Share Email Print

Proceedings Paper

Tree-dependent and topographic independent component analysis for fMRI analysis
Author(s): Oliver Lange; Anke Meyer-Base; Axel Wismueller; Monica Hurdal; DeWitt Sumners; Dorothee P. Auer
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Exploratory data-driven methods such as unsupervised clustering and independent component analysis (ICA) are considered to be hypothesis-generating procedures, and are complementary to the hypothesis-led statistical inferential methods in functional magnetic resonance imaging (fMRI). Recently, a new paradigm in ICA emerged, that of finding “clusters” of dependent components. This striking philosophy found its implementation in two new ICA algorithms: tree-dependent and topographic ICA. For fMRI, this represents the unifying paradigm of combining two powerful exploratory data analysis methods, ICA and unsupervised clustering techniques. For the fMRI data, a comparative quantitative evaluation between the two methods, tree-dependent and topographic ICA was performed. The comparative results were evaluated by (1) task-related activation maps, (2) associated time-courses and (3) ROC study. It can be seen that topographic ICA outperforms all other ICA methods including tree-dependent ICA for 8 and 9 ICs. However, for 16 ICs topographic ICA is outperformed by both FastICA and tree-dependent ICA (KGV) using an approximation of the mutual information the kernel generalized variance.

Paper Details

Date Published: 12 April 2004
PDF: 12 pages
Proc. SPIE 5439, Independent Component Analyses, Wavelets, Unsupervised Smart Sensors, and Neural Networks II, (12 April 2004); doi: 10.1117/12.541779
Show Author Affiliations
Oliver Lange, Florida State Univ. (United States)
Ludwig-Maximilians-Univ. Muenchen (Germany)
Anke Meyer-Base, Florida State Univ. (United States)
Axel Wismueller, Florida State Univ. (United States)
Ludwig-Maximilians-Univ. Muenchen (Germany)
Monica Hurdal, Florida State Univ. (United States)
DeWitt Sumners, Florida State Univ. (United States)
Dorothee P. Auer, Max Planck Institute of Psychiatry (Germany)

Published in SPIE Proceedings Vol. 5439:
Independent Component Analyses, Wavelets, Unsupervised Smart Sensors, and Neural Networks II
Harold H. Szu; Mladen V. Wickerhauser; Barak A. Pearlmutter; Wim Sweldens, Editor(s)

© SPIE. Terms of Use
Back to Top