Abstract
Primate Middle Temporal neurons have been discovered with asymmetric antagonistic surrounds (Xiao et al., 1997). The response rate of such a neuron (e.g., tuned to V deg/s, 0°) is reduced by a small stimulus moving at the same velocity and located either directly above or below the classical receptive field of the neuron. Perrone (JOV, 2012) has argued that this asymmetric inhibition is designed to reduce redundant velocity signals generated along the edges of moving objects. Redundancy reduction and asymmetric inhibition have also been postulated to underlie the motion oblique effect in humans (Perrone, Liston & Stone, VSS, 2014). We therefore sought to find evidence for asymmetric inhibition in humans using a motion detection paradigm. Observers fixated a small central cross while a window (10° wide x 1.4° high) containing a high contrast moving edge was presented in the near periphery (V = 5 deg/s, 0° for 1 sec). Perrone (2012) suggested that MT pattern neurons are the target of the asymmetric inhibition. Therefore the test stimuli consisted of small Gaussian-windowed patches (SD = 0.25°) of moving 120° plaids (4 c/deg, 5 deg/s, 0°) designed to isolate pattern neurons. The targets were located above or below the moving edge and appeared for 200 msecs. They had to be discriminated from ‘static’ plaids (sum of two counter-phase gratings) as the contrast was reduced. Motion detection thresholds were measured using a 2AFC staircase for cases in which the edge was moving (M) and when it was static (S). The threshold ratio (M/S) was used to probe for evidence of inhibitory processes. We have isolated a number of locations where the threshold value below the edge was significantly higher than that above it and vice versa, supporting the idea of an asymmetric inhibitory mechanism designed to ‘thin’ the velocity signals along moving edges.
Meeting abstract presented at VSS 2015