Interactive reportActivity in fusiform face area modulated as a function of working memory load1
Introduction
Pathology studies of prosopagnosic patients introduced the notion that a brain area might exist which is specialized for the perception of human faces [19]. Prosopagnosic patients have selectively impaired recognition and perception of faces compared with other visual images. These patients frequently demonstrate lesions in occipitotemporal areas of visual cortex, usually showing right-lateralized or bilateral damage. Primate single-unit recording experiments subsequently documented the presence of both face-selective and object-selective neurons in the inferior temporal cortex (IT) and superior temporal sulcus (STS) [8], [9], [27], [40], [43], [49]. In humans, the fusiform face area (FFA) is the suggested functional and anatomical homologue to primate face-selective IT [3], [32], [41], [42]; portions of human STS showing face- and gaze-selectivity are believed homologous to primate STS [30]. Investigations of the FFA have generally identified a small area of the fusiform gyrus which shows a greater response to visually presented faces than other visual stimuli [3], [32], [41], [42], [44]. Recent experiments have described the FFA as an ‘expert’ processor [25], but there is little evidence that this region might have functions beyond perception. With the exception of studies looking at the effects of attention in the FFA [38], [47], the FFA has been exclusively examined as a perceptual module.
We believe that the functions of the FFA should extend beyond the perception of faces or other stimuli requiring expert processing. Current models of object working memory predict a significant role for the FFA in working memory for faces. Working memory has been broadly defined as a cognitive system that permits short-term, active maintenance and manipulation of stimulus information when that information is not available in the environment [4], [5]. Data from primate lesion [7], [21], [24] and single-unit recording studies [15], [23], [24], [33], [35], [36], [37] have described different but complementary roles for IT and PFC in object working memory. While PFC neurons robustly maintain stimulus-specific information across the delay period of a working memory task [23], [24], [35], IT neurons show more subtle delay effects [15] and robust stimulus-specific effects at target presentation [24], [33], [36], [37]. The response properties of neurons in these two regions suggest that activity in IT is biased in favor of behaviorally relevant stimuli by virtue of top-down feedback from prefrontal structures [14], [18]. If the FFA is homologous to face-selective regions of IT then FFA should have working memory properties similar to IT when the task involves face stimuli.
In the present study, we hypothesized that the FFA would demonstrate increased activity with increased working memory load for human faces. If a positive bias on a subgroup of neurons within the FFA is associated with maintaining a behaviorally relevant face, it is reasonable to believe that maintenance of more faces would involve recruitment of additional subgroups of FFA neurons. Previous work has suggested the involvement of fusiform gyrus in face working memory [17], [26], [28], [29], [50], but in various ways these inferences have been confounded by the perceptual demands of the task. As the FFA is known to be active under conditions of face perception, these confounds make it difficult to state whether working memory was the source of the FFA activity in these studies. Since the present study used a working memory task which controlled for perceptual demands across task conditions, we were in a position to make stronger statements about changes in FFA activity specific to working memory.
To test our hypothesis, we used fMRI to measure neuronal activity in humans performing an n-back task for gray-scale faces (Fig. 1). Given the similarity to previous visuospatial n-back studies, we expected to replicate previous findings of increased activity with load in the PFC and posterior parietal areas [10], [11], [16], [31]. The novel aspects of the experiment were: (1) the use of face stimuli in the n-back task, and (2) a separate experiment to functionally define the FFA. These features allowed us to simultaneously analyze the PFC and the FFA, an easily identifiable portion of extrastriate visual cortex. Using the same logic employed by other investigators to correlate activity in frontal and parietal regions with working memory, we were able to determine that working memory load for faces modulated FFA activity.
Section snippets
Subjects
Nine right-handed subjects (age range 21–27, 5 male and 4 female) were recruited from the University of Pennsylvania Medical School. All participants were screened against medical, neurological, and psychiatric illnesses, and also for use of prescription medications. The day prior to the scan, subjects were behaviorally trained to an overall 85% performance criterion.
Behavioral task (Fig. 1)
Subjects were presented with an instruction (0-back, 1-back, or 2-back) for 4 s, followed by a sequence of 13 faces. The face
Behavioral results
Behavioral results are summarized in Table 1. As expected, subjects indicated that the 2-back task was much more difficult than either the 0-back or 1-back task. This increased difficulty is reflected in the lower accuracy and higher reaction time for the 2-back task. The overall accuracy of 86.2±3.2% was close to the 85% criterion used for training the subjects.
fMRI results in fusiform face area and fusiform object area
The FFA was identified in a separate localizer run in all nine subjects (Fig. 2). The FFA was defined as the set of voxels within the
Discussion and references
The n-back task is an elegant paradigm that has allowed experimenters to isolate areas of the brain which vary with working memory load-dependent aspects of the task [10], [11], [16], [31]. In n-back experiments, subjects perform a working memory task where they decide when a serially-presented stimulus matches the one presented n stimuli ago. By parametrically varying n, the task highlights processes that vary with the load of information maintained and/or manipulated in working memory. As a
Acknowledgements
Supported by the American Federation for Aging Research and NIH grants NS01762, AG13483 and AG15793 and NIH Training Grant #T32-GM07517 (Graduate Training in Systems and Integrative Biology). We thank Charan Ranganath for helpful comments on the manuscript.
References (51)
- et al.
The inferential impact of global signal covariates in functional neuroimaging analyses
Neuroimage
(1998) - et al.
Dissociating working memory from task difficulty in human prefrontal cortex
Neuropsychologia
(1997) - et al.
Selectivity between faces in the responses of a population of neurons in the cortex in the superior temporal sulcus of the monkey
Brain Res.
(1985) - et al.
A parametric study of prefrontal cortex involvement in human working memory
Neuroimage
(1997) - et al.
Effects of cooling inferotemporal cortex on performance of visual memory tasks
Exp. Neurol.
(1981) - et al.
Age-related changes in regional cerebral blood flow during working memory for faces
Neuroimage
(1998) - et al.
Distinguishing the functional roles of multiple regions in distributed neural systems for visual working memory [see comments]
Neuroimage
(2000) - et al.
Suppression of visual responses of neurons in inferior temporal cortex of the awake macaque by addition of a second stimulus
Brain Res.
(1993) - et al.
Analysis of fMRI time-series revisited — again [comment]
Neuroimage
(1995) - et al.
A trial-based experimental design for fMRI
Neuroimage
(1997)
Empirical analyses of BOLD fMRI statistics. I. Spatially unsmoothed data collected under null-hypothesis conditions
Neuroimage
Experimental design for brain fMRI
Face recognition in human extrastriate cortex
J. Neurophysiol.
Working memory
Science
Dementia and working memory
Q. J. Exp. Psychol. [A]
Delayed-matching and delayed-response deficit from cooling dorsolateral prefrontal cortex in monkeys
J. Comp. Physiol. Psychol.
Functional subdivisions of the temporal lobe neocortex
J. Neurosci.
Distribution of cortical activation during visuospatial n-back tasks as revealed by functional magnetic resonance imaging
Cereb. Cortex
Attribute-based neural substrates in temporal cortex for perceiving and knowing about objects
Nat. Neurosci.
The physiological basis of attentional modulation in extrastriate visual areas
Nat. Neurosci.
Responses of neurons in inferior temporal cortex during memory-guided visual search
J. Neurophysiol.
A neural basis for visual search in inferior temporal cortex [see comments]
Nature
Temporal dynamics of brain activation during a working memory task
Nature
Transient and sustained activity in a distributed neural system for human working memory
Nature
Neural mechanisms for visual memory and their role in attention
Proc. Natl. Acad. Sci. USA
Cited by (142)
Neurocognitive basis of model-based decision making and its metacontrol in childhood
2023, Developmental Cognitive NeuroscienceThalamocortical contributions to working memory processes during the n-back task
2023, Neurobiology of Learning and MemoryCitation Excerpt :In addition to working memory load, cortical activity can be modulated by the types of stimuli maintained in working memory. For instance, when stimuli from different visual categories are presented, category-selective regions in the inferior temporal cortex, such as the fusiform face area (FFA) and the parahippocampal place area (PPA; O’Craven & Kanwisher, 2000), show increased activity scaled with working memory load for the respective categories (Druzgal & D’Esposito, 2001; Lepsien & Nobre, 2007; Ranganath et al., 2004). It is thought that these category-specific modulations are under the influence of top-down biasing signals from frontoparietal regions (D’Esposito & Postle, 2015), which enhance activity that encodes categorical working memory information to sustain working memory representation (Hwang et al., 2019, 2020; Lee & D’Esposito, 2012; Miller et al., 2011; Miller & D’Esposito, 2005).
Clarifying the role of higher-level cortices in resolving perceptual ambiguity using ultra high field fMRI
2021, NeuroImageCitation Excerpt :Straightforward manipulation of task demands can produce changes in the amplitude of BOLD response in higher level areas, such as the FFA (Druzgal and D'Esposito, 2001; Kay and Yeatman, 2017; Vuilleumier et al., 2001; Wojciulik et al., 1998) or STS (Narumoto et al., 2001). Traditionally, however, these tasks manipulations are often indirectly related to the neural processing of the stimuli at hand, as they serve a more general purpose of directing attention towards (e.g. N-back) or away from the stimuli (e.g. fixation tasks, see Bokde et al., 2005; Druzgal and D'Esposito, 2001; Egner and Hirsch, 2005; Kay and Yeatman, 2017; Wojciulik et al., 1998). Changes in neural responses to identical visual inputs related to these unspecific changes in task demands often reveal broad contributions from attentional networks, including frontal and parietal regions (Szczepanski et al., 2013).
Dynamic range in BOLD modulation: lifespan aging trajectories and association with performance
2017, Neurobiology of AgingA social-semantic working-memory account for two canonical language areas
2023, Nature Human Behaviour
- 1
Published on the World Wide Web on 5 October 2000.