Developmental prosopagnosia: A case analysis and treatment study

Abstract

This paper reports a treatment case study focused on face perception impairments designed for AL, an 8-year-old child with prosopagnosia. AL's prosopagnosia was characterized by deficits at the level of structural encoding—that is, he was unable to achieve normal basic perception of faces. This impairment then impacted on all subsequent aspects of familiar- and unfamiliar-face processing. Detailed assessment of feature processing revealed impairments in perception of facial features with a dissociation between relatively good perception of the mouth feature and poor perception of eye and nose features. Interestingly, results also suggested at least partial internal representation of facial features despite long-standing deficits in perception of these features. A treatment programme focused on training in perception, and analysis of facial features and familiar-face naming was conducted. Treatment resulted in excellent face naming for familiar faces, a decreased reliance on nonfacial cues and a reduction in AL's tendency to misidentify unfamiliar faces as family members.

Ruth Brunsdon was supported by a Department of Education, Science and Training Australian Post Graduate Award and Lyndsey Nickels by an Australian Research Council QEII Fellowship during the preparation of this paper.

The authors are grateful to Dr Alan Taylor for statistical advice. We are indebted to our client AL and his family for their enthusiastic participation throughout the project. We would like to thank Nicole Gow, an undergraduate psychology student, who volunteered her time to help develop some of the face test stimuli.

Notes

¹ In this paper we use the term face recognition to refer to recognition of a visually presented face stimulus as familiar. This is not to be confused with recognition of a person as familiar. To this end we argue that face recognition occurs at the level of face recognition units (like Bruce & Young, 1986).

² Polster and Rapcsak's (1996) report on unfamiliar-face learning in RJ, a densely prosopagnosic adult, is not included in this review. Their study investigates the effects of different encoding instructions on recognition memory for unfamiliar faces, but does not represent a true rehabilitation study.

³ Sergent and Poncet (1990) also failed to demonstrate lasting improvements following category presentation in their case study of PV, an adult with acquired prosopagnosia.

⁴ AL was 7 years old when assessment commenced but turned 8 during the pretreatment assessment phase. Published age-based norms were used accordingly—that is, 7-year-old norms were used when AL was aged 7, and 8-year-old norms were used after his birthday.

⁵ Each unfamiliar face was matched (according to gender, age, hair-style, and presence or absence of a beard and/or glasses) as closely as possible to one of the familiar-face photos.

⁶ Due to copyright restrictions the actual stimuli used in this study could not be reproduced in Figures 2 and 3, however suitably equivalent stimuli have been constructed for illustrative purposes.

⁷ Response time was recorded using the timing function in Microsoft PowerPoint. Response time was measured from the moment of stimulus display to AL's verbal response.

⁸ GEE is an approach to the analysis of correlated response data, which is particularly useful when the responses are binary. It takes account of the correlations between observations and uses them when calculating both parameter estimates and their standard errors. Interested readers can refer to Hanley, Negassa, Edwardes, and Forrester (2003) for more information.

⁹ It could be argued that this finding was simply a task complexity effect (i.e., the visual complexity of looking at and comparing two faces is greater than that when just comparing two noses/eyes/mouths). We would argue against this for two reasons. First, AL was able to perform the “within-face” task relatively well with mouth stimuli (as described in the next section), which suggests that he was able to cope with the task demands. Second, he was able to perform other face discrimination tasks of a similar complexity (e.g., Benton Face Recognition Test, Items 1–6), suggesting that in general he was able to cope with complex front-view discrimination tasks.

¹⁰ We do not argue that AL's processing of the mouth region was normal (as the current control was not age matched), just that it was relatively intact when compared with other facial regions.

¹¹ Spence (1995, p. 51) states “my previous research (Milne & Spence, 1987) demonstrated that most children over the age of 8 are able to decode facial expressions correctly”.

¹² Although KD was not strictly a developmental case, her early insult (meningitis at 14 months followed by numerous shunt infections and surgical shunt revisions and also anaesthetic complications when aged 3) warrants her to be mentioned here.

¹³ Relatively intact visual object imagery has been reported in other cases of developmental visual agnosia and prosopagnosia characterized by structural encoding deficits (KD, Young & Ellis, 1989; EP, Nunn et al., 2001; LG, Ariel & Sadeh, 1996).