next up previous
Next: Introduction

Tensorial Extensions of Independent Component Analysis for Multi-Subject FMRI Analysis

FMRIB Technical Report TR04CB1
(A related paper has been accepted for publication in NeuroImage)

Christian F. Beckmann and Stephen M. Smith

Oxford Centre for Functional Magnetic Resonance Imaging of the Brain (FMRIB),
Department of Clinical Neurology, University of Oxford, John Radcliffe Hospital,
Headley Way, Headington, Oxford, UK
Corresponding author: beckmann [at]


We discuss model-free analysis of multi-subject or multi-session FMRI data by extending the single-session Probabilistic Independent Component Analysis model (PICA; [Beckmann and Smith, 2004]) to higher dimensions. This results in a three-way decomposition which represents the different signals and artefacts present in the data, in terms of their temporal, spatial and subject-dependent variations. The technique is derived from and compared with Parallel Factor Analysis (PARAFAC; [Harshman and Lundy, 1984]). Using simulated data as well as data from multi-session and multi-subject FMRI studies we demonstrate that the tensor-PICA approach is able to efficiently and accurately extract signals of interest in the spatial, temporal and subject/session domain. The final decompositions improve upon PARAFAC results in terms of greater accuracy, reduced interference between the different estimated sources (reduced cross-talk), robustness (against deviations of the data from modelling assumptions and against overfitting) and computational speed. On real FMRI 'activation' data, the tensor-PICA approach is able to extract plausible activation maps, time courses and session/subject modes as well as provide a rich description of additional processes of interest such as image artefacts or secondary activation patterns. The resulting data decomposition gives simple and useful representations of multi-subject/multi-session FMRI data that can aid the interpretation and optimisation of group FMRI studies beyond what can be achieved using model-based analysis techniques.

Keywords: Tensor decomposition; Independent Component Analysis; PARAFAC; functional magnetic resonance imaging;

next up previous
Next: Introduction
Christian Beckmann 2004-12-14