تشخیص چهره پنهان در پروزوپاگنوزیا مادرزادی به روش گروه

Abstract Introduction Even though people with congenital prosopagnosia (CP) never develop a normal ability to “overtly” recognize faces, some individuals show indices of “covert” (or implicit) face recognition. The aim of this study was to demonstrate covert face recognition in CP when participants could not overtly recognize the faces. Methods Eleven people with CP completed three tasks assessing their overt face recognition ability, and three tasks assessing their “covert” face recognition: a Forced choice familiarity task, a Forced choice cued task, and a Priming task.

1. Introduction Individuals who do not report a history of brain injury yet have severe difficulties recognising faces are known as developmental (DP) or congenital prosopagnosics (CP) (Behrmann and Avidan, 2005, McConachie, 1976 and Duchaine, 2000). As many as 2–3% of the general population may find it very difficult to recognize faces (Bowles et al., 2009 and Kennerknecht et al., 2006) and these difficulties can run in families (Schmalzl et al., 2008, Lee et al., 2010 and Grueter et al., 2007). The impairment can be restricted to the recognition of facial identity, with no impairment recognising other facial cues such as expression and eye gaze (Duchaine et al., 2003a, Duchaine et al., 2003b, Duchaine et al., 2009, Humphreys et al., 2007 and Lee et al., 2010) or discriminating between other similar objects (Duchaine and Nakayama, 2005). However, in some cases face recognition impairments co-exist with more general difficulties in non-face (i.e., object) recognition (Duchaine et al., 2007, Lobmaier et al., 2010 and Wilson et al., 2010) or with difficulties perceiving biological motion (Dobel et al., 2007 and Lange et al., 2009). Whereas individuals with CP have failed to develop an adequate face recognition system, those with acquired prosopagnosia (AP) find themselves unable to recognize faces after brain damage resulting from incidents such as stroke or anoxia (Barton, 2008). Individuals with AP have severe deficits of overt (explicit) face recognition, in that previously familiar faces no longer give rise to feelings of familiarity and patients are unable to provide names or other identifying autobiographical information. However, some APs demonstrate covert (or implicit) recognition (see Schweinberger and Burton, 2003 for review). Covert recognition in AP has been demonstrated with physiological measures, as well as with different behavioural techniques. The most commonly used physiological technique to assess covert face recognition is the measurement of autonomic activity through Skin Conductance Responses (SCRs). Some people with AP display larger SCRs for familiar than unfamiliar faces ( Tranel and Damasio, 1995; Bauer, 1984), and this presence of differential autonomic arousal in the absence of overt face recognition to familiar faces has been interpreted as an index of covert recognition. Behavioural tasks are also frequently used to assess covert face recognition. Behavioural techniques include forced choice tasks, like the Forced choice familiarity task, where participants indicate which of two simultaneously presented faces is familiar, and the Forced choice cued task, where a printed name (cue) must be matched to the correct face (see Barton, 2008 for review). Despite their complete inability to recognize familiar faces, some APs choose the familiar face (in forced choice familiarity tasks) and/or match the name to the correct face (in forced choice cued tasks) significantly more often than chance ( Young and Hellawell, 1988, Sergent and Signoret, 1992, Diamond et al., 1994 and De Haan et al., 1991). Another behavioural task commonly used to assess covert recognition in AP is a Priming task, in which categorizing a printed name (e.g., Tony Blair) as either an actor or a politician is facilitated after viewing the same face, rather than a face from the opposite category (e.g., actor George Clooney) ( Barton et al., 2004 and Young and De Haan, 1988). In behavioural covert face recognition tasks, a distinction can be made between direct and indirect tasks (Barton et al., 2004). In direct tasks, participants are asked to make identity related decisions that directly involve the presented faces (e.g., Forced choice cued tasks and Forced choice familiarity tasks), whereas on indirect tasks, identity related decisions are measured by performance on another task, such as name classification (e.g., Priming task). Although some patients display covert recognition on both direct and indirect tasks, others only demonstrate covert recognition on one type (i.e., patient 008 displayed covert face recognition with a Forced choice cued task but not a Priming task) (Barton et al., 2004). Dissociations have also been reported between direct tasks; for example patient PH displayed covert recognition when assessed with a Forced choice cued task but not with a Forced choice familiarity task (Young and De Haan, 1988). The fact that performance on different behavioural covert recognition tasks can dissociate, both between and within participants, suggests that different behavioural tasks tap into different aspects of covert recognition (Barton et al., 2004 and Barton, 2008). Although some APs show covert recognition on a least a subset of behavioural tasks (Barton et al., 2001, Barton et al., 2004, Young and De Haan, 1988 and Damian and Rahman, 2003), others fail to show any signs of behavioural covert face recognition (Sergent and Villemure, 1989 and De Haan and Campbell, 1991). It is not entirely clear why some patients show covert face recognition and others do not. However, there appears to be a link between overt processing and behavioural covert face recognition, with patients with more severe overt recognition deficits less likely to demonstrate behavioural covert recognition. This association has been used to argue that behavioural covert recognition relies on the same system that supports overt face recognition (Barton, 2008 and Schweinberger and Burton, 2003). Given that people with CP have never developed an adequate face recognition system, it is of considerable theoretical importance to determine whether covert recognition can be demonstrated in this population, and if so, the conditions under which it is present. Early case studies addressing this issue typically failed to demonstrate covert face recognition in CP. Case YT did not show covert recognition on an Interference task, which required categorizing a name superimposed on either related or unrelated faces (Bentin et al., 1999). Similarly, case AB did not display covert recognition when assessed with either a Matching task, a Priming task, or a Forced choice familiarity task (De Haan and Campbell, 1991). Results such as these support the proposal that covert face recognition is only apparent when previously intact face representations have been damaged (as in AP), but not when face representations have never been formed (as is presumably the case in CP) (Barton et al., 2004). More recent studies have, however, demonstrated behavioural covert face recognition in cases of CP. Case “C” demonstrated covert recognition when assessed with a Forced choice cued task (although not with a Forced choice familiarity task or a Priming Task) (Rivolta et al., 2010). The existence of covert recognition in CP has also been supported by the finding that visual scan paths to novel faces are also different to those made to famous (Case “AA”; Bate et al., 2008) or recently studied (Cases “MZ”, “RW” & “WS”; Bate et al., 2009) faces. In the only group study published to date, six people with CP were quicker, and more accurate, at classifying whether two sequentially shown faces were the same or different when the faces were famous than when they were unknown, demonstrating covert recognition (Avidan and Behrmann, 2008). Avidan and Behrmann did attempt to control for the fact that CPs can overtly recognize some faces (on average, ∼40%), by analysing only the faces of famous individuals which CPs could not overtly recognize a few weeks earlier. However, the overt recognition test only included one of the five images of each famous individual used in the matching task, leaving open the possibility that the participant might have been able to recognize some, or all, of the other images, particularly if they were more iconic images. The aim of the current study was to determine whether behavioural covert recognition could be reliably observed in a group of CPs. Our study differs from the only previous group study (Avidan and Behrmann, 2008) in three ways. First, we tested a larger sample size (n = 11). Second, we used three covert face recognition tasks that have been shown to be sensitive to covert recognition in AP. The use of multiple measures will allow us to examine whether covert recognition occurs on all, none, or just some measures in CP. Third, we implemented two crucial features in our experimental design to ensure that we could be confident that covert recognition was actually “covert”. That is, for each participant we assessed overt recognition of all the face images used in the tests of covert recognition, and subsequently excluded all overtly recognized faces from the analysis. In addition, participants were asked to report their confidence in their responses, allowing us to differentiate tasks where performance was associated with insight from those where it was not, and thus to gauge whether performance could confidently be described as “covert”. No association between accuracy and confidence ratings would suggest that participants did not have insight into their performance on the task, and as such the task truly reflects covert recognition. In order to evaluate the degree of impairment of face recognition difficulties, participants initially completed reliable tests of face perception and memory, and a test of famous face recognition. To investigate covert face recognition in our group of CPs, we tested each participant with three behavioural covert recognition tasks: (1) a Priming task, (2) a Forced choice cued task, and (3) a Forced choice familiarity task. Both direct (Forced choice cued task and Forced choice familiarity task) and indirect (Priming) tasks were selected to increase the likelihood of detecting potential differences in sensitivity between covert recognition tasks. Finally, we examine whether there were correlations between performance on the different covert face recognition tasks and whether there was a relationship between overt and covert performance.

4. Conclusions In the current study we describe a detailed assessment of covert face recognition skills in a group of 11 individuals with CP with no overt recognition of the face images presented. The results of our study, which are in line with some aspects of previous research in both AP and CP, confirm that covert face recognition is in fact a feature of CP, and that different tasks can tap into different aspects of covert face recognition. More specifically, our results suggest that the Forced choice familiarity task represents a sensitive task for the detection of covert face recognition, whereas the Priming task does not. Lastly, we propose that the Forced choice cued task does not really provide an index of covert face processing as such, but rather of provoked-overt recognition.