Re: correlating positions of wondering ROIs

[dpoburko]

Hi All,

   I am looking for some hand waiving suggestions to approach a problem.
I am imaging fluorescent puncta that move around a little bit over time.
As I automatically segment these puncta in each image, I am left with a
Constellation of ROIs. Over time, some ROIs come and go, but most remain
detected. However the extact center of mass coordindates move around a
little over time. Part of this is just because the location of the
fluorophores can move. But the general orientation of a given puncta
within the constellation of say 50 - 900 puncta remains largely the
same. This might be comparable to looking at protein spots on 2D gels,
etc. Can anyone suggests a reasonable method for tracking puncta/spots
over a series of a few images and assigning them IDs. In the end, my
goal is to generate an ROI set from each set of puncta and be able to
known which puncta is which over a series of ROI sets in light of their
slight movement (i.e. changing center of mass co-ordinates). I hope this
was relatively clear.

Many thanks for any suggestions,
Damon

--

Damon Poburko, PhD
Postdoctoral Research Fellow
Stanford University School of Medicine
Dept. of Molecular & Cellular Physiology
279 Campus Dr., Beckman B103, Stanford, CA 94305
Ph: 650 725 7564, fax: [phone]

[Gluender-3]

>Hi All,
>
>   I am looking for some hand waiving suggestions to approach a
>problem. I am imaging fluorescent puncta that move around a little
>bit over time. As I automatically segment these puncta in each
>image, I am left with a Constellation of ROIs. Over time, some ROIs
>come and go, but most remain detected. However the extact center of
>mass coordindates move around a little over time. Part of this is
>just because the location of the fluorophores can move. But the
>general orientation of a given puncta within the constellation of
>say 50 - 900 puncta remains largely the same. This might be
>comparable to looking at protein spots on 2D gels, etc. Can anyone
>suggests a reasonable method for tracking puncta/spots over a series
>of a few images and assigning them IDs. In the end, my goal is to
>generate an ROI set from each set of puncta and be able to known
>which puncta is which over a series of ROI sets in light of their
>slight movement (i.e. changing center of mass co-ordinates). I hope
>this was relatively clear.
>
>Many thanks for any suggestions,
>Damon
>
>--
>
>Damon Poburko, PhD
>Postdoctoral Research Fellow
>Stanford University School of Medicine
>Dept. of Molecular & Cellular Physiology
>279 Campus Dr., Beckman B103, Stanford, CA 94305
>Ph: 650 725 7564, fax: [phone]

You may have a look at

<http://www.gluender.de/Writings/WritingsTexts/HardText.html#Gl-1991-2>

where I prove the identity of the difference vector between the
centres of gravity of two arbitrary n-dimensional density functions,
with the centroid vector of their cross-correlation function.

Maybe its of some help.

Best
--

                   Herbie

          ------------------------
          <http://www.gluender.de>