To what extent occluded objects produce responses in early visual cortex remains less well understood. It thus seems plausible that it extracts shape and surface properties and is only indirectly involved in detecting illusory contours. In contrast, LOC neurons respond to a wide area of the visual scene, frequently spreading over both hemifields 19, 23 and LOC is responsive to objects. Activity in these early regions might then be interpreted as sharp illusory edges. This makes them ideally suited for assigning sharp object boundaries. This also concurs with the response properties of neurons in early visual cortex: V1/V2 neurons have small receptive fields 19, 20 that are selective for edge orientation 21, 22. This suggests that feedback signals from LOC into early visual areas are required for the phenomenal experience of illusory contours to arise. While disrupting neural processing in LOC eliminates the illusory contour percept shortly after stimulus onset, TMS over early visual cortex does so at a later time. Therefore, LOC activation may reflect a quick, crude surface construction or object recognition response rather than the phenomenal experience of illusory contours.įurther evidence for a multistage, recurrent process in the percept of illusory contours comes from transcranial magnetic stimulation (TMS) experiments 18. Comparable fMRI activation occurs in LOC for both stimulus conditions. Stimuli in which the sharp edges of the Kanizsa-type Pacmen are curved eliminate the perception of illusory contours but nonetheless induce an impression of an enclosed salient diamond-shaped region 7. While higher visual areas are preferentially activated by illusory contour stimuli, the LOC activation itself might reflect global completion processes that are independent of the illusory contour percept. These results therefore advocate a “top-down” mechanism, in line with computational models emphasizing higher level surface construction and corresponding feedback 17. The LOC is situated higher in the visual processing hierarchy than V2. The LOC is also preferentially activated in fMRI studies on occlusion 16. An fMRI study found that illusory contours induce the strongest activation in a region encompassing high extrastriate cortex 15. The lateral occipital complex (LOC) is located within extrastriate cortex and activated when participants view intact objects compared to scrambled versions 12, 13, 14. Whilst these studies provide strong evidence for early visual cortex activation in response to illusory contours, they do not clarify whether this activation is due to feedback from higher areas. These results converge to suggest that information about the contours themselves is present as early as V1. A more recent fMRI study demonstrated that the orientation of illusory contours is encoded across visual cortex, including V1 1. Optical imaging experiments suggested an inversion of orientation tuning for illusory contours in V1 but not V2: while tuning preferences for actual and illusory contours were correlated in V2, orientation tuning for illusory contours was orthogonal to that for actual contours in V1 3. In humans, regional cerebral blood flow within V2 is increased whilst viewing illusory contour stimuli 5. Experiments in cats and macaque monkeys indicate that neurons in early visual regions support illusory contours, and that V2 activation is more integral than V1 8, 11. Therefore, it remains unclear whether a separate or common neural mechanism mediates these different perceptual completion processes within early visual cortex.Įarly visual areas may complete occluded portions of an object topographically and send this information to higher cortical regions for object recognition. However, there is also conflicting evidence that V1 and V2 respond to both illusory and occlusion stimuli 11. Moreover, electrophysiological recordings in macaque monkeys showed that neuronal activity in early visual areas V1 and V2 was elevated by illusory contours but diminished by occlusion 10. The left and right hemispheres of the patients were equally able to complete illusory contours however, the right hemisphere preferentially supported occlusion. In that study, illusory contour or occlusion stimuli were placed in each hemifield. Further evidence for different neural mechanisms stems from the analysis of two callosotomy (“split brain”) patients 9. The perceptual differences between illusory contours and occlusion suggest that their neural activation should also be dissociable in some brain regions.

It remains controversial whether illusory contours and occlusion share a common neural mechanism and whether they reflect a bottom-up or top-down process. Several studies reported that both illusory contours and occlusion produce neural activity in early visual cortex 1, 2, 3, 4, 5, 6, 7, 8.

Occipital contour definition