Area and perimeter of cells

cellprofiler

#1

Dear Team,

In the nuclear translocation pipeline, the output file provides with lot of measurements.
Two of them are Area and Perimeter.

What I need to know is that nucleus is a 3D object, so it has to be a sphere, not a circle.
And sphere doesn’t have area/perimeter. It has curved surface area and volume.
So, what does area & perimeter of nuclei represents?
What formula is it using for calculation?

Kindly let me know so that I can take care of these issues.
Thanks
Mridul KK


#2

Hi,

Unless you’re using a specialized microscope to take Z-stacks (such as a confocal), your typical microscope image will be a planar cross-section of the object of interest. In these cases, the quality of the image will be dependent on the thickness of the focal plane.

Therefore, any morphological parameters will necessarily be a 2-D measurement. So when you see area, perimeter, etc mentioned, it is implicitly assumed to be cross-sectional area, cross-sectional perimeter, etc.

Hope this helps,
-Mark


#3

Thanks Mark for the explaination.

So, it means that for confocal microscopy, I can take the area as ‘curved surface area (4pieradiusradius)’ instead of 'cross-sectional area (pieradius*radius)’. Latter is usualy the case with other microscopes…right?
So, is ‘volumetric (3D)’ measurement is possible for confocal images as compared to normal perimeter value?

will like to know the answers of these.
Thanks once again.
Mridul KK


#4

If you have 3D images (that is, Z-stacks) and you use a 3D nucleus-identification algorithm (not currently available in CellProfiler) then you will be able to produce 3D measurements, such as volume.

But, given 2D image data (whether confocal or not), the only thing you can calculate is how much 2D area is occupied in the image by each nucleus, and what is the 2D perimeter around the nucleus within the image slice you are analyzing. So, these are the measurements that CellProfiler produces. If you believe that your nuclei are spheres (or some other particular shape) and that you are focused through the center of the nuclei in your images, you could extrapolate the volume measurements based on the area measurement using simple math - but of course it’s a big assumption to make that the nuclei really are perfectly spherical! So depending on your biological question, you probably simply want to report the 2D measurements rather than extrapolate 3D measurements based on questionable assumptions.

Hope this clarifies,
Anne


#5

Thanks Anne.
I understand what you are emphasizing on. Its clear to me.
Mridul KK