AUTHOR=Yue Kun , Webster Jason , Grabowski Thomas , Shojaie Ali , Jahanian Hesamoddin 

TITLE=Iterative Data-adaptive Autoregressive (IDAR) whitening procedure for long and short TR fMRI

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 18 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2024.1381722

DOI=10.3389/fnins.2024.1381722

ISSN=1662-453X

ABSTRACT=Functional magnetic resonance imaging (fMRI) has become a fundamental tool for studying brain function. However, the presence of serial correlations in fMRI data complicates data analysis and can lead to incorrect conclusions in some analysis pipelines. Traditional whitening procedures designed for data of longer repetition times (TRs) are inadequate for the increasing use of short-TR fMRI data. This study addresses the shortcomings of existing whitening methods by introducing an iterative whitening approach named "IDAR" (Iterative Data-adaptive Autoregressive model). IDAR employs high-order autoregressive (AR) models with flexible and data-driven orders, offering the capability to model complex serial correlation structures in both short-TR and long-TR fMRI datasets. We thoroughly demonstrate the effectiveness of IDAR by evaluating the residual serial correlations after whitening, examining type-I error rates, and assessing the statistical power in both task-based and resting-state fMRI settings.