A 7T fMRI correlate of visual perceptual filling-in in human visual cortex
Poster No:
2557
Submission Type:
Abstract Submission
Authors:
Kenshu Koiso1,2, Anna Razafindrahaba1, Mark Roberts1, Vincent van de Ven1, Federico De Martino1, Peter De Weerd1
Institutions:
1Maastricht University, Maastricht, the Netherlands, 2The University of Electro-Communications, Tokyo, Japan
First Author:
Kenshu Koiso
Maastricht University|The University of Electro-Communications
Maastricht, the Netherlands|Tokyo, Japan
Maastricht University|The University of Electro-Communications
Maastricht, the Netherlands|Tokyo, Japan
Co-Author(s):
Introduction:
We have used Troxler fading to probe fMRI correlates of visual surface perception in retinotopic visual cortex. In our version of this paradigm, subjects fixate a point away from a gray square surrounded by dynamic texture (1). After prolonged, steady fixation, the square is perceptually filled-in by the texture (perceptual filling-in) in the absence of physical changes to the stimulus. A neurophysiological study in monkeys (2) reported a signal increase in the representation of the square during perceptual filling-in. Here, we test whether a 7T fMRI correlate of perceptual filling-in is present in human visual cortex, and whether attention modulates its strength. We will also test the laminar distribution of the signal to evaluate the possibility that feedback contributes to the signal.
Methods:
Stimulus:
A dynamic texture was used consisting of small vertical bars refreshing at 30Hz. A 2.75 deg square texture-free, equiluminant gray square was placed over the texture, 5.5 deg away from the fixation spot. The luminance was set at 35 cd/m2 throughout the display.
Six or seven functional runs, four localizer runs, and anatomical image acquisition were conducted in a 2h session per participant. In functional runs, a block design with 30.36 sec stimulus-on and 20.24 sec stimulus-off was used. In the standard condition, Filling-In was Enabled (FIE), whereas in a control condition Filling-In was Prevented (FIP) by replacing the entire stimulus with an equiluminant full-screen gray for 0.2s every second (see 2). In both FIE and FIP conditions, attention was either directed to the square (by asking participants to press a button whenever they observed filling-in; ATT condition), or was directed away (using a difficult brightness detection task at fixation; NATT condition). Two blocks per condition were included in a run (~7 min).
In the localizer run, we retinotopically mapped the location of the patch using a retionotopic phase-encoding design. A 16 Hz flickering checkerboard frame of 0.17 deg moved from the center of the patch to 1.5 deg outside with a total of 17 phase encoded steps at a rate of 1 TR/step. This mapping stimulus was repeated eight times per run (~6 min).
Scanning:
We used a Siemens 7T MAGNETOM scanner with a 1TX/32RX head coil at Scannexus (Maastricht, NLs). For functional and localizer runs, we collected GE-BOLD data with a resolution of 0.8 mm isotropic, 66 slices, and a TR of 2.53 sec. For anatomical imaging, a MP2RAGE protocol with a resolution of 0.7 mm isotropic was used.
Participants:
Two young healthy females with normal or corrected eye acuity were scanned.
Analysis:
BrainVoyager 22.4 (Brain Innovation, Maastricht, NLs) was used for fMRI data analysis (3). Standard preprocessing was applied. We identified the patch cortical location using standard retinotopic mapping. In preliminary analyses, the response to the different experimental conditions within the cortical location of the patch was determined by averaging the BOLD response across repetitions and runs.
A dynamic texture was used consisting of small vertical bars refreshing at 30Hz. A 2.75 deg square texture-free, equiluminant gray square was placed over the texture, 5.5 deg away from the fixation spot. The luminance was set at 35 cd/m2 throughout the display.
Six or seven functional runs, four localizer runs, and anatomical image acquisition were conducted in a 2h session per participant. In functional runs, a block design with 30.36 sec stimulus-on and 20.24 sec stimulus-off was used. In the standard condition, Filling-In was Enabled (FIE), whereas in a control condition Filling-In was Prevented (FIP) by replacing the entire stimulus with an equiluminant full-screen gray for 0.2s every second (see 2). In both FIE and FIP conditions, attention was either directed to the square (by asking participants to press a button whenever they observed filling-in; ATT condition), or was directed away (using a difficult brightness detection task at fixation; NATT condition). Two blocks per condition were included in a run (~7 min).
In the localizer run, we retinotopically mapped the location of the patch using a retionotopic phase-encoding design. A 16 Hz flickering checkerboard frame of 0.17 deg moved from the center of the patch to 1.5 deg outside with a total of 17 phase encoded steps at a rate of 1 TR/step. This mapping stimulus was repeated eight times per run (~6 min).
Scanning:
We used a Siemens 7T MAGNETOM scanner with a 1TX/32RX head coil at Scannexus (Maastricht, NLs). For functional and localizer runs, we collected GE-BOLD data with a resolution of 0.8 mm isotropic, 66 slices, and a TR of 2.53 sec. For anatomical imaging, a MP2RAGE protocol with a resolution of 0.7 mm isotropic was used.
Participants:
Two young healthy females with normal or corrected eye acuity were scanned.
Analysis:
BrainVoyager 22.4 (Brain Innovation, Maastricht, NLs) was used for fMRI data analysis (3). Standard preprocessing was applied. We identified the patch cortical location using standard retinotopic mapping. In preliminary analyses, the response to the different experimental conditions within the cortical location of the patch was determined by averaging the BOLD response across repetitions and runs.
Results:
We successfully mapped the cortical location of the patch in both V1 and V2 in the two participants (Fig. 2A). Within the cortical location of the patch, the FIE-ATT condition showed a unique signal change compared to all other conditions (Fig. 2B). The signal increase during the block was in agreement with previous behavioral (1) and neurophysiological (2) findings.
Conclusions:
Our preliminary analysis demonstrates a correlate of perceptual filling-in using 7T fMRI at sub-millimiter resolution. These preliminary results seem to indicate that the expected response is only present when attention is directed to the patch. Future analyses will investigate the effect of filling-in and its interaction with attention at the laminar level.
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI)
Univariate Modeling
Novel Imaging Acquisition Methods:
BOLD fMRI
Perception, Attention and Motor Behavior:
Attention: Visual 2
Perception: Visual 1
Keywords:
Cortex
Cortical Layers
FUNCTIONAL MRI
HIGH FIELD MR
MRI
Perception
Univariate
Vision
Other - layer-fMRI
1|2Indicates the priority used for review
Provide references using author date format
2. De Weerd P, Gattass R, Desimone R, Ungerleider LG (1995), ‘Responses of cells in monkey visual cortex during perceptual filling-in of an artificial scotoma’, Nature vol 377 p731-734.
3. Goebel R, Esposito F, Formisano E (2006), ‘Analysis of FIAC data with BrainVoyager QX: From single subject to cortically aligned group GLM analysis and selforganizing group ICA’, Human Brain Mapping.