search

soft-matter / trackpy

Python particle tracking toolkit
http://soft-matter.github.io/trackpy
Other
445 stars 131 forks source link

3D Bubbles #624

Open calebclayreagor opened 4 years ago

calebclayreagor commented 4 years ago

I'm wondering what the best practice would be for dealing with 3D bubbles (or in my case, cells). Any guidance would be appreciated.

tacaswell commented 4 years ago

Is your problem primarily in finding them or tracking them?

alexlib commented 4 years ago

@calebclayreagor - it's possible to use openptv to get 3d positions from an optical system and track the results in trackpy. I can explain how. But it depends if you are using 3D optical system with triangulation.

calebclayreagor commented 4 years ago

@tacaswell I'm currently having trouble finding the cells, but I believe I have an acceptable solution. I'm working with z stacks from a confocal microscope (specifically, a structured illumination microscope) over a series of timepoints, and what I need is the position of the cell cytoplasm (approximately a bubble shape) in the xy plane. Due to the 3D positioning of the cells, I have to use different z levels to find the xy locations for different cells in the same series. However, I'm currently interested in tracking only one of the cells, so I will manually select the best z level for each experiment and use the 2D bubble workflow.

If anyone has additional comments on this approach, I'd love to hear them. My lab has been working on segmenting these videos for some time, with less than satisfactory results.

rbnvrw commented 4 years ago

That sounds like an interesting problem. Could you attach a few sample png's so we have a better feeling for what your data looks like?

nkeim commented 4 years ago

If you want to threshold a 3D image and find the centers of mass of the objects, the same way that the bubble tracking tutorial does for 2D images, you might look at scipy.ndimage.label() and the other functions in that module.