Cortical Activation during Visual Memory-guided and Hand-guided Saccades: an MEG study

L Ren, EK Cressman, G Blohm, DO Cheyne, JD Crawford

Our previous studies (Ren et al. J. Neurophysiol. 2006, Ren and Crawford Exp. Brain Res. 2007) examined behavioral aspects of saccades to hand-held targets. Here, we use Magnetoencephalography (MEG) to compare the spatial-temporal patterns of cortical activation during visual memory-guided and hand-guided saccades. MEG measures the magnetic fields generated by the brain, from which the source of activity can be mathematically modeled. Electro-oculography (EOG) and Electromyography (EMG) were used to record eye movement and hand movement onset respectively. Eight subjects performed three paradigms in a dark room: 1) visual memory-guided saccades: subjects saccaded toward a briefly flashed target after the central fixation LED was extinguished; 2) hand-guided saccades: a handheld target was flashed and then actively moved (in the dark) to the opposite visual field while subjects fixated the central LED. The fixation LED was then extinguished, and subjects saccaded toward the handheld target; 3) combined task: same as #2, but the handheld target flashed briefly at its final location before the fixation LED was extinguished. With the fine time resolution of MEG, we found that in paradigm 1 (in which only visual information was available), there was a shift of cortical activation from cuneus to precuneus during the memory delay interval from the visual stimulus offset to saccade onset. In paradigm 2, primary sensory-motor cortex activation was observed 100ms after the onset of the hand movement, followed by frontal cortex activation (frontal eye fields according to its function and talairach coordinates) at saccade onset. In paradigm 3, the cortical activation observed included the areas seen in both paradigms 1 and 2. However, paradigm 3 showed stronger parietal activation and weaker cuneus activation than paradigm 1, but stronger cuneus activation and weaker frontal activation than paradigm 2. Subtraction of the ‘vision’ data or the ‘hand’ data from the ‘combined’ data left similar, but not identical patterns of activation compared to the other hand/vision data set, suggesting that in this task sensory integration is more than the sum of its parts. In conclusion, the current MEG data trace the spatialtemporal sequence of brain activation during visual memory-guided and hand-guided saccades, and illustrate modality- specific pathways for human saccades.