Alternative-Based Thresholding for Pre-Surgical fMRI

• Activity Number: 362
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Imaging
• Author(s): Beatrijs Moerkerke*+ and Joke Durnez and Andreas Bartsch and Thomas Nichols
• Companies: Ghent University and Department of Data Analysis, Ghent University, Belgium and Department of Neuroradiology, University of Heidelberg, Heidelberg, Germany and University of Warwick
• Keywords: pre-surgical fMRI ; multiple testing ; alternative p-value ; false negatives

Abstract

fMRI is a non-invasive imaging technique that enables to locate task related areas in the brain. The procedure plays a major role in pre-surgical planning for patients with resectable brain lesions such as tumors.

For fMRI data analysis, the brain is divided in a huge number of volume units (voxels) and for each voxel, a statistical test is performed to decide on its activation status. In cognitive neurosciences, focus lies on stringent control of the false positive rate to account for the huge multiple testing problem. However, false negatives can be detrimental in clinical settings as they may lead to surgical resection of vital brain tissue. Consequently, we argue for a testing procedure with a stronger focus on preventing false negatives.

We present a thresholding procedure that combines 2 measures of significance for each voxel: a classical p-value which reflects evidence against the null hypothesis of no activation and an alternative p-value which reflects evidence against activation with a pre-specified size. This results in a layered statistical map for the brain where each layer measures evidence against the null, the alternative or both.