Drift correction software for time-series imaging?

[John_VU]

We are doing time-series measurements for multiple regions of many
wells on a well plate, and have encountered instrument drift when we
return to the same location. Has anyone else tried time series imaging
with repeated visits to different fields? If so, what did you find re
drift? Does anyone have Acapella, Columbus, or Image J software that
has been successful in correcting for this?

[Michelle Ocana]

We have found stage drift when returning to the same position on the
plate. I have been told that the problem lies in the hardware. The
stage does not have a piezo so the location is not exact. It can vary
over time and position so just plugging in a simple XY pixel shift
wouldn't work.

[John Wikswo]

Michelle,

Thanks for getting back to me!  Yes, it does appear to be a hardware problem.

How large are the drifts you see?  Have you been able to make time-series measurements?

Does PE offer any suggestions or explanations?

Where are you located?

John

--
John P. Wikswo
Gordon A. Cain University Professor
A.B. Learned Professor of Living State Physics
Director, Vanderbilt Institute for Integrative Biosystems Research and Education
Professor of Biomedical Engineering, Molecular Physiology & Biophysics, and Physics & Astronomy
Vanderbilt University
VU Station B 351807 (U.S. Postal Service)
  Nashville, TN 37235-1807 USA
6301 Stevenson Center (UPS and Fedex)
  Nashville, TN 37232 USA
Phone: (615) 343-4124 Fax: (615) 322-4977 ...
Email: john....@vanderbilt.edu
http://www.vanderbilt.edu/viibre/Wikswo.html

[John Wikswo]

Michelle -

Another key question is when did you inform PE about these drifts? We were told that they'd never seen them before, and ran tests on our Opera for a month.

Later, we learned that someone else was recently told by PE that they hadn't seen these drifts before, and they would need to do extensive tests.

I would really appreciate any help that you could provide as we try to sort this out.

John

Michelle,

Thanks for getting back to me!  Yes, it does appear to be a hardware problem.

How large are the drifts you see?  Have you been able to make time-series measurements?

Does PE offer any suggestions or explanations?

Where are you located?

John

On 8/2/2011 12:21 PM, Michelle Ocana wrote:

[Emmanuel Gustin]

John,

Are you observing small random variations in position due to the
limited precision of the stage, or a real systematic drift?

We did see serious systematic drift once on our Opera, eventually
accumulating to enough error to make the plate loader crash. That was
solved, after a long troubleshoot, by replacing the electronic cabling
of the stage: Probably the motors were dropping steps because of
microfractures in the wires. (It seems unlikely that you would be able
to do long time series if you had that problem.)

Stage error is hard to compensate for unless you have fixed markers in
the well. You could try adding small fluorescent beads, and hope that
they won't move... But usually the real problem in a long time series
is the motion of the cells, and people have developed cell tracking
algorithms to address that... Maybe these can solve your problem?

Emmanuel