fMRI Time Series - Signal and Noise Modelling
Steve Smith
Oxford University
FMRIB
In the first part of the talk, I will discuss the issue of autocorrelation in fMRI time series (“smoothness in the noise”) and describe an approach for modeling this in a flexible and robust manner, separately at each voxel.
In the second part, I will present recent advances in modeling the hemodynamic response function using constrained, optimized basis functions (iABF), which allow more efficient temporal fitting of the fMRI signal.
