Mesh refinement and timelapses

[Carlos Toscano Ochoa]

Hello there!

I'm trying to dominate microbeTracker for E.coli timelapse movies purposes. And I have the next problem: When the analysis is done, I realize that a lot of incorrect meshes are because the meshes from previous frames accumulate errors. To avoid this, I have to select all the cells in a frame, clic on "Refine" once or twice, and then analyze the next frame. Couldn't this be done automatically?? I mean, each time a frame is analyzed in a timelapse movie, Refine script should be run one or two times over all detected cells before proceed to the next frame. How can I do that??

I really want to forget about Schnitzcells or similar programs and I've being giving a chance to microbeTracker for more than want week. I want this to work, and use it for the next papers!!

Thanks!!

[Seán Murray]

This should be largely equivalent to increasing the parameter fitMaxIter.

Try also increasing fsmooth if it's too low.

Ensure drift has been corrected for and if the cells move too much between frames, have a look at moveall and fitStep1.

Sean

[Carlos Toscano Ochoa]

Hi! I've set fitMaxIter to 2000, fsmooth to 200, check the shift between frames and played with moveall and fitStep1. Nothing of that worked, but...

Changing forceWeights = 0.1 4 4, resulted in the perfect fit to the shapes. So, these are not relative forces (as [0.1 0.1 0.1] is not the same as [1 1 1]) so decreasing the first variable and increasing the other two, resulted in a perfect fit to the phase images of E.coli. I'll paste the whole parameters for E.coli to share with the community as soon as I check they can track the population in the correct way.

Carlos Toscano-Ochoa