Connectome Operations For FSL ExEcution (COFFEE): a turnkey pipeline for preprocessing of fMRI data

Numerous methods exist to analyze functional MRI (fMRI) data, but no software currently exists to enhance the commonly-used FSL statistical analysis software with preprocessing methods from the Human Connectome Project. Here we developed the Connectome Operations For FSL ExEcution (COFFEE) pipeline to integrate Human Connectome preprocessing, including Freesurfer brain extraction and registration, for FSL general linear model analysis of volumetric fMRI data. We tested COFFEE on a real fMRI dataset (N=26) wherein participants performed a precision drawing task during fMRI scanning. COFFEE preprocessing, compared to traditional FSL preprocessing, led to lower inter-individual variability across the brain, more precise brain extraction, and greater detected activation in sensorimotor areas contralateral to movement. The COFFEE pipeline provides a turnkey preprocessing alternative for FSL users with substantial advantages for analysis of task fMRI data. COFFEE software is available online for research use and further testing. ### Competing Interest Statement Author BAP and Washington University in St. Louis have a licensing agreement with PlatformSTL to commercialize the drawing program used in this study.