Linearity of the fMRI response in category‐selective regions of human visual cortex

Abstract

The goal of this study was to determine the linearity of the blood oxygen level-dependent (BOLD) response, as measured by functional magnetic resonance imaging (fMRI), in category-selective regions of human visual cortex. We defined regions of the temporal lobe that were selective to faces (fusiform face area, FFA) and places (parahippocampal place area, PPA). We then determined the linearity of the BOLD response in these regions to their preferred and nonpreferred stimuli. First, we tested the principle of scaling. As we increased the visibility of the stimulus, there was a corresponding linear increase in the fMRI signal in the FFA and PPA to their preferred stimulus (face and place, respectively). In contrast, responses in the FFA and PPA to the nonpreferred stimulus did not conform to the principle of scaling. Next, we asked whether the fMRI response in these regions of visual cortex conformed to the principle of additivity. To assess this, we determined whether the response to a long stimulus block could be predicted by adding the response to multiple shorter duration blocks. Although the fMRI response in the FFA and PPA was generally linear to the preferred stimulus, a more nonlinear response was apparent to the nonpreferred stimulus. In conclusion, the linearity of the BOLD response in the human ventral visual pathway varied across cortical region and stimulus category. This suggests that measures of linearity may provide a useful indication of neural selectivity in the brain.

Internal representations for face detection: An application of noise‐based image classification to BOLD responses

“…Second, it is reasonable to assume that any CI based upon activation in a single brain region may provide only a noisy and incomplete estimation of the overall internal template driving behavioral responses. For instance, in the case of face perception, its reliance upon an entire network of cortical regions is well‐documented [Gauthier et al, ; Haxby et al, ; Ishai et al, ; Rossion et al, ; Tsao et al, ] and consistent with the idea that these regions provide both redundant and complementary information for the purpose of face recognition [Fox et al, ; Gobbini and Haxby, ; Nestor et al, ]. Thus, the construction of hybrid CIs based from patterns of activation across multiple regions may ultimately provide a way to boost the quality of neurally‐derived CIs.…”

“…Second, face profile and object features encoded in the FFA are probably less diagnostic for their respective classes and, therefore, less robustly encoded—for instance, highly effective features for face detection like those shown in Figure b have difficulty in dealing with profiles. Third, the FFA does not appear to respond linearly to other categories than faces [Horner and Andrews, ] warning against the application of standard image classification to such cases. Therefore, we argue, the investigation of alternative types of stimuli in face‐selective regions is likely to be less informative.…”

“…This fact is important in that standard image classification is based on a linearity assumption: the magnitude/likelihood of a response increases linearly with the presence of a particular image structure. Recent research shows that face‐selective areas, but not other high‐level visual areas, exhibit this property in response to faces [Davidenko et al, in press; Horner and Andrews, ] warranting the application of image classification to their responses.…”

Measuring Cortical Activity During Auditory Processing With Functional Near-Infrared Spectroscopy

“…(21)). For example, Soltysik et al (23) reported that the auditory response obeys linearity for stimuli of a relatively long duration, but reveals nonlinear properties for short-duration stimuli (<10 s).…”

Modular processes in mind and brain

“…The failures of linearityinsuchstudies(suchasfailuresoftemporalsummation) generally don't tell us whether it is B(N ),or the neural response N , or both, that are nonlinear (Boynton & Finney, 2003, Miller, Luh, Liu, Martinez, Obata, et al, 2001Soltysik, Peck, White, Crosson, & Briggs, 2004). Generalization is problematic, because different brain regions may differ in linearity (Horner & Andrews, 2009;Soltysik et al, 2004). The authors of a study in which electrophysiological as well as fMRI measures were taken concluded that the source of nonlinearity in V1 is N ,notB(N ) (Wan,Riera,Iwata,Takahashi, Wakabahyashi, & Kawashima, 2006).…”