Comparison of MEG and BOLD Signals in a Visual-Tactile Discrimination Task

Abstract

Event Abstract Back to Event Comparison of MEG and BOLD Signals in a Visual-Tactile Discrimination Task Yiwen L. Hegner1*, Ying Lee1, Erick Ortiz2 and Christoph Braun3 1 University of Tuebingen, MEG Centre, Institute of Medical Psychology and Behavioural Neurobiology, Germany 2 University of Tuebingen, MEG and fMEG Center, Germany 3 University of Trento, Center for Mind/Brain Sciences, Italy Previous studies have shown a complex relationship between scalp electromagnetic activation and BOLD signals[1,2]. Accordingly, not only evoked fields but also event-related de-/synchronization are reflected in positive BOLD responses. In order to further explore the relationship between MEG and fMRI signals, a combined event-related fMRI and 275 channel (CTF) whole-head MEG study was conducted in 12 right-handed subjects. The experimental setup for both studies was identical except that longer inter-trial intervals were used in the fMRI study to account for the slow BOLD signal. Subjects were instructed to match one out of two visual patterns with a tactile pattern. The visual patterns were presented 1 s prior to the presentation of the tactile pattern, which lasted 0.6s and were presented simultaneously at both middle fingertips. One second after the offset of the tactile stimulus, a rectangular response cue encircled one of the two visual patterns for 1s and indicated which of the them had to be compared with the tactile stimulus. Half of the subjects responded to matching visual-tactile stimuli, the other half to non-match. There were two sessions and each session had 96 (48 easy, 48 difficult) trials. MEG source localization was performed using the multiple sparse prior method in SPM8. Both the source localization of the transient evoked tactile response (e.g. 40-60ms and 60-80ms post tactile stimulus onset) and of the induced oscillatory activity in the beta (16-22Hz) and gamma range (56-66Hz) during 20-80ms post tactile stimulus onset correspond largely with the location of the BOLD responses. The contrast between easy and difficult task conditions revealed that the sources of the beta activity match best with the same contrast of the BOLD activation. Our study thus demonstrates that in the current visual-tactile pattern matching task, the BOLD signals are mostly represented by the beta oscillatory activity shortly after the tactile stimulus onset.