Now I'm working on a project of trophoblast stem cells with my colleague. I got some images from him but I have some troubles with segmenting the cell shape based on e-cadherin marker. With e-cadherin we could have the membrane information for the cells, I was trying to get the cell shape with it. However it seems that the result is not so perfect as I expected. Here I attached the images and pipeline I'm using now.
In the folder, I attached the 3d stack images of DAPI and e-cadherin, and a picture of the result from my pipeline (the red channel is membrane and the green channel is the segmented cell shape, and I was using the slice no.13 to no.19 to make an averaged projection for analysis). As we can see fromthe picture, some of the cells are well segmented but not all of them. I want to know are there any tricks to improve this result? Or maybe there will be other methods for this kind of segmentation?
Thanks in advance.