Alignment of time lapse microscopy

[dwilmaerts]

Hi Oufti team,

I am trying to use Oufti to analyze time lapse microscopy of E. coli on LB agar pads (phase contrast). Images are taken every 15 min for approximately 10 -16 h. I am now using a test set of 5 frames, starting from t=0. In these test frames, cells elongate strongly but do not divide. 

I am running into some issues:

-          When I want to align the frames (there is a minor xy drift), oufti shuts down.

-          If I 'detect and analyse' “all frames”, segmentation does not work properly. If I run the analysis independent (this frame), segmentation works fine. Is this the result of the align-issue?

Thanks in advance!

Dorien

[Oufti]

Hi Dorien,

With regard to your issues: 

-When running the alignment, does this happen during the automatic or manual alignment? And what operating system do you use to run Oufti?

-Your problems with running Oufti across all frames likely stem from two issues. First, Oufti uses the previous frame in time-lapse mode to detect cells in the current one. In case of incorrectly detected cells, this can lead to error propagation across multiple frames. This can be overcome by splitting up your time-lapse analysis into shorter segments (by specifying the frames to be analyzed) and correcting the segmentation for each segment separately. Second, as you are dealing with elongated cells, standard Oufti parameters likely are inadequate, so make sure you optimize those (for example, on a single frame with elongated cells) before running your entire time-lapse analysis.

Hope this helps!

Thanks for using Oufti!

Oufti-team

Op woensdag 9 februari 2022 om 11:06:31 UTC+1 schreef dwilmaerts: