Object-centered attentional biases and object recognition contributions toscene segmentation in left- and right-hemisphere-damaged patients

Citation
Ma. Peterson et al., Object-centered attentional biases and object recognition contributions toscene segmentation in left- and right-hemisphere-damaged patients, PSYCHOBIOLO, 26(4), 1998, pp. 357-370
Citations number
52
Categorie Soggetti
Neurosciences & Behavoir
Journal title
PSYCHOBIOLOGY
ISSN journal
08896313 → ACNP
Volume
26
Issue
4
Year of publication
1998
Pages
357 - 370
Database
ISI
SICI code
0889-6313(199812)26:4<357:OABAOR>2.0.ZU;2-3
Abstract
Participants viewed elongated rectangular displays in which two regions sha red a central contour. In experimental stimuli, the central contour portray ed a known object on one, high-denotative, side. In control stimuli, no kno wn objects were portrayed on either side of the central contour but one sid e of each control stimulus was a scrambled version of one of the high-denot ative regions, matching it on all factors known to influence scene segmenta tion other than object recognition. For each display, participants decided whether the left or the right region was more likely to be an object. Parad oxically, both right-hemisphere- (RH) and left-hemisphere- (LH) damaged ind ividuals were more likely to see objects lying on the contralesional rather than the ipsilesional side of the central contour. This tendency is attrib uted to an object-centered attentional bias toward the central contour when objects lie on its contralesional side and away from the central contour w hen objects Lie on its ipsilesional side. Object-centered attentional biase s were stronger following RH than LH damage. Elderly control participants s howed a slight bias in the same direction as RH-damaged individuals. More h igh-denotative regions than scrambled regions were seen as objects, even wh en object-centered attention was biased away from the central contour carry ing the object recognition information. The latter result suggests that the object recognition processes contributing to scene segmentation are preatt entive.