초록 ▼

• 과밀효과(crowding effect)는 시각이 역동적으로 조절되는 맥락(context) 효과의 대표적인 현상임. 본 과제는 사람과 실험용 원숭이에서 single-cell 활동, LFP, fMRI, ERP, 등의 신경 활동을 측정하고 이 신호들의 관계를 분석하여 행동 상의 맥락효과가 발생하는 신경 기전을 규명하고자 한 것임.

• 행동과제를 수행하도록 훈련된 원숭이 V1 세포들의 수용장(receptive field) 내외에 위치한 중심-주변 자극의 신경표상 간의 상호작용을 spike 및 LFP 신호를 측정하여, 맥락효

• 과밀효과(crowding effect)는 시각이 역동적으로 조절되는 맥락(context) 효과의 대표적인 현상임. 본 과제는 사람과 실험용 원숭이에서 single-cell 활동, LFP, fMRI, ERP, 등의 신경 활동을 측정하고 이 신호들의 관계를 분석하여 행동 상의 맥락효과가 발생하는 신경 기전을 규명하고자 한 것임.

• 행동과제를 수행하도록 훈련된 원숭이 V1 세포들의 수용장(receptive field) 내외에 위치한 중심-주변 자극의 신경표상 간의 상호작용을 spike 및 LFP 신호를 측정하여, 맥락효과를 일으키는 시각 표상 간의 상호 작용이 발생하는 패턴과 거리를 분석. 수용장 외부 시각 자극에 대한 LFP 반응은 10mm 이상까지 관찰되어 신경 표상 간 상호작용의 범위가 광범함을 발견.

• 사람의 경우, 신경 표상 간 상호 작용은 수용장의 방위선호성에 따라서 불균등하게 발생하고 있음을 fMRI 신호를 통해서 발견하였으며, 시각 표상들이 대뇌 피질 상에서의 거리에 따라 체계적으로 변화함을 발견.

• 과밀효과를 양안 경합과 동시에 제시함으로써, 두 효과 간 상호작용을 측정한 결과, 과밀효과가 발생하는 단계가 방위 변별의 경우 V1에서 V4 사이인 것으로 추정됨.

(출처:보고서 초록 p.3)

Abstract ▼

● During this project period, we aimed to examine the contextual effects in visual processing at the diverse levels of neural and behavioral measures. These included spike and local field potential (LFP) signals in the primate V1, fMRI and ERP in humans, collected with experimental paradigms of visu

● During this project period, we aimed to examine the contextual effects in visual processing at the diverse levels of neural and behavioral measures. These included spike and local field potential (LFP) signals in the primate V1, fMRI and ERP in humans, collected with experimental paradigms of visual crowding.

● A focal stimulus outside the receptive field robustly induces LFP change, while the same stimulus evokes no spike activity. We determined how this subthreshold LFP induced by surround stimulus interacted with spike response to the RF stimulus in the awake monkey V1. We sequentially presented two identical Gabor stimuli with a variable stimulus onset asynchrony (SOA) the first one (S1, context stimulus) was presented outside RF inducing a subthreshold LFP change, and the second one (S2, test stimulus) was subsequently presented within RF generating a spiking response. This enabled us to manipulate the temporal relation between subthreshold LFP and evoked spike activity and to determine whether subthreshold LFP contributed to modulation of spike activity in a SOA-dependent manner. We found that the subthreshold LFP propagated a considerably long distance, estimated to be more than 10mm of cortical distance. The cross-correlation between the time course of subthreshold LFP and the pattern of SOA-dependency of spike activity was significant. These results indicate that signal integration is farther beyond the RF than previously estimated based on spike-triggered average, and suggest that subthreshold LFP, as a context, modulate spike activity in a SOA-dependent manner.

● We originally planned to conduct a series of fMRI experiments to understand cortical mechanisms of visual crowding effects. Because we found a critical problem in the original fMRI paradigm, however, we conducted two alternative sets of fMRI experiments to study contextual effects: (1) analysis of coaxially anisotropic spread of cortical responses to oriented stimuli in human early visual cortex; (2) analysis of functional architecture of noise and resting-state correlations in human early visual cortex. We developed a new task and stimuli that allow us to readily reveal the key phenomenon of visual crowding effects, which is mis-binding of local features between target ang crowding stimuli. We established the solid presence of co-axial anisotropy in point spread function of fMRI responses to unifocal stimuli at the fovea and multifocal stimuli at the periphery in human early visual cortex. We also demonstrated that the co-axial anisotropy was systematically dependent on the co-orientation relationship between stimuli and voxels' orientation preferences and eccentricity locations. Finally, we obtained a preliminary set of fMRI data indicating that local regions in human early visual cortex are systematically correlated in their noise responses as well as resting-state responses.

● We examined the stages of crowding effect in visual processing by inducing binocular rivalry to the flankers. The changes of neuronal responses were measured with ERP while the distance between the target and the flankers was quantitatively manipulated. We also performed experiments on ensemble statistics to find out the mechanisms underlying the center-surround interaction and crowding effect. We measured the bias and the precision of participants' orientation discrimination when both crowding and binocular rivalry occurred. We found that judgments of target orientations were more impaired when the impairment was due to the simultaneous induction of crowding and rivalry compared to when the impairments of only crowding and only rivalry were combined. These results suggest that brain areas with orientation selectivity from V1 to V4 are related to crowding effect. We also found that mean size can be calculated with the ratio of the entire area of presented items to their numerosity. Finally, we showed that statistical information can be extracted only from visible stimuli.

(출처:SUMMARY p.6)

목차 Contents

• 표지 ... 1제 출 문 ... 2보고서 초록 ... 3요 약 문 ... 4SUMMARY ... 6CONTENTS ... 8목차 ... 9제 1 장 연구개발과제의 개요 ... 10 제 1절 연구개발의 목표 ... 10 제 2절 연구개발의 중요성 ... 11제 2 장 국내외 기술개발 현황 ... 12 제 1절 국내외 기술개발 현황 ... 12 제 2절 구체적 연구사례 조사 ... 13제 3 장 연구개발수행 내용 및 결과 ... 16 제 1절 연구범위 및 이론적 실험적 접근방법 ... 16 제 2절 연구수행 내용 및 결과 ... 18제 4 장 목표달성도 및 관련분야에의 기여도 ... 40 제 1절 연구개발목표의 달성도 ... 40 제 2절 평가의 착안점에 따른 목표달성도에 대한 자체평가 ... 41제 5 장 연구개발결과의 활용계획 ... 42제 6 장 연구개발과정에서 수집한 해외과학기술정보 ... 43제 7 장 참고문헌 ... 44