Temporal integration of movement: the time-course of motion streaks revealed by masking

Abstract

Temporal integration in the visual system causes fast-moving objects to leave oriented 'motion streaks' in their wake, which could be used to facilitate motion direction perception. Temporal integration is thought to occur over ≈100 ms in early cortex, although this has never been tested for motion streaks. Here we compare the ability of fast-moving ('streaky') and slow-moving fields of dots to mask briefly flashed gratings either parallel or orthogonal to the motion trajectory. Gratings were presented at various asynchronies relative to motion onset (from -200 to +700 ms) to sample the time-course of the accumulating streaks. Predictions were that masking would be strongest for the fast parallel condition, and would be weak at early asynchronies and strengthen over time as integration rendered the translating dots more streaky and grating-like. The asynchrony where the masking function reached a plateau would correspond to the temporal integration period. As expected, fast-moving dots caused greater masking of parallel gratings than orthogonal gratings, and slow motion produced only modest masking of either grating orientation. Masking strength in the fast, parallel condition increased with time and reached a plateau after 77 ms, providing an estimate of the temporal integration period for mechanisms encoding motion streaks. Interestingly, the greater masking by fast motion of parallel compared with orthogonal gratings first reached significance at 48 ms before motion onset, indicating an effect of backward masking by motion streaks.

Figure 1. A schematic diagram of the temporal masking experiment.

Probe orientation was defined relative to the direction of motion, either parallel with the motion trajectory or orthogonal to it. Four motion directions were randomly interleaved to help prevent adaptation. Participants' task was to judge whether the grating appeared in one of two motion displays, upper or lower. The upper and lower windows always contained the same motion direction on a given trial, and various test asynchronies from the full set of 21, ranging from to ms relative to motion onset, were randomly interleaved in blocks of trials.

Figure 2. Masking elevations of a brief target grating from unmasked baseline as a function of presentation time relative to the onset (0 ms) of a 500 ms motion mask.

The plot shows the means of four observers, plotted with 1 standard error bars. To capture any effects of backward and forward masking, target gratings were presented as early as 190 ms before the motion mask began (indicated by negative timing), as well as up to 190 ms after the motion mask ended. Results are shown for fast and slow motion masks, and for target gratings parallel and orthogonal to the motion direction.

Figure 3. Results from a repeated-measures one-way ANOVA analysis of the averaged threshold elevations for each condition during the masking period.

Error bars show 1 standard error.

Figure 4. Masking specific to motion streaks – differences and first derivative.

a) The masking component specific to motion streaks plotted as a function of the grating target's asynchrony relative to motion onset. The plot shows group means (n = 7), flanked by 90% confidence intervals, of the difference between the fast parallel and fast orthogonal conditions. This contrast reveals the streak-specific masking component because while both conditions contain fast translating dots (and therefore motion streaks), masking occurs only in the parallel condition where target and mask are iso-oriented. b) The first derivative of the streak-specific masking component, calculated using the three-point method, plotted in panel A, flanked by 90% confidence intervals. The sustained increase in masking in panel A around 0 ms is indicated by the four consecutive positive slopes around 0 ms. The points either side of this series of four points are not significantly different from zero. Linearly interpolating between the lower confidence intervals, this elevated series of points is significant between and ms, indicating a temporal integration period of 77 ms.

Figure 5

a) A sustained impulse response function defined by Manahilov et al's (2003) equation (see equation 1, main text). The function plotted here has the following parameters: – the values Manahilov et al found best described sustained impulse responses at a frequency of 2 cyc/deg. At half-height, the impulse response has a full width of 41 ms. b) A single frame taken from the sequence of frames defining the fast translating blobs. c) The temporally smeared version of the blob stimulus that results from passing the fast translating blob image in panel b through the temporal impulse shown in panel a (i.e., performing a convolution integral). The output shown in panel c is the ‘streaky’ image that can be assumed to emerge following a simple linear filtering stage characterized by a sustained impulse response. d) The spatial tuning of the streak image in panel c. The figure shows the output of a sliding log Gabor filter computing the spatial energy at each spatial frequency from the minimum frequency to 12 cyc/deg in the direction orthogonal to the streaky elongations (i.e., vertically, in this case). The log Gabor had a spatial bandwidth of 1 octave and a narrow orientation bandwidth (1) oriented to sample vertically across the image shown in panel c. The peak frequency occurs at 1.6 cyc/deg and falls to half-height at 3.3 cyc/deg. e) The orientation tuning of the image in panel c at peak frequency. The data were obtained by rotating the log Gabor filter (1 octave spatial bandwidth, peak at 1.6 cyc/deg) in one-degree steps. Grey symbols show the filter output and the black line is the best-fitting Gaussian function (standard deviation = 22.2).