Purchase this article with an account.
R. Landman, H. Spekreijse, V. A. F. Lamme; A neural correlate of change blindness in V1. Journal of Vision 2001;1(3):128. doi: https://doi.org/10.1167/1.3.128.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Sudden changes are often not noticed when coupled with a brief disruption (‘change blindness’). Change detection under these circumstances appears to require focal attention. Little is known about what happens in the brain when a change is detected/missed. To explore the relation between brain activity and change detection, we trained monkeys to detect which of a varied number of rectangles changed its orientation across two subsequent presentations. We recorded multi-unit activity in the primary visual cortex while the animals performed the task. The percentage of correct responses indicated that change detection accuracy depended on the number of rectangles. With 4 rectangles on screen, about 50% of the changes were missed. Here we compared neuronal activity before the change occurred, between correct and incorrect trials. In correct trials, cells with their receptive field inside the rectangle that was going to change had a firing rate, higher then in incorrect trials, starting 95 ms after stimulus onset. In the background, activity was lower in correct than incorrect trials, starting 235 ms after stimulus onset. At non-changing rectangles there was no difference. Thus, figure/background segregation was greater in correct trials than incorrect trials, particularly at the location of the change. These results show that the magnitude of a V1 response related to pre-attentive figure-ground segregation predicts whether a change is going to be detected or not. This suggests that early, perhaps pre-attentive, processing is very important in change detection.
This PDF is available to Subscribers Only