Purchase this article with an account.
Marieke Mur, Douglas Ruff, Jerzy Bodurka, Peter Bandettini, Nikolaus Kriegeskorte; Ranking 96 object images by their activation of FFA. Journal of Vision 2008;8(6):408. doi: https://doi.org/10.1167/8.6.408.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The fusiform face area (FFA) is a region in human inferior temporal cortex that has been shown to respond most strongly to faces. Previous imaging studies only assessed category-average activation as they grouped stimuli into predefined natural categories. Here we ask whether there are particular non-face object images that elicit a strong FFA response or particular face images that elicit a weak response. In addition, we investigate whether specific faces consistently elicit a stronger FFA response than others. To address these questions, we rank 96 particular object images by the activation they elicit in FFA. Blood-oxygen-level-dependent fMRI measurements were performed at high resolution (voxel size 1.95×1.95×2 mm3), using a 3T scanner. First, FFA was defined conventionally at varying sizes using a separate block design experiment. Then, the activation in those voxels in response to 96 different object photos was measured in two separate sessions, as subjects performed a fixation-cross-color discrimination task. FFA responses to the 96 object images were ranked according to response amplitude. Group results (n=4) indicated that single-image activation of both left and right FFA was stronger for face images than most other object images (average choice probability = .92). This result was clearest for maximally face-selective FFA voxels. Right FFA responses were stronger and more robust against increasing size than left FFA responses. Activation in the parahippocampal place area (PPA) and early visual cortex (EVC) did not rank faces before most other object images. These results were consistent across sessions and subjects. Preliminary further analyses failed to show evidence for consistency of within-face ranking orders either within or across subjects. This would suggest that the FFA activation profile is flat across different individual faces.
This PDF is available to Subscribers Only