No abstract

Numerous neurological, developmental, and psychiatric conditions demonstrate impaired face recognition abilities, which can be socially debilitating. These impairments could be caused by either deficient face perceptual processes, such as reduced ability to integrate face parts into a whole, or deficient face memory processes, such as reduced ability to associate a face with semantic information. Research and clinical practice have focused more on developing face memory assessments, while it currently remains unclear which face perception assessments best captures perceptual deficits. A validated face perception measure could not only help with diagnosing the causes of face recognition deficits but could also help determine the most appropriate treatment. Here, we compare several available face perception assessments to identify those that can best assess perception deficits in developmental prosopagnosics. Thirty prosopagnosics and thirty age-matched neurotypical controls completed a battery of four face perception assessments, namely, computerized Benton Face Recognition Test, Cambridge Face Perception Test, University of Southern California Face Perception Test, and Telling Faces Together Test. They were also evaluated on two face recognition/ memory measures- Cambridge face memory test and famous faces memory test. We used logistic regression for the perception tests to predict prosopagnosic vs. control group membership and used multiple linear regressions to predict continuous objective and subjective measures of face recognition memory. Our results show that the Benton face test was the most reliable (α = 0.74), sensitive (AUC= 0.83), and predictive assessment of prosopagnosia diagnosis and face memory performance across the groups. The Cambridge face perception test also performed adequately well in terms of test sensitivity (AUC=0.80) and predicted face memory performance across the groups. Further, we found that face lighting change trials better predicted DP group membership and face recognition abilities than viewpoint-change trials. Together, these results have direct clinical application in assessing populations with face processing difficulties.

Previous face matching studies provide evidence that matching same identity faces (match trials) and discriminating different face identities (non-match trials) rely on distinct processes. For example, instructional studies geared towards improving face matching in applied settings have often found selective improvements in match or non-match trials only. Additionally, a small study found that developmental prosopagnosics (DPs) have specific deficits in making match but not non-match judgments. In the current study, we sought to replicate this finding in DPs and examine how individual differences across DPs and controls in match vs. non-match performance relate to featural vs. holistic processing abilities. 43 DPs and 27 controls matched face images shown from similar front views or with varied lighting or viewpoint. Participants also performed tasks measuring featural (eyes/mouth) and holistic processing (part-whole task). We found that DPs showed worse overall matching performance than controls and that their relative match vs. non-match deficit depended on image variation condition, indicating that DPs do not consistently show match- or non-match-specific deficits. When examining the association between holistic and featural processing abilities and match vs. non-match trials in the entire group of DPs and controls, we found a very clear dissociation: Match trials significantly correlated with eye processing ability (r=.48) but not holistic processing (r=.11), whereas non-match trials significantly correlated with holistic processing (r=.32) but not eye processing (r=.03). This suggests that matching same identity faces relies more on eye processing while discriminating different faces relies more on holistic processing.

The issue of the face specificity of recognition deficits in developmental prosopagnosia (DP) is fundamental to the organization of high-level visual memory and has been increasingly debated in recent years. Previous DP investigations have found some evidence of object recognition impairments, but have almost exclusively used familiar objects (e.g. cars), where performance may depend on acquired object-specific experience and related visual expertise. An object recognition test not influenced by experience could provide a better, less contaminated measure of DPs' object recognition abilities. To investigate this, in the current study we tested 30 DPs and 30 matched controls on a novel object memory test (NOMT Ziggerins) and the Cambridge Face Memory Test (CFMT). DPs with severe impairment on the CFMT showed no differences in accuracy or reaction times compared with controls on the NOMT. We found similar results when comparing DPs with a larger sample of 274 web-based controls. Additional individual analyses demonstrated that the rate of object recognition impairment in DPs did not differ from the rate of impairment in either control group. Together, these results demonstrate unimpaired object recognition in DPs for a class of novel objects that serves as a powerful index for broader novel object recognition capacity.

Previous face matching studies provide evidence that matching same identity faces (match trials) and discriminating different face identities (non-match trials) rely on distinct processes. For example, instructional studies geared towards improving face matching in applied settings have often found selective improvements in match or non-match trials only. Additionally, a small study found that developmental prosopagnosics (DPs) have specific deficits in making match but not non-match judgments. In the current study, we sought to replicate this finding in DPs and examine how individual differences across DPs and controls in match vs. non-match performance relate to featural vs. holistic processing abilities. 43 DPs and 27 controls matched face images shown from similar front views or with varied lighting or viewpoint. Participants also performed tasks measuring featural (eyes/mouth) and holistic processing (part-whole task). We found that DPs showed worse overall matching performance than controls and that their relative match vs. non-match deficit depended on image variation condition, indicating that DPs do not consistently show match- or non-match-specific deficits. When examining the association between holistic and featural processing abilities and match vs. non-match trials in the entire group of DPs and controls, we found a very clear dissociation: Match trials significantly correlated with eye processing ability (r=.48) but not holistic processing (r=.11), whereas non-match trials significantly correlated with holistic processing (r=.32) but not eye processing (r=.03). This suggests that matching same identity faces relies more on eye processing while discriminating different faces relies more on holistic processing.

The question of the face specificity of recognition deficits in prosopagnosia is fundamental to the organization of high-level visual processing and memory and has been vigorously debated numerous times over the last 20 years. This debate was recently revived in developmental prosopagnosics (DPs) by a meta-analysis reporting that a large percentage of DPs (ranging from 22-80%) have mild to major object recognition deficits (Geskin & Behrmann, 2017). Subsequent studies have also shown significantly lower DP group-level object recognition performance. However, previous investigations measuring object recognition have largely used familiar objects (e.g., cars) where performance may depend on object-specific experience, leaving the question open as to whether DPs would perform similarly when the stimuli are completely unfamiliar objects. To investigate this, in the current study we tested 30 DPs and 30 matched controls on a novel object memory test (NOMT Ziggerins) and the Cambridge Face Memory Test (CFMT). DPs were impaired on the CFMT but performed similarly to controls on the NOMT. We also compared DP NOMT performance to a larger matched sample of 274 web-based controls and found no difference in accuracy or reaction time. Individual analyses demonstrated that only one DP reached major impairment in object recognition on the NOMT, an impairment rate no different from that in either control group. Together, these results demonstrate essentially normal object recognition in DPs for a class of novel objects that serves as a powerful index for broader novel object recognition capacity. These results raise the possibility that DPs do not have object recognition deficits per se, but rather may have a decreased capacity to benefit from their experience with highly familiar object categories.

While age-related decline in face recognition memory is well established, the degree of decline in face perceptual abilities across the lifespan and their underlying mechanisms are incompletely characterized. In the current study, using the part-whole task, we sought to examine how age relates to facial feature discrimination ability and holistic face processing in a large sample of 3,341 online participants aged 18-69 years. We evaluated performance on the part-whole eye and mouth trials and the magnitude of the part-whole holistic advantage across the lifespan. We found that while discrimination of the eye region decreased beginning in the 50s, both mouth discrimination accuracy and the magnitude of the holistic advantage were stable with age. When investigating gender differences, we found that age-related declines in eye region accuracy were more pronounced in men than women, but this was not true for mouth accuracy or holistic processing. We discuss potential mechanistic explanations for this eye region-specific aging effect, including age-related hearing loss and its potential relationship with the age-related positivity effect.

Autism traits are commonly used as exclusionary criteria in studies of developmental prosopagnosia (DP). We investigated whether autism traits result in qualitatively different face processing in 43 DPs with high vs. low autism quotient (AQ) scores and 27 controls. Compared to controls, behavioral face recognition deficits were similar between the high and low AQ DP groups aside from worse emotion recognition in the high AQ DPs. Both DP groups showed reduced face selectivity in task-based fMRI, although higher AQ DPs showed decreased face selectivity in the posterior superior temporal sulcus. Resting-state fMRI showed similar face network connectivity between DP groups. This suggests that face processing is similar between the DP groups, with additional emotion processing deficits in higher AQ DPs.