Ghost Cells

[Bryant Walker]

Hello,

I am using MicrobeTracker to analyze some E. Coli cells in some microfluidic chips with dead end pockets. The channels/pockets are somewhat of a different contrast to the background, but the cells in the pockets are still very distinct. However when using MicrobeTracker ghost cells, contours that are produced around nothing are produced, actually many of these ghost cells are produced. I have tried using different parameters but still get these ghost cell contours. Now I do get the right contours around the actual cells, but the ghost cell contours just slow the program down and make it have to run for 3 hours instead of maybe couple of minutes for just the cells. So I am wondering if anyone else is having the same problem with getting ghost cell contours and/ or how to possibly fix this so my analysis time will not take 3 hours to run through 300 images? 

Thanks

Bryant Walker