Coherence dependence of high-density visual evoked potentials to global form and motion displays

Abstract

Coherence thresholds for global form and motion can be used to assess function of the extra-striate ventral and dorsal pathways respectively. Psychophysical thresholds are similar in adults, but show differential development, and motion thresholds are raised in a variety of developmental disorders.

Thresholds can also be estimated by measuring the amplitude of visual evoked potentials (VEPs) as a function of coherence, and extrapolating to zero amplitude. With a single occipital channel, VEP form and motion thresholds are similar, but motion amplitudes are about half those for form coherence (Braddick et al, VSS 2006). Here, we extend this approach to high-density (128 channel) VEPs to address three questions: (a) are the single-channel amplitude differences a result of non-optimal electrode location for form and/or motion? (b) are form and motion VEP thresholds the same when each is measured at its optimal location? (c) do any scalp locations show a non-linear response to coherence, as found in some brain regions with fMRI?

The form stimulus had dots arranged in short concentric arcs giving a static global circular pattern. The motion stimulus dots moved along similar arcs, producing global rotation. Coherence of each pattern type varied from 0% – 100% in separate recording blocks. These patterns alternated at 2 Hz with randomly-arranged dot trajectories with no global structure.

Both form and motion elicited significant posteriorly-located VEPs, showing significant linear regression with coherence. Amplitudes of the peak responses to form and motion were similar, but the locations differed; the motion peak was close to the midline, the form peak more lateral. Form and motion VEP thresholds derived from these peaks were similar to each other, but significantly greater than the psychophysical thresholds. Finally, we found no evidence of systematic 2nd- or 3rd-order non-linearities at any scalp location.

Footnotes

 Supported by Research Grant G0601007 from the Medical Research Council.