A novel group analysis for functional MRI of the human brain based on a two-threshold correlation (TTC) method

Tibor Auer, Attila Schwarcz, Tamas Doczi, Klaus Dietmar Merboldt, Jens Frahm

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

This work presents a new group analysis for functional MRI of human brain activation. The two-threshold correlation (TTC) method determines two statistical thresholds by estimating the noise distribution underlying the summed histogram of correlation coefficients (CC) from all sections and subjects. The probabilistic CC thresholds (p < 0.0001 for the identification of highly significant activation centers and p < 0.05 for limiting the iterative addition of directly neighboring voxels to these centers) are applied to the group CC maps for each section. These maps may be reconstructed by taking the maximum (MAX) or mean (MEAN) CC value of all subjects for a particular voxel. Experimental analyses involved functional echo-planar imaging of sequential finger-to-thumb opposition and silent word generation at 3 T (eight subjects). Preprocessing included motion correction, spatial filtering, and normalization to MNI space. While the results for the TTC MAX approach were very similar to those obtained for a standard SPM analysis, the TTC MEAN approach turned out to be more conservative emphasizing voxels that are activated in most rather than in only a few subjects. The new method is simple, fast, and robust by linking two thresholds in a physiologically meaningful manner.

Original languageEnglish
Pages (from-to)335-339
Number of pages5
JournalJournal of Neuroscience Methods
Volume167
Issue number2
DOIs
Publication statusPublished - Jan 30 2008

Keywords

  • Functional MRI
  • Group analysis
  • SPM

ASJC Scopus subject areas

  • Neuroscience(all)

Fingerprint Dive into the research topics of 'A novel group analysis for functional MRI of the human brain based on a two-threshold correlation (TTC) method'. Together they form a unique fingerprint.

  • Cite this