Optimizing data processing to improve the reproducibility of single‐subject functional magnetic resonance imaging

Abstract

Currently, the most important clinical application of functional magnetic resonance imaging (fMRI) is presurgical planning to prevent functional deficits (Bookheimer, 2007; Stippich, Blatow, & Garcia, 2015). fMRI data can be used to influence the surgical management of patients by determining the necessity of intraoperative mapping, the necessary extent of brain exposure, and the safest surgical route (Morrison et al., 2016). Achieving fMRI activation maps with high reliability is critical to preserving eloquent motor and language regions during presurgical planning (Nettekoven, Reck, Goldbrunner, Grefkes, & Lucas, 2018). Despite nearly three decades of research in fMRI, there is little standardization for imaging procedures, data processing, and interpretation of the results (Stippich et al., 2015). Recently, the American Society of Functional Neuroradiology (ASFNR) has published a white paper on recommended paradigms for presurgical language assessment (Black et al.

Evaluating the efficacy of multi-echo ICA denoising on model-based fMRI

“…Beyond R 2 , it is important that preprocessing and analysis choices lead to reproducible outcomes ( Botvinik-Nezer et al, 2020 ;Bowring et al, 2019 ;Lerma-Usabiaga et al, 2020 ;Soltysik, 2020 ). In our study, we leveraged the relative stability of voxel-wise retinotopic tuning to test whether ME-ICA led to similar parameter estimates from independent runs of data ( van Dijk et al, 2016 ).…”

Evaluating the efficacy of multi-echo ICA denoising on model-based fMRI

“…Beyond R 2 , it is important that preprocessing and analysis choices lead to reproducible outcomes (Botvinik-Nezer et al, 2020; Bowring et al, 2019; Lerma-Usabiaga et al, 2020; Soltysik, 2020). In our study, we leveraged the relative stability of voxel-wise retinotopic tuning to test whether ME-ICA led to similar parameter estimates from independent runs of data (van Dijk et al, 2016).…”