible to to that with ImageJ and if yes do I have to install a
>>> certain plugin? If not can you recomment another software?
>>>
>>> Thank you in advance for your time,
>>>
>>> Antonis
>>>
>>> --------------------------------------------
>>> Antonis Giakountis, PhD
>>> Prof. David Baulcombe group
>>> Department of Plant Sciences
>>> University of Cambridge
>>> Downing Street
>>> Cambridge CB2 3EA
>>> UK
>>>
>>>
>>
>>
>> --
>> G. Esteban Fernandez, Ph.D.
>>
>> Associate Director
>> Molecular Cytology Core Facility
>> University of Missouri
>> 120 Bond Life Sciences Center
>> Columbia, MO 65211
>>
>> http://www.biotech.missouri.edu/mcc
>>
>> (573)882-4895
>> (573)884-9676 fax
>>
>>

G. Esteban Fernandez

Re: Cell counting with imageJ

You can also use the the built-in functions under Analyze > Analyze
Particles to count automatically after thresholding the images. If
thresholding doesn't do a good job of including all the nuclei and
excluding background then it can get tricky. Good thresholding is the
key to good results with automatic counting. Also, if there are
nuclei touching each other you can separate them with Process > Binary
> Watershed. Maybe the ITCN plugin does the separation automatically?

-Esteban

On Fri, Nov 13, 2009 at 8:36 AM, Jason Bailey <[hidden email]> wrote:

> Antonis,
>
> I have had much luck with the Image-Based Tool for Counting Nuclei (ITCN)
> with mammalian cells. I imagine it could work well with yours as well. I
> typically do two counts--first count all nuclei labeled with Hoechst 33342,
> then count the nuclei labeled with a different dye (either ethidium or FITC,
> usually). Although this might require some changes to how you label your
> cells.
>
> If you must resort to manual counting, the "Cell Counter" plugin marks
> cells/nuclei as you click and lets you track multiple cell types as you go.
>
> Both are available at the ImageJ Plugins page:
> http://rsbweb.nih.gov/ij/plugins/index.html
>
> Good luck!
>
> Jason
>
> On Fri, Nov 13, 2009 at 9:25 AM, Antonis Giakountis <[hidden email]> wrote:
>
>> Hi Esteban,
>>
>> I really would prefer not to do it by hand. A typical experiment of mine
>> involves several timepoints (10-16) and at least five reps per line. Each
>> one of those has hundreds of cells and I was hoping for some stitistics out
>> of this. To make a long story sort I expect thousants of cells and I would
>> like to get an estimation of the percentage of them that has fluorescence
>> out of the total.
>>
>> In the worst case I could deal with user define-clinking as long as I could
>> mark the cells that are counted. As for the images I am aiming for
>> triple-stain nuclear immunofluorescence, but since this is done with the
>> confocal I can easily scan each channel separately, including the
>> transmitted-light one (BF).
>>
>> Antonis
>> ----- Original Message ----- From: "G. Esteban Fernandez" <
>> [hidden email]>
>>
>> To: <[hidden email]>
>> Sent: Friday, November 13, 2009 2:10 PM
>>
>> Subject: Re: Cell counting with imageJ
>>
>>
>> How many images/cells do you need to count? Is this something that
>>> you can do by hand in a matter of a couple of days or do you want an
>>> automated procedure? Often it takes a few days to get an automated
>>> procedure to work correctly, if it ever works at all, during which
>>> time you could've done the counting by hand. There are probably
>>> ImageJ plugins that keep track of counts as you click on cells in the
>>> images.
>>>
>>> Do you have images that match the fluorescence images to get total
>>> cell numbers? For example transmitted light (brightfield/phase/DIC)
>>> images?
>>>
>>> -Esteban
>>>
>>>
>>> On Fri, Nov 13, 2009 at 7:43 AM, Antonis Giakountis <[hidden email]>
>>> wrote:
>>>
>>>> Dear all,
>>>>
>>>> I am a plant Biologist and I would like to use ImageJ in order to count
>>>> cell numbers (diameter 10-20 microns on average) and to estimate percentage
>>>> of cells that have fluorescence.
>>>>
>>>> Is it possible to to that with ImageJ and if yes do I have to install a
>>>> certain plugin? If not can you recomment another software?
>>>>
>>>> Thank you in advance for your time,
>>>>
>>>> Antonis
>>>>
>>>> --------------------------------------------
>>>> Antonis Giakountis, PhD
>>>> Prof. David Baulcombe group
>>>> Department of Plant Sciences
>>>> University of Cambridge
>>>> Downing Street
>>>> Cambridge CB2 3EA
>>>> UK
>>>>
>>>>
>>>
>>>
>>> --
>>> G. Esteban Fernandez, Ph.D.
>>>
>>> Associate Director
>>> Molecular Cytology Core Facility
>>> University of Missouri
>>> 120 Bond Life Sciences Center
>>> Columbia, MO 65211
>>>
>>> http://www.biotech.missouri.edu/mcc
>>>
>>> (573)882-4895
>>> (573)884-9676 fax
>>>
>>>
>

Antonis Giakountis

Re: Cell counting with imageJ

In reply to this post by G. Esteban Fernandez

Thank you all for your time and kind help. I tried cell counter and seems to
work nice but I also want to try ITCN (have some problems launching it at
the moment though). Sooner or later I will get what I want.

With best wishes,

Antonis
----- Original Message -----
From: "G. Esteban Fernandez" <[hidden email]>
To: <[hidden email]>
Sent: Friday, November 13, 2009 2:10 PM
Subject: Re: Cell counting with imageJ

Jason Bailey-2

Re: Cell counting with imageJ

In reply to this post by G. Esteban Fernandez

Esteban: I think ITCN uses a process somewhat more complex than watershed.
It does not threshold the image, so it works quite well for counting nuclei
that are very close together, and even overlapping slightly. As I
understand, it uses the colvolution and find maxima functions instead. It
requires a little trial and error in setting up the input parameters
initially, but then the process moves pretty quickly, and it produces better
results than I ever managed using thresholding and watershed processing. It
also marks the image so that you can do some quality control.

Antonis: It is a little picky--make sure the image is open before you try to
launch ITCN. Also, optimize the input parameters with a smaller ROI--if your
images are large, it takes a lot of processing power.

Jason

On Fri, Nov 13, 2009 at 9:46 AM, G. Esteban Fernandez <
[hidden email]> wrote:

> You can also use the the built-in functions under Analyze > Analyze
> Particles to count automatically after thresholding the images. If
> thresholding doesn't do a good job of including all the nuclei and
> excluding background then it can get tricky. Good thresholding is the
> key to good results with automatic counting. Also, if there are
> nuclei touching each other you can separate them with Process > Binary
> > Watershed. Maybe the ITCN plugin does the separation automatically?
>
> -Esteban
>
>
> On Fri, Nov 13, 2009 at 8:36 AM, Jason Bailey <[hidden email]>
> wrote:
> > Antonis,
> >
> > I have had much luck with the Image-Based Tool for Counting Nuclei (ITCN)
> > with mammalian cells. I imagine it could work well with yours as well. I
> > typically do two counts--first count all nuclei labeled with Hoechst
> 33342,
> > then count the nuclei labeled with a different dye (either ethidium or
> FITC,
> > usually). Although this might require some changes to how you label your
> > cells.
> >
> > If you must resort to manual counting, the "Cell Counter" plugin marks
> > cells/nuclei as you click and lets you track multiple cell types as you
> go.
> >
> > Both are available at the ImageJ Plugins page:
> > http://rsbweb.nih.gov/ij/plugins/index.html
> >
> > Good luck!
> >
> > Jason
> >
> > On Fri, Nov 13, 2009 at 9:25 AM, Antonis Giakountis <[hidden email]>
> wrote:
> >
> >> Hi Esteban,
> >>
> >> I really would prefer not to do it by hand. A typical experiment of mine
> >> involves several timepoints (10-16) and at least five reps per line.
> Each
> >> one of those has hundreds of cells and I was hoping for some stitistics
> out
> >> of this. To make a long story sort I expect thousants of cells and I
> would
> >> like to get an estimation of the percentage of them that has
> fluorescence
> >> out of the total.
> >>
> >> In the worst case I could deal with user define-clinking as long as I
> could
> >> mark the cells that are counted. As for the images I am aiming for
> >> triple-stain nuclear immunofluorescence, but since this is done with the
> >> confocal I can easily scan each channel separately, including the
> >> transmitted-light one (BF).
> >>
> >> Antonis
> >> ----- Original Message ----- From: "G. Esteban Fernandez" <
> >> [hidden email]>
> >>
> >> To: <[hidden email]>
> >> Sent: Friday, November 13, 2009 2:10 PM
> >>
> >> Subject: Re: Cell counting with imageJ
> >>
> >>
> >> How many images/cells do you need to count? Is this something that
> >>> you can do by hand in a matter of a couple of days or do you want an
> >>> automated procedure? Often it takes a few days to get an automated
> >>> procedure to work correctly, if it ever works at all, during which
> >>> time you could've done the counting by hand. There are probably
> >>> ImageJ plugins that keep track of counts as you click on cells in the
> >>> images.
> >>>
> >>> Do you have images that match the fluorescence images to get total
> >>> cell numbers? For example transmitted light (brightfield/phase/DIC)
> >>> images?
> >>>
> >>> -Esteban
> >>>
> >>>
> >>> On Fri, Nov 13, 2009 at 7:43 AM, Antonis Giakountis <[hidden email]>
> >>> wrote:
> >>>
> >>>> Dear all,
> >>>>
> >>>> I am a plant Biologist and I would like to use ImageJ in order to
> count
> >>>> cell numbers (diameter 10-20 microns on average) and to estimate
> percentage
> >>>> of cells that have fluorescence.
> >>>>
> >>>> Is it possible to to that with ImageJ and if yes do I have to install
> a
> >>>> certain plugin? If not can you recomment another software?
> >>>>
> >>>> Thank you in advance for your time,
> >>>>
> >>>> Antonis
> >>>>
> >>>> --------------------------------------------
> >>>> Antonis Giakountis, PhD
> >>>> Prof. David Baulcombe group
> >>>> Department of Plant Sciences
> >>>> University of Cambridge
> >>>> Downing Street
> >>>> Cambridge CB2 3EA
> >>>> UK
> >>>>
> >>>>
> >>>
> >>>
> >>> --
> >>> G. Esteban Fernandez, Ph.D.
> >>>
> >>> Associate Director
> >>> Molecular Cytology Core Facility
> >>> University of Missouri
> >>> 120 Bond Life Sciences Center
> >>> Columbia, MO 65211
> >>>
> >>> http://www.biotech.missouri.edu/mcc
> >>>
> >>> (573)882-4895
> >>> (573)884-9676 fax
> >>>
> >>>
> >
>
>
>
> --
> G. Esteban Fernandez, Ph.D.
>
> Associate Director
> Molecular Cytology Core Facility
> University of Missouri
> 120 Bond Life Sciences Center
> Columbia, MO 65211
>
> http://www.biotech.missouri.edu/mcc
>
> (573)882-4895
> (573)884-9676 fax
>

Lukas Hoffmann

Re: Cell counting with imageJ

To All,

I am currently writing just such a program for
autocounting fluorescently stained cells. I will be happy to give you
this program for free when I complete it, if you request.
But, I do not know when it will be complete. I am an undergrad working
20 hrs a week who is going to take time off for christmas break. The project is in its early stages therefore I cannot predict if
I will have success. I am no expert on this subject. But there is no
problem for you if I fail, of course.

My
basic idea is, you count a few images by hand and tell the program what
the count "should" be. Then the program changes the way it does the
thresholding & image processing in many different ways, and
compares all of its test counts with the correct count. When the
program finds one type of image processing that has a low error rate
compared to the manual counting, it uses that type of processing for
the rest of your images.

For this to work, all of your images
must be similarly obtained: same type of cells, same lighting
conditions, etc. If you have significantly different images, then the
tests have to be re run.

If you are interested in collaborating,
please send me some of your images that you have manually counted. I
can use them to refine my program so that it works on my lab's images
and your lab's images.

Lukas Hoffmann
Carnegie Mellon University
Center for the Neural Basis of Cognition
Urban Lab

> Date: Fri, 13 Nov 2009 10:26:21 -0500
> From: [hidden email]
> Subject: Re: Cell counting with imageJ
> To: [hidden email]
>
> Esteban: I think ITCN uses a process somewhat more complex than watershed.
> It does not threshold the image, so it works quite well for counting nuclei
> that are very close together, and even overlapping slightly. As I
> understand, it uses the colvolution and find maxima functions instead. It
> requires a little trial and error in setting up the input parameters
> initially, but then the process moves pretty quickly, and it produces better
> results than I ever managed using thresholding and watershed processing. It
> also marks the image so that you can do some quality control.
>
> Antonis: It is a little picky--make sure the image is open before you try to
> launch ITCN. Also, optimize the input parameters with a smaller ROI--if your
> images are large, it takes a lot of processing power.
>
> Jason
>
>
> On Fri, Nov 13, 2009 at 9:46 AM, G. Esteban Fernandez <
> [hidden email]> wrote:
>
> > You can also use the the built-in functions under Analyze > Analyze
> > Particles to count automatically after thresholding the images. If
> > thresholding doesn't do a good job of including all the nuclei and
> > excluding background then it can get tricky. Good thresholding is the
> > key to good results with automatic counting. Also, if there are
> > nuclei touching each other you can separate them with Process > Binary
> > > Watershed. Maybe the ITCN plugin does the separation automatically?
> >
> > -Esteban
> >
> >
> > On Fri, Nov 13, 2009 at 8:36 AM, Jason Bailey <[hidden email]>
> > wrote:
> > > Antonis,
> > >
> > > I have had much luck with the Image-Based Tool for Counting Nuclei (ITCN)
> > > with mammalian cells. I imagine it could work well with yours as well. I
> > > typically do two counts--first count all nuclei labeled with Hoechst
> > 33342,
> > > then count the nuclei labeled with a different dye (either ethidium or
> > FITC,
> > > usually). Although this might require some changes to how you label your
> > > cells.
> > >
> > > If you must resort to manual counting, the "Cell Counter" plugin marks
> > > cells/nuclei as you click and lets you track multiple cell types as you
> > go.
> > >
> > > Both are available at the ImageJ Plugins page:
> > > http://rsbweb.nih.gov/ij/plugins/index.html
> > >
> > > Good luck!
> > >
> > > Jason
> > >
> > > On Fri, Nov 13, 2009 at 9:25 AM, Antonis Giakountis <[hidden email]>
> > wrote:
> > >
> > >> Hi Esteban,
> > >>
> > >> I really would prefer not to do it by hand. A typical experiment of mine
> > >> involves several timepoints (10-16) and at least five reps per line.
> > Each
> > >> one of those has hundreds of cells and I was hoping for some stitistics
> > out
> > >> of this. To make a long story sort I expect thousants of cells and I
> > would
> > >> like to get an estimation of the percentage of them that has
> > fluorescence
> > >> out of the total.
> > >>
> > >> In the worst case I could deal with user define-clinking as long as I
> > could
> > >> mark the cells that are counted. As for the images I am aiming for
> > >> triple-stain nuclear immunofluorescence, but since this is done with the
> > >> confocal I can easily scan each channel separately, including the
> > >> transmitted-light one (BF).
> > >>
> > >> Antonis
> > >> ----- Original Message ----- From: "G. Esteban Fernandez" <
> > >> [hidden email]>
> > >>
> > >> To: <[hidden email]>
> > >> Sent: Friday, November 13, 2009 2:10 PM
> > >>
> > >> Subject: Re: Cell counting with imageJ
> > >>
> > >>
> > >> How many images/cells do you need to count? Is this something that
> > >>> you can do by hand in a matter of a couple of days or do you want an
> > >>> automated procedure? Often it takes a few days to get an automated
> > >>> procedure to work correctly, if it ever works at all, during which
> > >>> time you could've done the counting by hand. There are probably
> > >>> ImageJ plugins that keep track of counts as you click on cells in the
> > >>> images.
> > >>>
> > >>> Do you have images that match the fluorescence images to get total
> > >>> cell numbers? For example transmitted light (brightfield/phase/DIC)
> > >>> images?
> > >>>
> > >>> -Esteban
> > >>>
> > >>>
> > >>> On Fri, Nov 13, 2009 at 7:43 AM, Antonis Giakountis <[hidden email]>
> > >>> wrote:
> > >>>
> > >>>> Dear all,
> > >>>>
> > >>>> I am a plant Biologist and I would like to use ImageJ in order to
> > count
> > >>>> cell numbers (diameter 10-20 microns on average) and to estimate
> > percentage
> > >>>> of cells that have fluorescence.
> > >>>>
> > >>>> Is it possible to to that with ImageJ and if yes do I have to install
> > a
> > >>>> certain plugin? If not can you recomment another software?
> > >>>>
> > >>>> Thank you in advance for your time,
> > >>>>
> > >>>> Antonis
> > >>>>
> > >>>> --------------------------------------------
> > >>>> Antonis Giakountis, PhD
> > >>>> Prof. David Baulcombe group
> > >>>> Department of Plant Sciences
> > >>>> University of Cambridge
> > >>>> Downing Street
> > >>>> Cambridge CB2 3EA
> > >>>> UK
> > >>>>
> > >>>>
> > >>>
> > >>>
> > >>> --
> > >>> G. Esteban Fernandez, Ph.D.
> > >>>
> > >>> Associate Director
> > >>> Molecular Cytology Core Facility
> > >>> University of Missouri
> > >>> 120 Bond Life Sciences Center
> > >>> Columbia, MO 65211
> > >>>
> > >>> http://www.biotech.missouri.edu/mcc
> > >>>
> > >>> (573)882-4895
> > >>> (573)884-9676 fax
> > >>>
> > >>>
> > >
> >
> >
> >
> > --
> > G. Esteban Fernandez, Ph.D.
> >
> > Associate Director
> > Molecular Cytology Core Facility
> > University of Missouri
> > 120 Bond Life Sciences Center
> > Columbia, MO 65211
> >
> > http://www.biotech.missouri.edu/mcc
> >
> > (573)882-4895
> > (573)884-9676 fax
> >

_________________________________________________________________
Windows 7: I wanted simpler, now it's simpler. I'm a rock star.
http://www.microsoft.com/Windows/windows-7/default.aspx?h=myidea?ocid=PID24727::T:WLMTAGL:ON:WL:en-US:WWL_WIN_myidea:112009

Gabriel Landini

Re: Cell counting with imageJ

In reply to this post by Antonis Giakountis

On Friday 20 November 2009, Lukas Hoffmann wrote:
> My basic idea is, you count a few images by hand and tell the program what
> the count "should" be. Then the program changes the way it does the
> thresholding & image processing in many different ways, and
> compares all of its test counts with the correct count. When the
> program finds one type of image processing that has a low error rate
> compared to the manual counting, it uses that type of processing for
> the rest of your images.

The problem with that approach is that it is counting objects without
knowing whether they are the same in the gold-standard count and in the test
count.

If you have 10 objects (each one is a cell) in the gold standard and only 1 of
those cells is recognised in the test image, but the cell is fragmented by the
segmentation procedure in 10 pieces, you would be assuming that the program is
counting perfectly when it is not.

Counts alone is not enough. You need to know that the segmented blobs somewhat
match (in position, size, shape) the gold standard blobs, only then you could
count them, but otherwise the assumption that they are the same in both images
is not satisfied.

Regards

G.

Lukas Hoffmann

Re: Cell counting with imageJ

G,

You're right, I never thought of that. For the sample images, users could mark the cells' locations with the ImageJ manual cell counter plugin. Then it runs the tests. If there are multiple particles close to the marker (like when cell is fragmented into 10 pieces), those are scored as a duplicate count. If a particle is not close to a marker, that is scored as a false count. Closeness is defined as 'inside the user-set cell radius'. So the tests should measure # of particles counted, number of duplicate counts, number of false counts. It would choose the image processing algorithm that optimizes all three. The user would not have extra work since he has to do some manual counting anyway.

Does that sound better?

Lukas Hoffmann

> Date: Sat, 21 Nov 2009 09:17:39 +0000
> From: [hidden email]
> Subject: Re: Cell counting with imageJ
> To: [hidden email]
>
> On Friday 20 November 2009, Lukas Hoffmann wrote:
> > My basic idea is, you count a few images by hand and tell the program what
> > the count "should" be. Then the program changes the way it does the
> > thresholding & image processing in many different ways, and
> > compares all of its test counts with the correct count. When the
> > program finds one type of image processing that has a low error rate
> > compared to the manual counting, it uses that type of processing for
> > the rest of your images.
>
> The problem with that approach is that it is counting objects without
> knowing whether they are the same in the gold-standard count and in the test
> count.
>
> If you have 10 objects (each one is a cell) in the gold standard and only 1 of
> those cells is recognised in the test image, but the cell is fragmented by the
> segmentation procedure in 10 pieces, you would be assuming that the program is
> counting perfectly when it is not.
>
> Counts alone is not enough. You need to know that the segmented blobs somewhat
> match (in position, size, shape) the gold standard blobs, only then you could
> count them, but otherwise the assumption that they are the same in both images
> is not satisfied.
>
> Regards
>
> G.

_________________________________________________________________
Bing brings you maps, menus, and reviews organized in one place.
http://www.bing.com/search?q=restaurants&form=MFESRP&publ=WLHMTAG&crea=TEXT_MFESRP_Local_MapsMenu_Resturants_1x1

Gabriel Landini

Re: Cell counting with imageJ

On Sunday 22 November 2009, Lukas Hoffmann wrote:

> You're right, I never thought of that. For the sample images, users could
> mark the cells' locations with the ImageJ manual cell counter plugin.
> Then it runs the tests. If there are multiple particles close to the
> marker (like when cell is fragmented into 10 pieces), those are scored as
> a duplicate count. If a particle is not close to a marker, that is scored
> as a false count. Closeness is defined as 'inside the user-set cell
> radius'. So the tests should measure # of particles counted, number of
> duplicate counts, number of false counts. It would choose the image
> processing algorithm that optimizes all three. The user would not have
> extra work since he has to do some manual counting anyway.

I think that more important is to look into how well do the segmented pixel
groups match the expected segmentation (i.e. a ground-truth or gold standard
image).
If I understand the above correctly one risks a single noise pixel very near a
non-detected cell to be accepted as correct.

Have a look here for some interesting ideas (the last part of the paper):
Sezgin, M & Sankur, B (2004), "Survey over Image Thresholding Techniques and
Quantitative Performance Evaluation", Journal of Electronic Imaging 13(1):
146-165.

Cheers

G.