D Birtles

Brain responses to global perceptual coherence

Coherence thresholds for concentric patterns of contour segments or dot trajectories have revealed differential development and vulnerability of extrastriate form and motion pathways respectively (Braddick et al, Neuropsychologia, 2003). Here we explore the properties of visual evoked potentials (VEPs) specific to form and motion coherence. Coherent contour or motion arrays alternate with incoherent arrays at 0.5-4 Hz. A 2nd harmonic VEP signal (F2) may arise from local changes occuring at each transition, but a first harmonic (F1) must reflect differential responses to coherence and incoherence, indicating global processing. F1 amplitude is approximately linear with coherence for both form and motion, confirming that F1 measures global processing. F2 is independent of coherence for form, and nearly so for motion, implying that it reflects local changes. Occipital F1 amplitudes for motion: form coherence are approximately 2:1. However, this does not reflect underlying sensitivity: psychophysical form and motion coherence thresholds are similar, and extrapolation of the linear region of F1 amplitude vs coherence reaches zero close to threshold coherence in each case. F1 amplitude for both displays increases gradually from 0.5-4 Hz. However, this is opposite to psychophysical coherence sensitivity, which falls over the same range. We conclude: (a) coherence-dependent VEPs provide measures of global form and motion processing, practical for infants and non-verbal (e.g. autistic) children; (b) this response is elicited efficiently at relatively high presentation rates; (c) the level tapped by the VEP does not provide the limiting factor on temporal integration in global perceptual processing.