Hi all, I have two fMRI series of the same subject with different TRs (0.72s, 1s). I wonder how can I resample / interpolate one scan to match the other?
I would strongly suggest that each series is analyzed separately. You can then conduct a higher level analysis that looks for common signal. If you have many datasets like this, you could add the TR value as a nuisance regressor.
Contrast
When you change the TR of a MRI scan, you are also changing the relative brightness of different tissues (gray matter, white matter, css, etc.). This is because the longer you wait between excitation, the more recovery you will have. In MRI, this longitudinal relaxation is referred to as T1:
Aliasing
The sampling rate directly impacts or Nyquist frequency, and we need to be aware of aliasing.
Thanks for your detailed explanation. But currently I just want to do a temporal alignment between the two in order to synchronize some other behavior measure (i.e. scan 1 with TR=0.72s has the temporal behavior data I want to analyze, while scan 2 with TR=1s doesn’t have that data, I want to get some idea of how that behav. data might look like in scan 2). From what I learnt, it’s related to harmonization techniques. But for now, I just want something simple and quick.
I guess I will just go head using spline interpolation if there are no better practices.