I think that the question you should be asking is not so much will my reconstruction be consistent, but will my *scans* be consistent. (I'm not familiar with that particular reconstruction software, but assuming you reconstruct in the same way, you will have consistent reconstructions.) As to things like ring artifacts, and beam hardening... these are dependent on your scanner, scan parameters, corrections being used, and your sample (size, thickness, density, etc) itself. Your CT scanner operator should be your first contact, I think.

Magdalena Gill Thanks. I referred to my data as "reconstructed" 2D data because they were not the output of the micro-CT device, those slices were "reconstructed" in NRecon from the device output (shadow images). Maybe I did not use a good terminology though, I am not sure about that. I know that the scan parameters were the same, however, within NRecon, we need to set the parameters such as beam hardening, ring artifact, etc before reconstructing the slices. So, I was wondering whether the parameters set in NRecon should be exactly the same ( i.e., misalignment compensation, ring artifacts reduction and beam hardening correction ).

Your terminology was correct in terms of what you call reconstructed - the CT scanner raw output is just radiographs. Most scanners come with their own propriotary reconstruction software that turns those radiographs into a 3D volume (and allows you to export that volume as 2D slices for further processing.) However, once you are within your reconstruction software - NRecon in your case - whatever parameters you are setting aren't *scan* parameters, they are *reconstrucion* parameters. And those *may* need to be different for different samples and different scans for optimal results. For instance: beam hardening may vary betweeen scans if the density of your sample is different, or if the scans used different X-ray energy settings. In such a case you would need to use different settings in NRecon to optimize your reconstruction in each case. So when I say scan parameters, I mean SCAN parameters - the kv, mA, magnification, number of radiographs, length of scan, number of averaged radiographs, geometric unsharpness, etc. None of these things are set in NRecon, but instead are set before the CT scan is performed. Finally, your sample matters too - the same parameters used on a dense sample and a less dense one, or one that is homogenous and one that is heterogenous, may yield different scans. This is why I suggest checking in with the person conducting the scans (as in, running the CT machine) to check how consistent their scans are. If they ARE consistent, then you *probably* can use the same NRecon settings for each scan. If they aren't, however, you will most certainly need to optimize each scan individually, which I understand may bring into question how comparable they will be.

Thank you Magdalena Gill .Your answer is very clear.