fMRI responses to perceptually completed regions in the human lateral occipital complex: Bounding illusory contours are not a necessary condition

Damian A. Stanley, Nava Rubin

Research output: Contribution to journalArticle

Abstract

It was previously reported that regions in the human Lateral Occipital Complex (LOC) show BOLD fMRI responses to illusory Kanizsa diamonds (Mendola et al. 1999), but it was not known whether the bounding illusory contours (ICs) were necessary to elicit the response. To test this, we generated new stimuli by rounding the corners of the Kanizsa 'pacmen' inducers and misaligning them slightly. Although ICs were no longer perceived, the configuration of the inducers nevertheless created an impression of an enclosed, Salient Region (SR). Baseline stimuli (no SR) were generated by flipping the inducers outwards. Observers viewed alternating blocks of SR and baseline stimuli (16 sec blocks, 256 sec runs). In separate runs, observers viewed Kanizsa (IC) diamonds alternating with facing-out inducers. IC/SR size was either 6.5 or 13 deg. EPI images were taken on a Siemens Allegra head-only 3T scanner (16 near-coronal slices, 3×3×3 mm voxels, TR=2s, TE=30ms). Regions of interest (ROIs) were pre-defined as the voxels in LO cortex (inferior and middle occipital gyri, lateral and inferior occipital sulci, and the lunate sulcus when present) which produced a greater BOLD response to pictures of objects than scrambled objects. Eight observers were tested (5 males; 6 naive). The across-observer average time-course in the ROIs was computed. The results showed a significant elevation of activity for both the IC and SR stimuli, compared to their baseline stimuli. There was no significant difference between the two conditions: the amplitude of fMRI response was just as strong for the SR stimuli. We conclude that the previously reported LOC responses to Kanizsa-type stimuli did not require the presence of bounding illusory contours. LOC response to Salient Regions may be the result of fast but crude region-based segmentation processes. Further research is needed to test if the more refined contour-based processes involve early cortex, as suggested by electrophysiological studies.

Original languageEnglish (US)
JournalJournal of Vision
Volume2
Issue number7
DOIs
StatePublished - 2002

Fingerprint

Occipital Lobe
fexofenadine
Diamond
Magnetic Resonance Imaging
Head
Research

ASJC Scopus subject areas

  • Ophthalmology

Cite this

fMRI responses to perceptually completed regions in the human lateral occipital complex : Bounding illusory contours are not a necessary condition. / Stanley, Damian A.; Rubin, Nava.

In: Journal of Vision, Vol. 2, No. 7, 2002.

Research output: Contribution to journalArticle

@article{4158ec3e24ea4a5c8d77f5ee8c3d8742,
title = "fMRI responses to perceptually completed regions in the human lateral occipital complex: Bounding illusory contours are not a necessary condition",
abstract = "It was previously reported that regions in the human Lateral Occipital Complex (LOC) show BOLD fMRI responses to illusory Kanizsa diamonds (Mendola et al. 1999), but it was not known whether the bounding illusory contours (ICs) were necessary to elicit the response. To test this, we generated new stimuli by rounding the corners of the Kanizsa 'pacmen' inducers and misaligning them slightly. Although ICs were no longer perceived, the configuration of the inducers nevertheless created an impression of an enclosed, Salient Region (SR). Baseline stimuli (no SR) were generated by flipping the inducers outwards. Observers viewed alternating blocks of SR and baseline stimuli (16 sec blocks, 256 sec runs). In separate runs, observers viewed Kanizsa (IC) diamonds alternating with facing-out inducers. IC/SR size was either 6.5 or 13 deg. EPI images were taken on a Siemens Allegra head-only 3T scanner (16 near-coronal slices, 3×3×3 mm voxels, TR=2s, TE=30ms). Regions of interest (ROIs) were pre-defined as the voxels in LO cortex (inferior and middle occipital gyri, lateral and inferior occipital sulci, and the lunate sulcus when present) which produced a greater BOLD response to pictures of objects than scrambled objects. Eight observers were tested (5 males; 6 naive). The across-observer average time-course in the ROIs was computed. The results showed a significant elevation of activity for both the IC and SR stimuli, compared to their baseline stimuli. There was no significant difference between the two conditions: the amplitude of fMRI response was just as strong for the SR stimuli. We conclude that the previously reported LOC responses to Kanizsa-type stimuli did not require the presence of bounding illusory contours. LOC response to Salient Regions may be the result of fast but crude region-based segmentation processes. Further research is needed to test if the more refined contour-based processes involve early cortex, as suggested by electrophysiological studies.",
author = "Stanley, {Damian A.} and Nava Rubin",
year = "2002",
doi = "10.1167/2.7.352",
language = "English (US)",
volume = "2",
journal = "Journal of Vision",
issn = "1534-7362",
publisher = "Association for Research in Vision and Ophthalmology Inc.",
number = "7",

}

TY - JOUR

T1 - fMRI responses to perceptually completed regions in the human lateral occipital complex

T2 - Bounding illusory contours are not a necessary condition

AU - Stanley, Damian A.

AU - Rubin, Nava

PY - 2002

Y1 - 2002

N2 - It was previously reported that regions in the human Lateral Occipital Complex (LOC) show BOLD fMRI responses to illusory Kanizsa diamonds (Mendola et al. 1999), but it was not known whether the bounding illusory contours (ICs) were necessary to elicit the response. To test this, we generated new stimuli by rounding the corners of the Kanizsa 'pacmen' inducers and misaligning them slightly. Although ICs were no longer perceived, the configuration of the inducers nevertheless created an impression of an enclosed, Salient Region (SR). Baseline stimuli (no SR) were generated by flipping the inducers outwards. Observers viewed alternating blocks of SR and baseline stimuli (16 sec blocks, 256 sec runs). In separate runs, observers viewed Kanizsa (IC) diamonds alternating with facing-out inducers. IC/SR size was either 6.5 or 13 deg. EPI images were taken on a Siemens Allegra head-only 3T scanner (16 near-coronal slices, 3×3×3 mm voxels, TR=2s, TE=30ms). Regions of interest (ROIs) were pre-defined as the voxels in LO cortex (inferior and middle occipital gyri, lateral and inferior occipital sulci, and the lunate sulcus when present) which produced a greater BOLD response to pictures of objects than scrambled objects. Eight observers were tested (5 males; 6 naive). The across-observer average time-course in the ROIs was computed. The results showed a significant elevation of activity for both the IC and SR stimuli, compared to their baseline stimuli. There was no significant difference between the two conditions: the amplitude of fMRI response was just as strong for the SR stimuli. We conclude that the previously reported LOC responses to Kanizsa-type stimuli did not require the presence of bounding illusory contours. LOC response to Salient Regions may be the result of fast but crude region-based segmentation processes. Further research is needed to test if the more refined contour-based processes involve early cortex, as suggested by electrophysiological studies.

AB - It was previously reported that regions in the human Lateral Occipital Complex (LOC) show BOLD fMRI responses to illusory Kanizsa diamonds (Mendola et al. 1999), but it was not known whether the bounding illusory contours (ICs) were necessary to elicit the response. To test this, we generated new stimuli by rounding the corners of the Kanizsa 'pacmen' inducers and misaligning them slightly. Although ICs were no longer perceived, the configuration of the inducers nevertheless created an impression of an enclosed, Salient Region (SR). Baseline stimuli (no SR) were generated by flipping the inducers outwards. Observers viewed alternating blocks of SR and baseline stimuli (16 sec blocks, 256 sec runs). In separate runs, observers viewed Kanizsa (IC) diamonds alternating with facing-out inducers. IC/SR size was either 6.5 or 13 deg. EPI images were taken on a Siemens Allegra head-only 3T scanner (16 near-coronal slices, 3×3×3 mm voxels, TR=2s, TE=30ms). Regions of interest (ROIs) were pre-defined as the voxels in LO cortex (inferior and middle occipital gyri, lateral and inferior occipital sulci, and the lunate sulcus when present) which produced a greater BOLD response to pictures of objects than scrambled objects. Eight observers were tested (5 males; 6 naive). The across-observer average time-course in the ROIs was computed. The results showed a significant elevation of activity for both the IC and SR stimuli, compared to their baseline stimuli. There was no significant difference between the two conditions: the amplitude of fMRI response was just as strong for the SR stimuli. We conclude that the previously reported LOC responses to Kanizsa-type stimuli did not require the presence of bounding illusory contours. LOC response to Salient Regions may be the result of fast but crude region-based segmentation processes. Further research is needed to test if the more refined contour-based processes involve early cortex, as suggested by electrophysiological studies.

UR - http://www.scopus.com/inward/record.url?scp=4243192361&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4243192361&partnerID=8YFLogxK

U2 - 10.1167/2.7.352

DO - 10.1167/2.7.352

M3 - Article

VL - 2

JO - Journal of Vision

JF - Journal of Vision

SN - 1534-7362

IS - 7

ER -