Would like to manually calculate surface areas

I have a rather unusual request and being a blatant newbie (I’ve used CellProfiler before to produce quantitative live-dead profiles, very successfully, but this I’m finding slightly more challenging), forgive me if the solutions are obvious.

I am using an interferometer to map the cell heights of dehydrated (prepared as for SEM) osteoblasts cultured on glass. While I’m quite happy with the manner in which the instrument software produces height data, I’m also interested in the surface area covered by the cells and there it is considerably less helpful. It outputs .tif screencaps of the cell, colour coded by height. Unfortunately, there’s sample-to-sample variation in the height scheme, so there doesn’t appear to be a simple way of getting CellProfiler to identify just the cell. I can crop the image to include only the relevant parts and can draw the outline of the cell, using the IdentifyObjectsManually module on CellProfiler, but after that, I start getting a little stuck. I know exactly how large the image is and ideally, I’d like to be prompted to enter a pixel -> micron scaling factor, cut out only the area I’m interested in and have CellProfiler calculate the area covered. Even more ideally, I’d love the eccentricity of the object too.

Kindly find attached an example image and a crude initial pipeline I have. Suggestions gratefully received.


Crude Interferometry Measurement.cp (2.81 KB)

Just answered my own question regarding the scaling, thanks to the FAQ. The rest is still a work-in-progress.


Hi Naa-Dei,

I wonder if you might have better luck if you output the screenshot as a grayscale intensity image, rather than color. CellProfiler is optimized for fluorescence images (light objects on dark background) so if you can get the image in that form, IdentifyPrimaryObjects might have a better shot at detecting the cells without manual intervention. Otherwise, the measurement modules you included should work just fine.