A new swept-parameter visual evoked potential (VEP) paradigm -- the Jitter Spatial Frequency (JSF) Sweep VEP -- is introduced that allows for efficient mapping of the spatiotemporal development of the developmental motion asymmetry (DMA). The paradigm involves presentation of vertical sinewave gratings undergoing 90o horizontal oscillatory displacements while their SF is swept during each VEP trial. JSF sweep VEPs, utilizing logarithmic spatial sweeps over 2 to 5 octaves, were recorded from 28 infants (8 - 43 weeks). Symmetric (second-harmonic, F2) and asymmetric (F1) components of the VEP were measured.
F1 responses typically had a bandpass profile as a function of spatial frequency, exhibiting both low- and high-spatial frequency (SF) rolloffs. RE and LE F1 responses have a 180-deg phase relationship, the signature of direction-labeled cortical detectors having opposite signs of nasalward/temporalward directional bias (Norcia et al., 1991; Hamer & Norcia , 1994). F2 responses tended to be more low-pass in form, exhibiting only a high-frequency rolloff.
Three spatial cutoffs can be derived from a single JSF sweep VEP by linear extrapolation to zero microvolts along the rising or falling limbs of the amplitude responses. Two high-SF cutoffs, for F1 and F2, are interpreted as estimates of the spatial resolution of direction selective (DS), motion mechanisms and non-DS, pattern mechanisms, respectively. A low-SF cutoff for F1 can also be estimated by extrapolation along the low-SF amplitude limb. We interpret this cutoff as an estimate of the boundary between low spatial frequencies, where the directionally selective mechanisms have matured and no longer have a biased response (F1 response absent), and higher spatial frequencies, for which the response is still immature (directionally biased, asymmetric, F1 present).
For 6 Hz oscillatory displacements, we find that the low-SF F1 cutoffs develop at 0.5 octaves/month, twice the rate of either conventional (phase-reversal) grating acuity development (Norcia & Tyler, 1985) or the high-SF F1 and F2 cutoffs measured in the present study. This result implies that no single mechanism can account for the developmental progression of the DMA at both low- and high-SFs. At 10 Hz, the DMA exhibited no significant development.
At both 6 Hz and 10 Hz, the F2 cutoffs occur at significantly higher spatial frequencies than the F1 cutoffs. This result suggests that infants' (DS) motion-sensitive mechanisms do not extend to as high spatial frequencies as (non-DS) pattern-sensitive mechanisms.
Finally, the high-SF cutoff for F1 provides an indirect estimate of the lower velocity limit of the direction-labeled cells underlying the DMA. This limit decreases with age from ~2 deg/sec at 8 weeks, to 0.75 deg/sec at 33 weeks, within the range of lower velocity limits for direction-of-motion identification in normal adults (~0.5 - 1 deg/sec; Graham, 1989), and close to prior VEP estimates in infants (0.6 deg/sec; Hamer & Norcia, 1994).