ImageJ - Vesicle movement analysis

[Tal Shprung]

Dear all,

I apologize if i repeat a previously asked question, but i was not sure how
to search for the solution.

I load to the program two consecutive images (10sec apart) of a live cell
with its vesicle fluorescently painted.

what i get is two images with many vesicles (round shaped objects).

what i need is to end up with some kind of number to describe the movement
of the vesicle (small movement/ short or long distance/

a fraction of the vesicles moving)

What i currently do is turn the pictures to Binary and adjust threshold to
auto and then subtract the second picture from the first.

I measure the area resulted and divide that from the first picture (time
0sec).

My problem is that by turning the picture to Binary, some close vesicles
appear as one big object.

How can i separate these vesicles? does anyone has a suggestion on a
different way of analysis?

Thank you very much.

Tal Shprung

P.S.

I want to thank the people behind this wonderful (and free) program. You
have made my research possible.

[Martin Jaekel]

Hi Tal,

you could maybe use the watershed algorithm to separate those close
objects. you can find the watershed function in one of the process menu
options i think.

best,
martin

On Feb 20 2008, Tal Shprung wrote:

>Dear all,
>
>I apologize if i repeat a previously asked question, but i was not sure how
>to search for the solution.
>
>I load to the program two consecutive images (10sec apart) of a live cell
>with its vesicle fluorescently painted.
>
>what i get is two images with many vesicles (round shaped objects).
>
>what i need is to end up with some kind of number to describe the movement
>of the vesicle (small movement/ short or long distance/
>
>a fraction of the vesicles moving)
>
>What i currently do is turn the pictures to Binary and adjust threshold to
>auto and then subtract the second picture from the first.
>
>I measure the area resulted and divide that from the first picture (time
>0sec).
>
>My problem is that by turning the picture to Binary, some close vesicles
>appear as one big object.
>
>How can i separate these vesicles? does anyone has a suggestion on a
>different way of analysis?
>
>Thank you very much.
>
>Tal Shprung
>
>P.S.
>
>I want to thank the people behind this wonderful (and free) program. You
>have made my research possible.
>

[Tal Shprung]

Hi Martin,

The watershed is not enough, it does seperate the objects but very slightly.

I am looking for something that, still in the colored picture, will find the
strongest points within each vesicle's center of mass and decress the
vesicle's size a bit, to seperate the vesicles.

On Feb 20, 2008 10:24 AM, M. Jaekel <[hidden email]> wrote:

> Hi Tal,
>
> you could maybe use the watershed algorithm to separate those close
> objects. you can find the watershed function in one of the process menu
> options i think.
>
> best,
> martin
>
> On Feb 20 2008, Tal Shprung wrote:
>
> >Dear all,
> >
> >I apologize if i repeat a previously asked question, but i was not sure
> how
> >to search for the solution.
> >
> >I load to the program two consecutive images (10sec apart) of a live cell
> >with its vesicle fluorescently painted.
> >
> >what i get is two images with many vesicles (round shaped objects).
> >
> >what i need is to end up with some kind of number to describe the
> movement
> >of the vesicle (small movement/ short or long distance/
> >
> >a fraction of the vesicles moving)
> >
> >What i currently do is turn the pictures to Binary and adjust threshold
> to
> >auto and then subtract the second picture from the first.
> >
> >I measure the area resulted and divide that from the first picture (time
> >0sec).
> >
> >My problem is that by turning the picture to Binary, some close vesicles
> >appear as one big object.
> >
> >How can i separate these vesicles? does anyone has a suggestion on a
> >different way of analysis?
> >
> >Thank you very much.
> >
> >Tal Shprung
> >
> >P.S.
> >
> >I want to thank the people behind this wonderful (and free) program. You
> >have made my research possible.
> >
>

[Michael Schmid]

Hi Tal,

you could try Process>Binary>Find Maxima with a noise tolerance
that you have to find by trial and error (use preview).

You can also convert the image to grayscale (or use one color
channel that has the best contrast). In a grayscale image you
can set a threshold to suppress background objects in
Find Maxima.

Michael
________________________________________________________________

On 20 Feb 2008, at 13:08, Tal Shprung wrote:

> Hi Martin,
>
> The watershed is not enough, it does seperate the objects but very  
> slightly.
>
> I am looking for something that, still in the colored picture, will  
> find the
> strongest points within each vesicle's center of mass and decress the
> vesicle's size a bit, to seperate the vesicles.
>
> On Feb 20, 2008 10:24 AM, M. Jaekel <[hidden email]> wrote:
>
>> Hi Tal,
>>
>> you could maybe use the watershed algorithm to separate those close
>> objects. you can find the watershed function in one of the process  
>> menu
>> options i think.
>>
>> best,
>> martin
>>
>> On Feb 20 2008, Tal Shprung wrote:
>>
>>> Dear all,
>>>
>>> I apologize if i repeat a previously asked question, but i was  
>>> not sure
>> how
>>> to search for the solution.
>>>
>>> I load to the program two consecutive images (10sec apart) of a  
>>> live cell
>>> with its vesicle fluorescently painted.
>>>
>>> what i get is two images with many vesicles (round shaped objects).
>>>
>>> what i need is to end up with some kind of number to describe the
>> movement
>>> of the vesicle (small movement/ short or long distance/
>>>
>>> a fraction of the vesicles moving)
>>>
>>> What i currently do is turn the pictures to Binary and adjust  
>>> threshold
>> to
>>> auto and then subtract the second picture from the first.
>>>
>>> I measure the area resulted and divide that from the first  
>>> picture (time
>>> 0sec).
>>>
>>> My problem is that by turning the picture to Binary, some close  
>>> vesicles
>>> appear as one big object.
>>>
>>> How can i separate these vesicles? does anyone has a suggestion on a
>>> different way of analysis?
>>>
>>> Thank you very much.
>>>
>>> Tal Shprung
>>>
>>> P.S.
>>>
>>> I want to thank the people behind this wonderful (and free)  
>>> program. You
>>> have made my research possible.
>>>
>>

[Tal Shprung]

Thank you very much, i will try that

Tal

On Feb 20, 2008 2:35 PM, Michael Schmid <[hidden email]> wrote:

> Hi Tal,
>
> you could try Process>Binary>Find Maxima with a noise tolerance
> that you have to find by trial and error (use preview).
>
> You can also convert the image to grayscale (or use one color
> channel that has the best contrast). In a grayscale image you
> can set a threshold to suppress background objects in
> Find Maxima.
>
> Michael
> ________________________________________________________________
>
> On 20 Feb 2008, at 13:08, Tal Shprung wrote:
>
> > Hi Martin,
> >
> > The watershed is not enough, it does seperate the objects but very
> > slightly.
> >
> > I am looking for something that, still in the colored picture, will
> > find the
> > strongest points within each vesicle's center of mass and decress the
> > vesicle's size a bit, to seperate the vesicles.
> >
> > On Feb 20, 2008 10:24 AM, M. Jaekel <[hidden email]> wrote:
> >
> >> Hi Tal,
> >>
> >> you could maybe use the watershed algorithm to separate those close
> >> objects. you can find the watershed function in one of the process
> >> menu
> >> options i think.
> >>
> >> best,
> >> martin
> >>
> >> On Feb 20 2008, Tal Shprung wrote:
> >>
> >>> Dear all,
> >>>
> >>> I apologize if i repeat a previously asked question, but i was
> >>> not sure
> >> how
> >>> to search for the solution.
> >>>
> >>> I load to the program two consecutive images (10sec apart) of a
> >>> live cell
> >>> with its vesicle fluorescently painted.
> >>>
> >>> what i get is two images with many vesicles (round shaped objects).
> >>>
> >>> what i need is to end up with some kind of number to describe the
> >> movement
> >>> of the vesicle (small movement/ short or long distance/
> >>>
> >>> a fraction of the vesicles moving)
> >>>
> >>> What i currently do is turn the pictures to Binary and adjust
> >>> threshold
> >> to
> >>> auto and then subtract the second picture from the first.
> >>>
> >>> I measure the area resulted and divide that from the first
> >>> picture (time
> >>> 0sec).
> >>>
> >>> My problem is that by turning the picture to Binary, some close
> >>> vesicles
> >>> appear as one big object.
> >>>
> >>> How can i separate these vesicles? does anyone has a suggestion on a
> >>> different way of analysis?
> >>>
> >>> Thank you very much.
> >>>
> >>> Tal Shprung
> >>>
> >>> P.S.
> >>>
> >>> I want to thank the people behind this wonderful (and free)
> >>> program. You
> >>> have made my research possible.
> >>>
> >>
>