Abstract
Humans have several mechanisms for the visual perception of motion, including one that is luminance-based (first-order) and another that is luminance-independent (second-order). Recent psychophysical studies have suggested that significant interaction occurs between these two neural processes. We investigated whether such interactions are represented as neural activity measured by magnetoencephalography (MEG). The second-order motion of a drifting sinusoidal grating, which is defined by the speed of the dot motion, did not generate a response. Apparent motion (AM) of the square area, defined by the speed of randomly moving dots, evoked a magnetic response whose latency and amplitude changed with the distance that the area moved (a second-order characteristic), though the response properties were significantly different from those for the first-order AM. AM, defined by both first- and second-order attributes, evoked an MEG response and the latencies and the amplitudes were distributed between those for the first- and second-order motions. The cortical source of the response was estimated to be around MT+. The results show a distinct difference in the neural processing of the second-order motion that cannot be explained by the difference in visibility, and they indicate that the interaction of the neural processes underlying first- and second-order motion detection occurs before the MEG response. Our study provides the first physiological evidence of a neural interaction between the two types of early motion detection.