An evaluation of thresholding techniques in fMRI analysis

• Published in: NeuroImage (Orlando, Fla.) Fla.), 2004-05, Vol.22 (1), p.95-108
• Main Authors: Logan, Brent R., Rowe, Daniel B.
• Format: Article
• Language: English
• Subjects: Algorithms; Bayes Theorem; Benjamini–Hochberg procedure; Bonferroni procedure; Computer Simulation; Experiments; False discovery rate; Family-wise error rate; Fingers - physiology; Humans; Hypotheses; Image Processing, Computer-Assisted - statistics & numerical data; Magnetic Resonance Imaging - statistics & numerical data; Movement - physiology; Permutation; Random fields; Random variables; Resampling; Spatial correlation; Statistics; Studies
• ISSN: 1053-8119; 1095-9572
• DOI: 10.1016/j.neuroimage.2003.12.047

This paper reviews and compares individual voxel-wise thresholding methods for identifying active voxels in single-subject fMRI datasets. Different error rates are described which may be used to calibrate activation thresholds. We discuss methods which control each of the error rates at a prespecified level α, including simple procedures which ignore spatial correlation among the test statistics as well as more elaborate ones which incorporate this correlation information. The operating characteristics of the methods are shown through a simulation study, indicating that the error rate used has an important impact on the sensitivity of the thresholding method, but that accounting for correlation has little impact. Therefore, the simple procedures described work well for thresholding most single-subject fMRI experiments and are recommended. The methods are illustrated with a real bilateral finger tapping experiment.