Continuous carry-over fMRI experiments present stimuli in an unbroken, sequential manner, and can be used to estimate simultaneously the mean difference in neural activity between stimuli (for the purpose of distributed pattern analysis) as well as the effect of one stimulus upon another (carry-over effects). Continuous carry-over designs with serially balanced sequences are particularly well suited to the characterization of "similarity spaces," in which the perceptual similarity of stimuli is related to the structure of neural representation both within and across voxels.

This page provides the resources necessary to understand the approach and to design your own experiments.

Methods

• The basic method: Continuous carry-over designs for fMRI, NeuroImage, Volume 35, Issue 4, May 2007, pages 1489-1494.
• Extension of the approach to study conjoint and independent neural tuning: DM Drucker, WT Kerr, GK Aguirre. (2009) Distinguishing conjoint and independent neural tuning for stimulus features with fMRI adaptation. Journal of Neurophysiology. June;101(6):3310-24 [ Link to supplementary materials ] [ Link to Faculty of 1000 evaluation ]
• Introduction to continuous carry-over designs for fMRI, a talk presented at Human Brain Mapping 2007 (and in a revised form in 2009)

Applications

• DM Drucker, GK Aguirre. (2009). Different spatial scales of shape similarity representation in lateral and ventral LOC. Cerebral Cortex, (TBA).
• I Charest, C Pernet, F Crabbe, P Belin. (submitted) Investigating the representation of voice gender using a continuous carry-over fMRI design. Human Brain Mapping 2009.

Carry-over experiments should present stimuli in a sequence which is at least first-order counterbalanced, meaning every stimulus precedes and follows every other. Second order counterbalancing is useful to guard against some modeling assumptions of the approach. Two types of sequences are useful for this purpose:

M-Sequences were introduced for use in fMRI by Buracas and Boynton. A limitation of m-sequences is that they are not available for all n of stimuli one may wish to study.

Type 1 Index 1 sequences, initially described by Finnery and Outhwaite in 1956 also provide first-order counterbalance, and are available for all n of stimuli greater than 5.

Given a particular similarity matrix to be detected in the carry-over effects, an assumed hemodynamic response function, and some scanning parameters (TR and stimulus ISI), one may search for an optimum type 1 index 1 sequence.

We have provided pre-made sequences for certain stimulus spaces.

We have provided for download the data, covariates, and results, from a continuous carry-over study.