Contextual modulation of the high-resolution 3T GE BOLD fMRI response

    Research output: Contribution to journalArticlepeer-review


    Background and Aims. Inhibitory neural processes play a key role in human behavior: it is hard to find a perception or behavior that is not modulated by context, and the mechanisms of contextual modulation are frequently, if not dominantly, inhibitory. In human vision, contextual modulation is apparent in the suppression of visual features that are not distinguished from a scene background, those which lack contrast along low-level dimensions such as brightness, orientation or color. We have begun a series of experiments measuring the hemodynamic response to low-level contextual modulation in early visual areas. Methods. In this particular experiment, four target stimuli (Gabor patches) were presented at 3 degrees eccentricity (one in each quadrant of the visual field). Subjects were required to perform a contrast discrimination task on the target stimuli: the contrast of one of the four Gabors was augmented during one of two stimulus presentation intervals (a standard two-interval forced-choice task). In preparatory studies before the fMRI experiment, this psychophysical measurement allowed us to infer the early visual contrast response function to the target under three conditions: targets alone, targets with parallel flanking Gabor patches that suppressed the early visual response (measured as increased discrimination thresholds), and targets with orthogonal flanking Gabor patches (which did not suppress the early visual response). Responses to the targets were suppressed by 30% when targets presented at 35% contrast were surrounded by 50% contrast parallel flankers. In the corresponding fMRI experiment, the contrast discrimination task motivated a constant level of effort and alertness during all three experimental conditions and allowed experimental verification of the presence of surround suppression during the scanning. For the fMRI experiment, the three experimental conditions were presented in an event-related design with an average inter-trial-interval of 4.5 s. The posterior occipital cortex was imaged using a gradient echo (GE) EPI pulse sequence at 3 Tesla, with 1.5 mm isotropic resolution and a volume repetition time of 1.5 s. A general linear model was used to estimate the event-related response in each voxel to each stimulus type. Results. The amplitude of the BOLD response to the targets with parallel flankers was not decreased, even though both the psychophysical experiments and the electrophysiological literature indicate neural activity in this region of primary visual cortex is suppressed. Conclusions. Even considering the possible confounds of hemodynamic blurring, this finding questions the utility of standard GE BOLD techniques for studying contextual modulation on a fine scale. Although there was no observed correlation between BOLD peak amplitude and inferred neural activity in primary visual cortex, the different stimulus configurations elicited interesting differences in the shape of the hemodynamic response: the addition of suppressive flankers modulated the relative contributions of positive and negative portions of the response. Further experiments are investigating what the shape of the hemodynamic response function can tell us about contextual interactions in high resolution BOLD fMRI.

    Original languageEnglish (US)
    JournalJournal of Cerebral Blood Flow and Metabolism
    Issue numberSUPPL. 1
    StatePublished - Nov 13 2007

    Fingerprint Dive into the research topics of 'Contextual modulation of the high-resolution 3T GE BOLD fMRI response'. Together they form a unique fingerprint.

    Cite this