Re: standard deviation in an analysed image

Posted by sebastian on Feb 07, 2015; 5:09am
URL: http://imagej.273.s1.nabble.com/standard-deviation-in-an-analysed-image-tp5011470p5011496.html

dear herbie, good day.

my question is, when i take a stacked image through confocal for the cells. what i do is, i try to see the max intensity of the stack and then create an roi around each cells ( for eq i have 20 cells in an image) and for each cells i perform measure. this result gives me a standard deviation and integrated density along with many other measures for 20 cells. this standard deviation what i understand is the deviation of all the pixels from the mean gray value for a cell. therefore in total if i consider all the 20 cells and i bypothetically calculate an average integrated density for the expression of my protein in one cell as it is a highly hetergenous system. hence my question is, do i need to take an average of the standard deviation shown in this steps for 20 cells and mark that in my plot for the integrated density of the image or do i need to recalculate the deviation manually in 20 cells. the problem is if i manually calculate the deviation, the result is tooo high for the standard deviation as it is a highly heterogenous system. but the p value in between the result is significant so if i show high deviation, is it correct? what standard deviation has to be plotted in the graph for integrated density of the image.

Regards,
Sebastian John