I want to know what the smallest defect a particular CT system is capable of imaging, for example a small hole in a phantom. As I understand it, the spatial resolution of a CT system is the smallest separation at which two points can be distinguished as separate entities. Therefore as long as the spatial resolution is say an order lower than the diameter of the hole then I think I should be able to see a fair representation of the hole. From what I've read the spatial resolution of a CT system is determined from the modulation transfer function (MTF). I don't see how, once the MTF graph has been produced (and I presume there is software which can do this?), you can read the spatial resolution from it? I expected you would just return a distance x? I have a reasonable but by no means complete understanding of what's going on here and any further explanation would be greatly appreciated. I am engineering student who is using micro-CT to analyse defects in steels, resources for industrial CT are very limited so I hoped to find some help from the medical world! Ultimately I want to know how can I quantify what the smallest hole I could hope to see is. I would be interested to know as well how the spatial resolution relates to the pixel size in the scan image.