Examining the role of the core-face and extended network in  familiar face-identity processing in typically developed adults and developmental prosopagnosia subtypes

Dr. med. Sofia Martina Meyer

AutorIn

Dr. med. Sofia Martina Meyer

Titel

Examining the role of the core-face and extended network in familiar face-identity processing in typically developed adults and developmental prosopagnosia subtypes

Untertitel

a functional MRI study

Zitierfähige Url:

https://nbn-resolving.org/urn:nbn:de:bsz:14-qucosa2-947034

Erstveröffentlichung

2024

Datum der Einreichung

23.04.2024

Datum der Verteidigung

29.10.2024

Abstract (EN)

Introduction: Face-identity recognition is crucial for human communication and is likely based on a hierarchical process, initiated by face perception, followed by face-identity recognition and semantic association (Bruce & Young, 1986). At the neural level, this maps to cognitive models proposing a core-face and extended network that interact to support recognition. Posterior core-face regions process face perception (occipital face area: OFA; posterior fusiform face area: pFFA) and anterior regions (anterior FFA: aFFA; anterior temporal lobe face area: aTL-FA) familiar face-identity recognition. The extended network supports semantic association. Developmental prosopagnosia (DP) is a disorder marked by profound difficulties in recognizing faces, even highly familiar faces, in the absence of brain injury. To date, there is limited consensus regarding the underlying neural mechanisms of the disorder. Behavioral evidence highlights heterogeneity in DP, implying potential subtypes - apperceptive DP with face perception deficits, and associative DP with intact face perception, but difficulty recognizing familiar faces. Here, we applied a functional magnetic resonance imaging (fMRI) design with natural ambient images of familiar and unfamiliar faces to investigate and contrast the neural mechanisms of familiar and unfamiliar face-identity recognition in typically developed (TD) individuals and DPs who were meticulously characterized into potential behavioral subtypes. Hypotheses: Based on the hierarchical model of face-identity processing we predicted: H1) TD individuals recruit anterior core-face regions (aTL-FA and aFFA) during familiar face-identity processing and posterior core-face regions (OFA and pFFA) during unfamiliar face-identity processing. H2) DP individuals can be characterized into distinct behavioral phenotypes, i.e., apperceptive and associative subtypes, based on performance in face perception and face memory tasks. H3) The behavioral subtypes of DP are reflected in distinct neural response profiles: apperceptive DPs show overall reduced responses in all core-face regions to familiar and unfamiliar faces, while associative DPs show intact responses in posterior core-face regions, but reduced recruitment of anterior regions. Methods: To examine the core-face and extended network, functional imaging data was acquired using a multi-echo gradient echo planar imaging sequence optimized to image regions susceptible to signal drop-out, including the aTL, during the performance of a face-identity recognition task. In total, we recruited 33 TD and 18 DP participants. Before the fMRI task, DP cases were behaviorally characterized into an apperceptive and associative subtype according to their results on face memory and face perception tasks. Each DP subtype group was compared to an equal number of age- and sex-matched controls from the TD group. All participants were familiarized with one of two sets of face-identities via video. These identities were used as familiar face stimuli in the face-identity recognition task inside the scanner, with the untrained set serving as unfamiliar face stimuli. In the fMRI task, familiar or unfamiliar faces were presented in separate blocks (i.e., block design). In each block, participants determined if ‘test faces’ matched an initially presented ‘target face’. fMRI analyses focused on blood oxygenation level dependent (BOLD) responses during familiar and unfamiliar face-identity recognition within pre-defined regions of interest (ROIs): bilateral core-face network (OFA, pFFA, aFFA, aTL-FA) and extended network. Results: Behavioral deficits in 12 individuals were in accordance with an apperceptive DP subtype and 6 with an associative DP subtype. Behaviorally, in the face-identity recognition fMRI task, DPs performed significantly worse than matched controls during familiar (U = 12, p < 0.001) and unfamiliar (U = 12, p < 0.001) trials. On the neural level: the TD group recruited all pre-defined core-face regions during familiar and unfamiliar face-identity recognition. The OFA was more engaged during unfamiliar than familiar face-processing and responses in this region correlated with unfamiliar face-identity recognition performance. Unexpectedly, the aTL-FA was not more sensitive to familiar than unfamiliar faces. In DPs, distinct response profiles in apperceptive and associative subtypes were observed. Apperceptive DPs’ within-group analysis showed the aTL-FA is more sensitive to familiar than unfamiliar faces, while the OFA showed no increased sensitivity to unfamiliar faces. Compared to matched controls, they demonstrated overall reduced BOLD responses in specific core-face ROIs (i.e., bilateral OFA, pFFA, and aFFA) to unfamiliar faces (relative to familiar), most evident in the left OFA. In the same core-face ROIs increased responses to familiar than unfamiliar faces were detected. Responses in the bilateral OFA of apperceptive DPs positively correlated with unfamiliar face-identity recognition task performance. Associative DPs’ within-group analysis found no responses in the aTL-FA to either familiar or unfamiliar faces. Relative to matched controls, they had reduced responses in the right aFFA to both familiar and unfamiliar faces, as well as reduced left OFA responses to unfamiliar faces. In the extended network, both DP subtypes showed compensatory increased responses relative to their matched controls during familiar face-identity processing, with the apperceptive DP group exhibiting greater recruitment. Conclusion: In the TD brain, the present findings partially support a hierarchical model of face-identity processing: there was stronger recruitment of the bilateral OFA during unfamiliar face-identity recognition, underlining the increased perceptual processing necessary for recognizing unfamiliar faces. However, no elevated recruitment of anterior ROIs (i.e., aFFA, aTL-FA) during familiar face-identity recognition was found. In DPs, we find the characterization into subtypes is manifested in distinct neural response profiles. In apperceptive DPs, the behavioral face perception deficit may be attributed to reduced responses to unfamiliar faces in bilateral core-face ROIs (OFA, p/aFFA). The increased responses to familiar faces in core-face regions like the aTL-FA and extended network of this group may indicate compensatory recruitment or an interplay between the core-face and extended network. In associative DPs, we propose the exclusive face memory deficit may be reflected in reduced/missing responses in anterior core-face regions, i.e., aFFA and aTL-FA. Unexpectedly, the OFA of associative DPs does not respond like the TD group with higher sensitivity to unfamiliar faces. Associative DPs potentially recruit posterior core-face regions as compensation. Altogether, the findings emphasize the importance of bilateral posterior core-face regions for face perception, while challenging the notion of a strictly hierarchical face network. Based on these observations, we pose new questions regarding the organization of face-identity processing in TD and DP individuals and suggest a direct comparison of DP subtypes for future investigations.

Freie Schlagwörter (DE)

angeborene Prosopagnosie, fMRT, Gesichtererkennung

Freie Schlagwörter (EN)

prosopagnosia, fmri, face-identity,

Klassifikation (DDC)

610

Klassifikation (RVK)

YH 2504

GutachterIn