Brain Imaging and Behavior DOI 10.1007/s11682-013-9285-5

NEUROIMAGING AND REHABILITATION SPECIAL ISSUE

Neuroimaging and facial affect processing: implications for traumatic brain injury Dawn Neumann & Michelle A. Keiski & Brenna C. McDonald & Yang Wang

# Springer Science+Business Media New York 2013

Abstract The ability to recognize others’ emotions is critical to successful interpersonal interactions. Given its importance, there has been an extensive amount of research using functional magnetic resonance imaging (fMRI) to investigate the neurobiological mechanisms associated with facial affect recognition in healthy individuals, and some in patient populations with affective disorders. Findings from these studies reveal that the underlying mechanisms involve a distributed neural network, engaging structures within limbic and subcortical regions, prefrontal cortex, temporal and parietal lobes, and occipital cortex. In the last several decades, researchers have become increasingly interested in how emotion recognition is affected after a traumatic brain injury (TBI), which often involves damage to these structures, as well as the neural circuitry connecting them. Not surprisingly, research has reliably demonstrated that facial affect recognition deficits are common after TBI. To date, however, no neuroimaging studies have investigated facial affect recognition D. Neumann (*) Department of Physical Medicine and Rehabilitation, Rehabilitation Hospital of Indiana, Indiana University School of Medicine, 4141 Shore Drive, Indianapolis, IN 46254, USA e-mail: [email protected] M. A. Keiski Department of Physical Medicine and Rehabilitation, Indiana University Center for Neuroimaging, Indiana University School of Medicine, 355 W. 16th St., GH Suite 4100, Indianapolis, IN 46202, USA e-mail: [email protected] B. C. McDonald : Y. Wang Department of Radiology and Imaging Sciences, Indiana University Center for Neuroimaging, Indiana University School of Medicine, 355 W. 16th St., GH Suite 4100, Indianapolis, IN 46202, USA B. C. McDonald e-mail: [email protected] Y. Wang e-mail: [email protected]

deficits in the TBI population. Consequently, the purpose of this paper is to consider how functional magnetic resonance imaging (fMRI) might inform our knowledge about affect recognition deficits after TBI, and potentially enhance treatment approaches. Keywords Neuroimaging . Emotion . Facial affect recognition . Traumatic brain injury Most people engage in interpersonal interactions on a daily basis, whether it is with family members, a significant other, friends, colleagues, or general acquaintances. A large part of our interpersonal interactions involves an exchange of nonverbal affect; hence, we express our emotions and recognize others’ emotions, through facial and vocal expressions, gestures, and postures (Nowicki and Mitchell 1998). Recognizing how our social counterpart feels is important for guiding our own emotions and context-appropriate behavior (Fusar-Poli et al. 2009). Of the nonverbal cues, facial expressions are believed to be the most prominent for communicating emotions (Posamentier and Abdi 2003; Radice-Neumann et al. 2007). As such, it is not surprising that studies have found facial affect recognition deficits to be linked with inappropriate social interactions and behaviors, as well as emotion dysregulation (Croker and McDonald 2005; Hopkins et al. 2002; Hornak et al. 1996; Knox and Douglas 2009; Mueser et al. 1996; Nowicki and Mitchell 1998). Given the importance of facial affect recognition, it has been a strong focus for many neuroimaging and behavioral studies in both healthy controls and populations with affective disorders. Since the mid-1990s, an abundance of functional neuroimaging studies have been conducted in an effort to understand the neurobiological mechanisms that underlie facial affect perception and recognition (Fusar-Poli et al. 2009; Sabatinelli et al. 2011). As a result, several critical reviews (Adolphs 2002; Calder and Young 2005; Phillips et al. 2003; Posamentier and

Brain Imaging and Behavior

Abdi 2003) and meta-analyses (Fusar-Poli et al. 2009; Sabatinelli et al. 2011) have emerged to help consolidate the findings across healthy populations. Additionally, neuroimaging has also been used to study affect recognition deficits in several patient populations, such as those with Parkinson’s disease (Delaveau et al. 2009; Ibarretxe-Bilbao et al. 2009; Tessitore et al. 2002), schizophrenia, and autism (Fakra et al. 2008; Piggot et al. 2004; Wang et al. 2004). Findings from these neuroimaging studies have helped advance the knowledge and treatment of emotion recognition deficits in these populations with affect disorders. In the last several decades, researchers have become increasingly interested in how emotion recognition is affected after a traumatic brain injury (TBI). Survivors of TBI are often confronted with a variety of physical, cognitive, emotional and behavioral deficits (Draper et al. 2007; Hoofien et al. 2001; McDonald et al. 2002). A growing body of literature demonstrates that facial affect recognition is frequently compromised in people with TBI (Babbage et al. 2011; Jackson and Moffat 1987; McDonald and Flanagan 2004; Milders et al. 2003; Radice-Neumann et al. 2007; Radice-Neumann et al. 2009). In fact, a recent meta-analysis on facial affect recognition indicates that between 13 and 39 % of people with moderate to severe TBI are significantly impaired at recognizing emotions from facial expressions (Babbage et al. 2011). With an estimated 5.3 million Americans living with a chronic TBI-related disability, the potential number of people in the U.S. alone who might have this deficit is alarming (“Centers for Disease Control and Prevention and last accessed July 2 2013,”). This presents a significant concern as these deficits are thought to be an important component of larger problems that frequently occur after TBI. For instance, people with TBI commonly experience emotional and behavioral problems, as well as impoverished relationships (Eames and Wood 2003; Gainotti 1993; Lezak and O’Brien 1988; Oddy et al. 1985). Although a whole host of factors are likely to be contributing to poor social and behavioral outcomes after TBI, findings from several studies suggest that impaired affect recognition is indeed a relevant factor (Cooper et al. 2013; Knox and Douglas 2009; RadiceNeumann et al. 2007; Ryan et al. 2013). For instance, two recent studies found a significant correlation between emotion perception and poor social participation in participants who had had a brain injury (Cooper et al. 2013; Knox and Douglas 2009). Consequently, it has been acknowledged that there is a strong need to learn about and to treat facial affect impairments in the TBI population (Driscoll et al. 2011). Although a few interventions have been designed to improve facial affect recognition after TBI (Bornhofen and McDonald 2008a, b; McDonald et al. 2008; Radice-Neumann et al. 2009), very little is currently known about the mechanisms that underlie this defict in people with TBI.

To date, no neuroimaging studies have investigated facial affect recognition deficits in the TBI population. Consequently, the purpose of this paper is to consider how functional magnetic resonance imaging (fMRI) might inform our knowledge about affect recogntion deficits after TBI, and potentially enhance treatment approaches. In order to accomplish this, we will discuss several relevant topics. First, we begin with a brief discussion of facial affect recognition theories, in which we will describe different elements or cognitive/emotional processing components thought to contribute to the successful recognition of others’ emotions; hereafter, referred to as “elements of facial affect recognition”. This will be followed by an overview of facial affect recognition deficits after TBI, and the current state of treatments for this population. Next, we review outcomes from fMRI studies, highlighting brain regions and networks most consistently associated with facial affect processing. The next section integrates the elements of facial affect recognition with the neuroimaging findings, in order to highlight the associations between brain structures and function that are relevant to facial affect recognition. The remaining sections discuss the implications, benefits, and challenges of fMRI research to advance the current knowledge and treatments for improving facial affect recognition after TBI, as well as a discussion of the potential utility and pitfalls of conducting neuroimaging studies to investigate facial affect recognition deficits in people with TBI.

Facial affect recognition theories The following elements have been postulated to be important contributors to facial affect recognition: Perception, Emotion Replication and Experience, and Conceptual Understanding (Adolphs 2002; Phillips 2003; Phillips et al. 2003; RadiceNeumann et al. 2007). Each element is briefly described below. Perception Beyond early, granular-level visual processing, most theories regarding facial affect recognition postulate that faces are largely processed as a global percept, such that the features are perceived as a collective unit; this is referred to as global or holistic processing (Haxby et al. 2000; Posamentier and Abdi 2003; Radice-Neumann et al. 2007). There is also evidence to suggest that facial features are, to a degree, processed one feature at a time (i.e., eyebrows, eyes, mouth) (Adolphs 2002; Posamentier and Abdi 2003; Radice-Neumann et al. 2007); this is referred to as separable processing. This is consistent with the theory that certain face features are more emotionally salient and informative than others (Adolphs 2002; Phillips 2003; Radice-Neumann et al. 2007). For example, evidence indicates that the eyes of the face are a very emotionally

Brain Imaging and Behavior

salient feature, and thus believed to receive more visual attention than other features (Adolphs 2008). Emotion replication and experience Another element thought to contribute to facial affect processing is reproduction of the emotion in question (Adolphs 2002; Phillips 2003). This theory suggests that people better recognize how others are feeling when they can generate (replicate) and feel (experience) that same emotion within themselves. In other words, they recognize the emotion by personally identifying with a similar emotional experience, which is in part accomplished by replicating the facial expression (Bastiaansen et al. 2009). This assumes accompanying autonomic responses (i.e., systemic arousal, heart rate, body temperature) and changes in body state (i.e., our muscle tone; facial expression) that are concordant with the emotion. As subtle as the replication may be, numerous studies suggest that people often mirror others’ expressions, either just internally, or even to some degree externally (Bastiaansen et al. 2009; Carr et al. 2003; Iacoboni and Mazziotta 2007). Interestingly, findings from studies also show that imagined and executed actions have overlapping neural substrates (Decety 1996). Several studies further demonstrate that just observing an emotional face will often trigger an autonomic response (involving arousal and changes in heart rate) that is assumed to reflect changes in one’s emotional state (Critchley et al. 2005; Hopkins et al. 2002). Conceptual understanding Throughout the course of ones’ development and personal experiences, it is expected that schematic representations of emotions are constructed along with an associated conceptual understanding of that emotion (Damasio 1994). It is believed that people use past history and their knowledge about emotions/emotional expressions and surrounding contexts to identify how others are feeling (Adolphs 2002). For example people might recall watching a movie when an actor with raised eyebrows and wide eyes screamed because of a threatening situation; therefore, people are likely to assume that a person with these same traits and a similar situation is also fearful. In essence, one generates a concept of the emotion (i.e., fear) through associated knowledge, which in turn helps to recognize others’ facial expressions.

TBI and facial affect recognition A TBI is the result of an external physical force being applied to the head, whether a direct blow and/or accelerationdeceleration forces without a direct impact, that causes a

change in brain function and/or brain pathology (“Centers for Disease Control and Prevention and last accessed July 2 2013,”). A TBI can either be a closed or penetrating injury. A penetrating injury occurs when the brain is pierced or penetrated by an object, such as a bullet or knife, mainly resulting in localized damage within and surrounding the object’s trajectory within the brain. In contrast, closed TBIs (most often from motor vehicle accidents and falls) characteristically result in damage that is more distributed throughout the brain. These events may cause the brain to strike bony surfaces within the cranial vault, resulting in contusions not only at the site of impact, but also the opposite side of the skull (Coup-contrecoup injuries). (“Centers for Disease Control and Prevention and last accessed July 2 2013,”). In addition to the contusions, the movement of the brain and/or rotational forces frequently cause axons inside the brain to shear, leading to diffuse axonal injuries (DAIs) and ultimately the disruption of various neural pathways and circuits (Scheid et al. 2003). The orbital frontal cortex, temporal lobes, limbic system, and the cortical-cortical and cortical-subcortical connections to and from these regions are often vulnerable to the damage caused by a TBI; discussed later in further detail, these areas and networks are known to be important for processing and recognizing facial affect (Dima et al. 2011; Scheid et al. 2003; Singh et al. 2010). Studies investigating facial affect perception and recognition in people with TBI have found that they are often impaired at identifying (labeling), matching, and discriminating facial expressions (Green et al. 2004; Ietswaart et al. 2008; Jackson and Moffat 1987; Milders et al. 2003; RadiceNeumann et al. 2009; Spell and Frank 2000). Some studies have suggested that people with TBI predominantly have trouble recognizing negative emotions (Croker and McDonald 2005; Hopkins et al. 2002; Jackson and Moffat 1987). However, it is currently undetermined if these findings are actually attributable to the valence of the stimuli (negative versus positive), or whether negatively valenced facial expressions are just more difficult to recognize than positively valenced facial expressions. In contrast to past speculations, one recent study found evidence that people with TBI have difficulty recognizing positive affect as well (Zupan and Neumann 2013). In this study, the authors classified people with TBI as impaired or not impaired at facial affect recognition using scores on a standardized test. People with TBI who were impaired at facial affect recognition had significantly more recognition errors for all emotions, including happy, compared to people with TBI who had normal affect recognition. Finally, newer research is beginning to suggest that people with TBI also seem to have difficulty recognizing neutral faces, often misinterpreting them for a negative emotion (Zupan and Neumann 2013). Deficits in facial affect recognition after TBI do not appear to be a simple perceptual impairment or a deficit in general

Brain Imaging and Behavior

face perception. Research suggests that face perception and facial affect recognition are dissociable (Winston et al. 2004). Moreover, most participants in affect recognition studies have not been found to have problems with general face perception. Additionally, most studies excluded participants for visual neglect and/or significant visuoperceptual difficulties. Certain cognitive functions have been correlated with facial affect recognition. A recent publication investigated the relationships between facial affect recognition and cognitive functioning in 70 adults with moderate to severe TBI (Yim et al. 2013). Findings from this study indicated that attention, nonverbal memory, working memory, delayed verbal memory and speed of processing were significantly related to facial affect recognition. An executive functioning test (set shifting task) was not correlated with any of the facial affect recognition measures. Given the significance and frequency of impaired affect recognition, a handful of affect recognition interventions have recently been developed and tested for adults with TBI (Bornhofen and McDonald 2008a, b; McDonald et al. 2009; McDonald et al. 2008; Radice-Neumann et al. 2009). The few treatments that have been published appear to address the core elements of facial affect recognition theory: perception, emotion replication and experience, and conceptual understanding. More specifically, previously developed interventions often incorporated the following approaches: teaching participants to attend to relevant facial features and to associate the facial characteristics with an emotion (i.e., wide eyes and raised eyebrows=fear); to feel the emotion being expressed (i.e., mimicry); and learning to associate the emotion with conceptual information (i.e., associate emotions with past situations and feelings)(Bornhofen and McDonald 2008a, b; Radice-Neumann et al. 2009) In addition, most interventions employed learning strategies typically used to overcome attention and memory challenges when teaching people with TBI new information (i.e., errorless learning, vanishing cues, repetition and rehearsal, self-instruction, and gradual introduction to more complex stimuli and tasks) (Bornhofen and McDonald 2008a, b; Radice-Neumann et al. 2009). In general, participants significantly improved their ability to recognize emotions from facial expressions post-treatment. (Bornhofen and McDonald 2008a, b; Radice-Neumann et al. 2009) One of the studies compared the effectiveness of their facial affect recognition intervention to a control group trained on basic cognitive skills, such as attention, speed of processing, verbal and nonverbal memory, and executive functioning. Participants who received the facial affect recognition intervention were significantly better at recognizing facial expressions post-treatment than the control group (Neumann et al. 2013) . This suggests that although some aspects of cognition are related to affect recognition, treatment is more effective if it teaches elements that are more directly related to the skill.

One study compared two strategies to determine which was more beneficial at assisting participants with TBI to better recognize facial affect (McDonald et al. 2009). In this study, there were two conditions in which subjects had to identify emotions from facial expressions: one in which participants were instructed to focus their attention on relevant facial features, and the other in which participants were instructed to mimic the expressions. Participants with TBI did not improve their ability to recognize facial affect from either instruction. Findings from this study suggest that there is a gap between emotion identification and participants’ attention to relevant features or their mimicry of an emotion. In other words, just attending to relevant visual features or mimicking an emotional expression alone does not appear to be sufficient for improving facial affect recognition after TBI. There may be several reasons for this gap: there may still be a perceptual component involved (i.e., not processing facial expressions globally); or an element of emotional experience (i.e., disconnection between replication and experience); or a conceptual element in which participants lack the ability to associate emotional labels with relevant features and mimicry; or a combination of all of the above. Thus, it would be a premature misinterpretation of the findings to assume that training visual attention or mimicry should not be included as part of a treatment program; it just implies that participants with TBI need more assistance to bridge the gap between these elements and emotion identification. Due to design limitations, it is difficult to draw any further conclusions from this study.

Functional neuroimaging findings on facial affect processing There has been an extensive amount of research using functional magnetic resonance imaging (fMRI) to investigate facial affect processing since the mid-1990s. However, varying results across individual studies have made it difficult to derive meaning from these findings. In order to synthesize the results from the plethora of studies, two meta-analyses incorporated over 100 neuroimaging studies on facial affect processing (Fusar-Poli et al. 2009; Sabatinelli et al. 2011). Pooling the data from such a large number of studies allows us to begin to develop a better understanding of the areas of the brain that are most consistently and significantly activated during facial affect processing tasks. These meta-analyses are an excellent place to start to make sense out of the literature. However, because these meta-analyses combine the results from a variety of facial affect recognition tasks, we also take a closer look at a few individual studies that compare brain responses to tasks that are designed to engage more perceptual processing (ie. matching) versus tasks that are designed to engage more cognitive processing (ie. labeling). These distinctions are important since people with TBI have

Brain Imaging and Behavior

been found to have difficulty with both types of tasks; learning what mechanisms are critical for each could be helpful when considering neuroimaging studies for TBI. Finally, we briefly review neuroimaging results in other populations with facial affect recognition deficits. The type of fMRI studies and findings from other patient populations may be a useful guide for fMRI studies in people with TBI.

Meta-analyses This section summarizes the findings from two major fMRI meta-analyses that have been conducted on emotional face processing in healthy controls (Fusar-Poli et al. 2009; Sabatinelli et al. 2011). Fusar-Poli and colleagues conducted a PubMed search from 1990 to May 2008 and found 105 fMRI experiments that used emotional face processing paradigms in healthy participants that met their criteria, and used an activation likelihood estimation (ALE) meta-analytic procedure to identify the locations of consistent brain activations (voxel coordinates) elicited across studies (Turkeltaub et al. 2002). Sabatinelli et al. ran a PubMed search from 1995 to 2009 and pooled reports from 100 fMRI facial affect processing studies that met their criteria. They also used the ALE method to analyze these data. It should be noted that both meta-analyses included studies with a wide array of emotional paradigms that were not explicitly listed by the authors. Examples of some of the paradigms include passive viewing, perceptual matching, labeling, and target identification. Consequently, the results indicate areas of the brain that are involved in numerous types of emotion processing and incorporate neural structures responsible for various aspects of affect recognition under various conditions. Table 1 summarizes the major findings from both meta-analyses.

Fusar-Poli et al. (2009) contrasted brain activation to emotional faces with a baseline fixation point, and found that emotional faces elicited significantly more activation (pneutral faces

Amyg Cingulate Putamen Amyg

Insula MTG FG

Amyg Amygdala; FG fusiform gyrus; IFG inferior frontal gyrus; IOG inferior occipital gyrus; LG lingual gyrus; MFG middle frontal gyrus; MOG middle occipital gyrus; MPFC medial prefrontal cortex; MTG medial temporal gyrus; and SFG superior frontal gyrus

Brain Imaging and Behavior Table 2 Neural activation in response to different emotional expressions compared to neutral. Reconstructed from Fusor-Poli et al. findings (2009) Happy

Sad

Angry

Fearful

Disgust

Amyg (R,L) Ant. Cing (R) FG (L)

Amyg (R) LG (L)

Insula (L) IOG (R)

Amyg (R,L) FG (R) MFG (R)

Insula (R,L) Thalamus (R) FG (L)

Amyg Amygdala; Ant. Cing. anterior cingulate; FG fusiform gyrus; IOG inferior occipital gyrus; LG lingual gyrus; MFG middle frontal gyrus

and overlapping area(s) within the limbic system. The amygdala was particularly important for processing fearful, happy, and sad faces; however, it was strongest for fearful faces. The insula was specifically engaged when processing disgusted and angry faces, but activation was stronger in response to disgusted faces.

Functional neuroimaging: perceptual and cognitive tasks of facial affect recognition In this next section, we discuss several individual studies in order to consider how brain activations are affected by perceptual versus cognitive tasks of facial affect processing. More specifically, we will focus on an emotional matching task (perceptual) and an emotional labeling task (cognitive). Because it has been found that people with TBI have difficulty in both matching and labeling facial expressions, it is important to consider the neuroimaging results on these specific tasks in normal populations, as well as other populations who also have affect recognition deficits. Because the metaanalyses in the previous sections reported findings that were pooled across various facial affect processing tasks, they were not able to comment about different brain reactivity in response to the various task types. A common and widely-used example of a perceptual emotion matching task requires the participant to select a facial expression that matches a target face from two alternatives presented below the target facial expression (Hariri et al. 2000). Hariri and colleagues designed this task based on the assumption that participants would use the perceptual characteristics of the facial expressions (i.e., wide eyes and raised eyebrows) to match the target face, rather than actually interpreting or labeling the emotion. This assumption has been supported by many studies that show stronger activations in the neural structures associated with visual attention and processing of the emotional features when performing this particular task compared to those elicited when labeling the emotion (see below). A control task often has participants match geometric shapes instead of facial expressions. The other task of interest is a cognitive labeling task. A typical example of this would be to display a picture of a facial

expression, and to ask the participant to select between two emotional labels (e.g., angry or sad) which best describes the emotion being expressed. This task was designed to evaluate the mechanisms that underlie participants’ conceptual knowledge of the expressed emotion. Therefore, neuroimaging during this type of cognitive task is expected to indicate the neural structures associated with conceptual understanding of the emotion being expressed, as well as the structures associated with visual perceptual processing of the stimulus. Studies comparing the facial affect perceptual matching task to a control matching condition (geometric shape matching) have found that the affective matching condition elicited greater activation in bilateral amygdala (Drabant et al. 2009; Hariri et al. 2000; Hariri et al. 2002; Lieberman et al. 2007; Wang et al. 2004; Wright and Liu 2006), thalamus (Hariri et al. 2000), fusiform gyrus (Hariri et al. 2000; Hariri et al. 2002; Wang et al. 2004; Wright and Liu 2006), parahippocampal gyrus (Hariri et al. 2002), and prefrontal cortex (Hariri et al. 2002) than the shape task. Hariri et al. (2000) also compared activation differences for the affective matching task with the affective labeling task. Compared to the labeling condition, the matching task engaged greater bilateral activation in the amygdala. Conversely, the labeling condition elicited greater activation in the right prefrontal cortex compared to the matching task. Because the matching and labeling tasks are assumed to elicit perceptual and cognitive processing respectively, the interpretation of these findings has been that perceptual processing of emotions has a stronger relationship with limbic activation, whereas cognitive processing of emotions has a stronger association with the prefrontal cortex. While there is never certainty about the way in which individuals actually process a stimulus despite the intention of the task (i.e., match perceptual features, label the emotion, or a blend of the two), the areas of brain activation support the processing hypotheses. In other words, it is logical that labeling an emotion would elicit activation in the prefrontal cortex, an area typically correlated with conceptual knowledge and associations (Vandenberghe et al. 1996; Wagner et al. 1997). The fact that less activation was found in the prefrontal cortex during the perceptual matching task does not mean that participants did not cognitively interpret the expressed emotion; it simply suggests that this type of process is less predominant in the matching task than it is in the cognitive task. Several studies have examined the correlations between prefrontal cortical activations and amygdala activations during the perceptual matching and cognitive labeling tasks. For these tasks, activity in the amygdala was negatively associated with activity in the prefrontal cortex. For matching tasks, activity was greater in the amygdala and smaller in the prefrontal cortex. Conversely, for the labeling task activity was smaller in the amygdala, which was inversely associated with greater activity in the prefrontal cortex (Hariri et al. 2000).

Brain Imaging and Behavior

Similar results have been supported in several other studies (Fakra et al. 2008; Foland-Ross et al. 2010; Iidaka et al. 2001; Stein et al. 2007), and the latter findings have been interpreted to mean that cognitive labeling of emotions helps to regulate affect by attenuating the emotional limbic responses. In essence, these findings move beyond facial affect processing to emotion regulation, which has very pertinent implications for people with TBI who are often reported to have problems with emotion dysregulation (Alderman 2003; Eames and Wood 2003; Prigatano 1992). In summary, these findings suggest that the amygdala and limbic regions have the strongest activations during affective perceptual tasks. In contrast, the prefrontal cortex is more engaged during affective labeling tasks. In the normal population, the prefrontal cortex and limbic system appear to work in conjunction with one another, and the prefrontal cortex may help to minimize emotional responses associated with the amygdala. fMRI findings in non-TBI populations with affective disorders fMRI studies have also been used to investigate facial affect recognition in populations that characteristically have these deficits, such as schizophrenia and autism (Adolphs et al. 2001; Critchley et al. 2000b; Fakra et al. 2008; Hubl et al. 2003). This section highlights a few of these studies to demonstrate how fMRI has been used to elucidate underlying mechanisms associated with facial affect processing deficits in the non-TBI population. In one study, a facial affect matching task was used to compare differences in brain activations between participants with schizophrenia and healthy controls (Fakra et al. 2008). Compared to controls, participants with schizophrenia had significantly less amygdala activation and more prefrontal cortical activity. The authors interpreted the decreased limbic activity as an indication that participants with schizophrenia had a weaker emotional response to the facial expressions than controls. They suggested that participants with schizophrenia had more prefrontal activity because they were exerting more cognitive energy to complete the task than controls. Fakra and colleagues (2008) also investigated the relationship between activity in the prefrontal cortex with the amygdala during cognitive labeling tasks in participants with schizophrenia and normal controls. An inverse correlation was found between the amygdala and prefrontal cortical activity for the controls (with greater activity in the prefrontal cortex), but not for the participants with schizophrenia. The authors inferred from these findings that participants with schizophrenia had diminished emotion regulation through the prefrontal-limbic circuitry. Similar to findings in participants with schizophrenia, children with autism have also been found to have decreased

activation in the amygdala when processing facial affect compared to controls (Critchley et al. 2000a; Hubl et al. 2003). Other fMRI studies in children with autism point to abnormal activity in the fusiform gyrus and the superior parietal lobule. An fMRI study using an affective labeling task to study facial affect recognition in children with autism reported less activation in the fusiform gyrus compared to controls (Critchley et al. 2000a); consequently it was suggested that the children with autism were not globally processing the faces. This finding was replicated and expanded on in a study that compared brain activation in participants with autism and controls in response to face and object processing (Hubl et al. 2003). Similar to the Critchley et al. findings, participants with autism had less activation in the fusiform gyrus compared to controls when processing faces. In addition, they also found that participants with autism had more activation in the superior parietal lobule than controls; interestingly, this region had greater activation in the controls when they were processing objects. The authors identified the superior parietal lobule as a region critical for visual search tasks, suggesting it is activated when a person is scanning pieces of a stimulus feature by feature, rather than as a gestalt.

Integrating fMRI findings with elements of facial affect recognition It is clear that facial affect recognition is a complex process relying on distributed neural networks throughout the brain. In an earlier section of this paper, the various theoretical elements of facial affect recognition were discussed. Many of these elements have been tied to specific neuroanatomical structures that have been consistently activated in the fMRI studies on facial affect processing. The findings from the fMRI metaanalyses on facial affect processing serve as an anchor, demonstrating which areas of the brain are important in processing emotion from faces. In order to make these fMRI findings more meaningful, in the following section we discuss the relationship between brain structure and elements of function. Information is detailed in Fig. 1. The product of this section will then be used as a guiding reference for our discussion in the section on TBI rehabilitation. Perception The occipital cortex is responsible for visually processing facial expressions at a granular level (Adolphs 2002), whereas higher level visual processing is believed to occur within other regions of the brain (Haxby et al. 2000; Posamentier and Abdi 2003). For instance, global processing of the face, such that the expression is analyzed as a whole, has been associated with increased activity in the fusiform gyrus (Haxby et al. 2000; Tarr and Gauthier 2000). Conversely, the act of visual

Brain Imaging and Behavior Fig. 1 This figure portrays the relationship between the elements of facial affect recognition (function) with neuroanatomical structures consistently activated in fMRI studies of facial affect recognition. Overlapping regions depict structures that have been implicated in more than one of the functional elements of facial affect recognition. Because there are many rich neuronal connections amongst these structures, it is expected that each element influences the other, which is depicted with the bidirectional arrows

ELEMENTS OF FACIAL AFFECT RECOGNITION Perception (Basic visual processing, global percept, attention to emotionally salient features e.g., eyes)

ASSOCIATED BRAIN STRUCTURES

Occipital Cortex (IOG, MOG, LG) Fusiform Gyrus Amygdala

Emotion Replication and Experience (Real and simulated; change in body state and autonomic reaction, such as arousal and heart rate)

Conceptual Understanding (Previous experience and knowledge about an emotion)

scanning, which would be expected if one were analyzing features separately (separable processing), has been associated with the superior parietal lobule (Hubl et al. 2003). Separable processing of features has also been correlated with increased neural activity in the superior temporal sulcus and inferior temporal gyrus (Haxby et al. 2000; Hoffman and Haxby 2000; Posamentier and Abdi 2003). Finally, the amygdala has been associated with visual attention to emotionally salient features, especially the eye region of a face (Adolphs et al. 2005; Haxby et al. 2000; Posamentier and Abdi 2003). Although it is hard to discern at this point whether the amygdala is directly involved with visually attending to the eye region, or if it is indirectly involved via the emotional reaction to the eyes, there is still a correlation between the amygdala and attention to this facial feature. In general, the neuroimaging studies on facial affect processing have found common activations in the occipital cortex, fusiform gyrus, and the amygdala. Functionally, this suggests that processing of facial affect engages areas associated with basic visual processing (occipital cortex); global processing (fusiform gyrus); and attention to salient face features, specifically the eyes (amygdala). Neuroimaging studies on children with autism, a population that characteristically has emotion perception deficits, found reduced activation in the fusiform gyrus and the amygdala during facial affect processing tasks (Corbett et al. 2009; Critchley et al. 2000a; Schultz et al. 2003). Functionally, this suggests that they are not processing faces globally or attending to emotionally salient features to the degree that healthy control subjects were. Emotion replication and emotion experience Several areas of the brain have been associated with certain aspects of emotion replication, whether it was an actual replication or simulation. The basal ganglia, including the caudate

Basal Ganglia (Putamen) Cerebellum Inferior Parietal Lobule Thalamus Insula Cingulate Inferior Frontal Gyrus Medial Prefrontal Cortex

and lentiform nucleus (i.e., putamen and globus pallidus), premotor and motor cortex, and cerebellum (Chakravarthy et al. 2010; Doya 2000; Morecraft et al. 2004; Sato et al. 2004) are important for coordinating voluntary movement and may be involved in replicating emotional expressions (Carr et al. 2003; Leslie et al. 2004). In addition, studies have found evidence of neurons that are specifically activated when mirroring other people’s actions, including facial expressions (“mirror neurons”). These cells have been found in the following regions: the inferior frontal cortex, superior temporal gyrus, insula, amygdala, and rostral inferior parietal lobule (Carr et al. 2003; Iacoboni and Mazziotta 2007; Morecraft et al. 2004; Sato et al. 2004). It has been suggested that the thalamus, amygdala, insula, cingulate cortex, and prefrontal cortex are involved in the generation of autonomic (i.e., systemic arousal and heart rate) and body state responses associated with an emotional state (Phillips 2003). Emotional experience has been associated with the amygdala (Phillips 2003; Schneider et al. 1997), insula (Adolphs 2002; Damasio et al. 2000), somatosensory-related cortices (including supramarginal and angular gyri, which receive somatosensory input) (Adolphs et al. 2000; Damasio et al. 2000) and orbital frontal and ventromedial prefrontal cortices (Hornak et al. 2003; Hornak et al. 1996). The insular and somatosensory cortices are essential for one’s own visceral and somatic awareness of autonomic and body state changes, as well as reconstructing a simulated representation associated with an emotional state (Adolphs et al. 2000). Orbital frontal and ventromedial prefrontal cortices are also important for producing actual and simulated emotional responses. Damage to the orbital and ventromedial prefrontal cortices has been associated with dampened subjective emotional experiences (Hornak et al. 2003; Hornak et al. 1996). Additionally, the conceptual knowledge associated with this region is believed to be necessary for spurring simulated emotional responses (Adolphs 2002).

Brain Imaging and Behavior

The neuroanatomical structures typically activated during fMRI studies during facial affect processing that are consistent with those implicated in emotion replication and experience include the following regions: basal ganglia, cerebellum, insula, amygdala, cingulate, inferior parietal lobule, somatosensory cortices (supramarginal gyrus), inferior frontal gyrus, and ventromedial prefrontal cortices. Activation in these regions suggests that facial affect processing involves several aspects of emotional mimicking that may include actual autonomic changes and alterations in body state, or a simulated response. Studies show that focal damage to the amygdala, insula, somatosensory cortex, and prefrontal cortex often result in impaired facial affect recognition (Adolphs et al. 2000; Adolphs et al. 1999; Calder et al. 2000; Hornak et al. 1996). Conceptual understanding Several studies have demonstrated that the prefrontal cortex and inferior frontal gyrus are involved in retrieval of semantic memories (concept-based knowledge), including tasks that associate visual attributes with concepts (Vandenberghe et al. 1996; Wagner et al. 1997). The medial and inferior prefrontal cortices have rich connections with the temporal lobe, amygdala, insula, and somatosensory cortices (Damasio 1994). Consequently, this enables the emotional information that has been processed in these other regions of the brain to be integrated in the prefrontal cortex to help form emotional representations and associations. The prefrontal cortex and inferior frontal gyrus are commonly activated during fMRI tasks that require participants to label an emotional expression. This suggests participants are retrieving semantic knowledge that is conceptually associated with the emotional expression. As can be seen in Fig. 1, some of the neuroanatomical structures appear to be involved in multiple elements of facial affect recognition. For instance, it is suggested that the amygdala is involved in both Perception and Emotion Replication and Experience. Thus, it seems that there is some overlap in the neural circuits associated with each element of facial affect recognition. Additionally, because there are many feedforward and feedback neural connections between the various brain regions involved with the different elements of facial affect recognition, it is expected that each element influences the other to a certain degree. For instance, we know that there are neural connections between the structures associated with perception, emotion replication and experience, and conceptual understanding (Hariri et al. 2000). As such, it is logical that perception would influence emotion replication and experience as well as conceptualization of emotion; emotion replication and experience would influence perception and conceptualization of emotion; and conceptualization of emotion will drive perception and emotional replication and

experience. Consequently, each element of facial emotion perception does not exist in isolation from the others, and each has the power to influence the others.

Implications for studying and treating facial affect processing deficits after TBI We know from studies using behavioral assessments to evaluate emotion perception that people with TBI often have significant deficits with labeling and matching facial affect (Babbage et al. 2011). We also know that a few interventions exist that address most of the elements of facial affect recognition, and that these interventions have been found to be moderately effective at reducing impairments of emotion perception. So, what would be some benefits of using neuroimaging to study these deficits in the TBI population? We propose that the benefits would include advancements in knowledge, more detailed assessment, and refined treatment approaches. Improving knowledge One benefit would be to advance our knowledge of the brain and the mechanisms that underlie this dysfunction secondary to TBI in particular. Similar to the neuroimaging studies in participants with schizophrenia and autism, we could begin to identify the neurobiological mechanisms and the elements of facial affect recognition that are impaired after TBI. Because certain brain structures have been associated with certain functions, we can determine which elements of facial affect recognition are actually impaired after TBI - whether it is perception, emotion replication and experience, conceptual knowledge, or all of the above. In addition to examining activity in certain brain regions of interest, we could also use neuroimaging to investigate the functioning of different circuits or neural networks associated with facial affect recognition in participants with TBI, as has been done in other populations. We know that a moderate to severe TBI can cause damage to one or many regions of the brain, as well as cause diffuse axonal damage that can either weaken or sever the axonal connections amongst structures (“Centers for Disease Control and Prevention and last accessed July 2 2013,”). Because of the interconnected neural circuitry, damage to one area of the brain has the potential to affect functions associated with other areas of the brain. Consequently, if we look at Fig. 1 as a reference, we might assume that a patient with damage to the medial prefrontal cortex may not only have problems with the conceptual element of facial affect recognition, but also possibly have some level of impaired emotion replication and experience, as well as perception. Neuroimaging studies can evaluate functional connectivity to provide us with more insight about how the

Brain Imaging and Behavior

different networks typically involved with facial affect recognition are functioning after TBI. In other words, correlational analyses can provide insights about how brain structures are interacting with one another (i.e., prefrontal-limbic circuitry), so that we are not looking at isolated activity (Hariri et al. 2000). Another potential ‘knowledge’ benefit is the same as it is for many other neuroimaging studies –to investigate evidence of neuroplasticity and regeneration. Do facial affect recognition interventions alter underlying brain functioning in people with TBI, and does this correlate with functional behavioral changes? Many of these regions in the brain are also involved in other higher aspects of social cognition and emotion regulation (Adolphs 2001; Banks et al. 2007). If we see evidence of changes to these structures and networks, this might be an indicator that people have the necessary foundation to begin working on more advanced social skills and emotional deficits.

Improving treatment and outcomes Even though there are currently a handful of successful interventions, this does not mean that we should no longer strive to improve the effectiveness and efficiency of the existing interventions. It is possible that the information to be gained from neuroimaging will help to refine these treatments in a way that more specifically addresses the problem. This is important for several reasons. When working with people with TBI, they can easily get overwhelmed with too much information; therefore it is a necessity to make interventions as simple and streamlined as possible. Additionally, because insurance only covers a limited amount of therapy time/sessions, it is critical to make any intervention as efficient as possible. Thus, another benefit of neuroimaging would be to use the information to improve treatments for these deficits, and consequently the outcomes of patients with TBI.

Improving assessment With the development and optimization of sensitive and robust neuroimaging paradigms, it is possible that neuroimaging techniques could one day be used to identify deficits specific to a particular individual. Consequently, therapeutic interventions for affect recognition deficits could be tailored to individuals’ needs. Alternatively, neuroimaging studies could be used to link brain activation with behavioral task performance so that ultimately, more cost-effective and widely available neuropsychological testing methods could be developed to identify which aspects are contributing to patients’ facial affect recognition deficits.

Implications from past neuroimaging findings for treating facial affect deficits in TBI We know that facial affect recognition is widely distributed throughout the brain. Given the diffuse and heterogeneous damage that often occurs secondary to TBI, most facial affect recognition interventions have attempted to incorporate training to address all of the elements of facial affect recognition. There is, however, one aspect of facial affect recognition that we know to be important from the fMRI findings, which current facial affect recognition interventions do not seem to be addressing. From the description of the published interventions, it is difficult to determine if any of them have actually trained participants to process facial expressions as a gestalt (global percept). Although we do not know at this point if this is a problem for people with TBI, we do know it is an important aspect of processing facial affect given the reliable activation of the fusiform gyrus during most emotion perception tests. As such, this may be one aspect of facial affect recognition that future interventions may want to consider including in their training. Studying facial affect recognition deficits after TBI with fMRI We believe there is potential for fMRI to be used to better understand emotion perception deficits after TBI, similar to the way it has been used to understand facial affect recognition impairments in participants with autism and schizophrenia(Critchley et al. 2000a; Fakra et al. 2008). fMRI studies should be evaluating participants with TBI who actually have facial affect recognition impairments and those who do not have such impairments, relative to each other and healthy controls. We propose that neuroimaging should be used to investigate some of the following research questions. 1. Which areas of the brain show activation differences between participants with TBI and normal controls during a facial affect processing task? Can these differences in activation be used to derive inferences about the element(s) of facial affect processing that are disrupted due to TBI, either at the individual or group level? Regions of interest should include those listed in Table 1 and Fig. 1, including the amygdala, fusiform gyrus, and medial prefrontal cortex. Figure 1 can be used to identify which element of affect recognition is associated with areas that have significant activation differences. For instance, decreased activation in the fusiform gyrus would suggest participants are having difficulty processing faces globally; this would be further supported in the context of increased activation in the superior parietal lobule, as observed in participants with autism (Critchley et al. 2000b; Hubl et al. 2003). Differences in activation

Brain Imaging and Behavior

observed in the occipital cortex would indicate difficulties with more basic visuoperceptual processing. If the participants have differences in regions of the brain associated with motor and premotor activity, they may need training in learning how to mirror facial expressions. It is important to remember that some structures are implicated in multiple functions associated with facial affect recognition. For example, differences in amygdala response could suggest several different problems. If participants have decreased amygdala activation, it may mean they are having trouble with emotion replication and experience, or not attending to emotionally salient features, most likely the eyes. Additionally, the medial prefrontal cortex and inferior frontal gyrus are implicated in both Conceptual Understanding and Emotion Replication and Experience. This is why investigators are encouraged to also evaluate participants’ perceptual processing with eye-tracking and other behavioral measures, physiological responses to emotional stimuli (systemic arousal via galvanic skin response; heart rate; and respiration), and subjective assessments of emotional self-awareness. 2. Do participants with TBI have typical brain activations for some emotional facial expressions, but not others? If so, which emotions have atypical activations and in which part of the brain? Differences in the limbic region may mean that participants have more difficulty with replicating and experiencing some emotions, such as fear, compared to others. For instance, research has shown that people with depression often have more difficulty recognizing neutral and mildly happy facial expressions (Leppanen et al. 2004; Surguladze et al. 2005). 3. Studies should examine the functional connectivity between the medial prefrontal cortex and amygdala for matching and labeling tasks (Fig. 1), similar to studies in controls and those in schizophrenia (Fakra et al. 2008; Hariri et al. 2000). Functional correlations between these regions are believed to be an indicator of emotion regulation through a top-down process in which the prefrontal cortex controls emotional responses by reducing amygdala reactivity (Hariri et al. 2000).

Neuroimaging gaps and limitations for studying affect recognition deficits in TBI One of the limitations of neuroimaging is the limited understanding of how more or less activity in a brain region should be interpreted. For instance, we still do not know what it means that several studies found that people with schizophrenia had more prefrontal cortical activity than controls during perceptual matching tasks (Fakra et al. 2008). General assumptions can and have been made, but at this point, we cannot know the true meaning. It also complicates matters

that so many regions of the brain have been associated with multiple functions of facial affect recognition. For instance, the amygdala is associated with perception and attention to emotionally salient features, as well as emotion replication and experience. We may have to find behavioral and physiological tasks to correlate with these studies in order to help to make this distinction. However, for now, we would need to train both strategies instead of focusing on one. There are also neuroimaging challenges specific to TBI. The most obvious issue is the heterogeneity of the brain damage. While patients with TBI often have damage to the frontal lobes, temporal lobes, and limbic structures, the extent of damage in these areas may differ across subjects, causing greater within-group variability. In addition, it is common for persons with TBI to have problems with depression and anxiety (Morton and Wehman 1995). These emotional disorders and the medications for these symptoms may confound neural activation, especially in the amygdala. Until these complications can be addressed, neuroimaging studies on participants with TBI should be interpreted cautiously.

Conclusion People with TBI often have trouble processing emotional information from faces. This deficit is associated with inappropriate social interactions, behaviors, and emotion dysregulation, as well as poor psychosocial outcomes. It is important to understand the nature of this impairment for the many reasons described earlier in this review. A large number of neuroimaging studies have been conducted in non-TBI populations, which have helped to elucidate some of the mechanisms that contribute to facial affect processing. We hope that the information detailed in this paper can be a guiding reference for future studies wanting to use fMRI to study facial affect recognition deficits in the TBI population.

References Adolphs, R. (2001). The neurobiology of social cognition. Current Opinion in Neurobiology, 11(2), 231–239. Adolphs, R. (2002). Recognizing emotion from facial expressions: Psychological and neurological mechanisms. Behavioral and Cognitive Neuroscience Reviews, 1(1), 21. Adolphs, R. (2008). Fear, faces, and the human amygdala. Current opinion in neurobiology, 18(2), 166–172. Adolphs, R., Damasio, H., Tranel, D., Cooper, G., & Damasio, A. R. (2000). A role for somatosensory cortices in the visual recognition of emotion as revealed by three-dimensional lesion mapping. Journal of Neuroscience, 20(7), 2683–2690. Adolphs, R., Gosselin, F., Buchanan, T. W., Tranel, D., Schyns, P., & Damasio, A. R. (2005). A mechanism for impaired fear recognition after amygdala damage. Nature, 433(7021), 68–72.

Brain Imaging and Behavior Adolphs, R., Sears, L., & Piven, J. (2001). Abnormal processing of social information from faces in autism. Journal of Cognitive Neuroscience, 13(2), 232–240. Adolphs, R., Tranel, D., Hamann, S., Young, A. W., Calder, A. J., Phelps, E. A., & Damasio, A. R. (1999). Recognition of facial emotion in nine individuals with bilateral amygdala damage. Neuropsychologia, 37(10), 1111–1117. Alderman, N. (2003). Contemporary approaches to the management of irritability and aggression following traumatic brain injury. Neuropsychological Rehabilitation, 13(1), 211–240. Babbage, D. R., Yim, J., Zupan, B., Neumann, D., Tomita, M. R., & Willer, B. (2011). Meta-analysis of facial affect recognition difficulties after traumatic brain injury. Neuropsychology, 25(3), 277. Banks, S. J., Eddy, K. T., Angstadt, M., Nathan, P. J., & Phan, K. L. (2007). Amygdala–frontal connectivity during emotion regulation. Social cognitive and affective neuroscience, 2(4), 303–312. Bastiaansen, J., Thioux, M., & Keysers, C. (2009). Evidence for mirror systems in emotions. Philosophical Transactions of the Royal Society, B: Biological Sciences, 364(1528), 2391–2404. Bornhofen, C., & McDonald, S. (2008a). Comparing strategies for treating emotion perception deficits in traumatic brain injury. Journal of Head Trauma Rehabilitation, 23(2), 103–115. Bornhofen, C., & McDonald, S. (2008b). Treating deficits in emotion perception following traumatic brain injury. Neuropsychological Rehabilitation, 18(1) Calder, A. J., Keane, J., Manes, F., Antoun, N., & Young, A. W. (2000). Impaired recognition and experience of disgust following brain injury. Nature Neuroscience, 3(11), 1077–1078. Calder, A. J., & Young, A. W. (2005). Understanding the recognition of facial identity and facial expression. Nature Reviews Neuroscience, 6(8), 641–652. Carr, L., Iacoboni, M., Dubeau, M. C., Mazziotta, J. C., & Lenzi, G. L. (2003). Neural mechanisms of empathy in humans: a relay from neural systems for imitation to limbic areas. Proceedings of the National Academy of Sciences, 100(9), 5497. Centers for Disease Control and Prevention, last accessed July 2, 2013. (July 2, 2013). Retrieved from http://www.cdc.gov/ncipc/tbi/TBI. htm Chakravarthy, V., Joseph, D., & Bapi, R. S. (2010). What do the basal ganglia do? A modeling perspective. Biological cybernetics, 1–17. Cooper, C. L., Phillips, L. H., Johnston, M., Radlak, B., Hamilton, S., & McLeod, M. J. (2013). Links between emotion perception and social participation restriction following stroke. Brain Injury, 0, 1–5. Corbett, B. A., Carmean, V., Ravizza, S., Wendelken, C., Henry, M. L., Carter, C., & Rivera, S. M. (2009). A functional and structural study of emotion and face processing in children with autism. Psychiatry Research: Neuroimaging, 173(3), 196–205. Critchley, H. D., Daly, E., Phillips, M., Brammer, M., Bullmore, E. T., Williams, S., & Murphy, D. G. (2000a). Explicit and implicit neural mechanisms for processing of social information from facial expressions: a functional magnetic resonance imaging study. Human Brain Mapping, 9, 93–105. Critchley, H. D., Daly, E. M., Bullmore, E. T., Williams, S. C., Van Amelsvoort, T., Robertson, D. M., & Murphy, D. G. (2000b). The functional neuroanatomy of social behaviour: changes in cerebral blood flow when people with autistic disorder process facial expressions. Brain, 123(Pt 11), 2203–2212. Critchley, H. D., Rotshtein, P., Nagai, Y., O’Doherty, J., Mathias, C. J., & Dolan, R. J. (2005). Activity in the human brain predicting differential heart rate responses to emotional facial expressions. NeuroImage, 24(3), 751–762. Croker, V., & McDonald, S. (2005). Recognition of emotion from facial expression following traumatic brain injury. Brain Injury, 19(10), 787–799. Damasio, A. (1994). Descartes’ error: emotion, reason, and the human brain. New York: HarperCollins Publishers Inc.

Damasio, A. R., Grabowski, T. J., Bechara, A., Damasio, H., Ponto, L., Parvizi, J., & Hichwa, R. D. (2000). Subcortical and cortical brain activity during the feeling of self-generated emotions. Nature neuroscience, 3, 1049–1056. Decety, J. (1996). Do imagined and executed actions share the same neural substrate? Cognitive Brain Research, 3(2), 87–93. Delaveau, P., Salgado-Pineda, P., Witjas, T., Micallef-Roll, J., Fakra, E., Azulay, J. P., & Blin, O. (2009). Dopaminergic modulation of amygdala activity during emotion recognition in patients with Parkinson disease. Journal of clinical psychopharmacology, 29(6), 548. Dima, D., Stephan, K. E., Roiser, J. P., Friston, K. J., & Frangou, S. (2011). Effective connectivity during processing of facial affect: evidence for multiple parallel pathways. The Journal of Neuroscience, 31(40), 14378–14385. Doya, K. (2000). Complementary roles of basal ganglia and cerebellum in learning and motor control. Current opinion in neurobiology, 10(6), 732–739. Drabant, E. M., McRae, K., Manuck, S. B., Hariri, A. R., & Gross, J. J. (2009). Individual differences in typical reappraisal use predict amygdala and prefrontal responses. Biological Psychiatry, 65(5), 367–373. Draper, K., Ponsford, J., & Schönberger, M. (2007). Psychosocial and emotional outcomes 10 years following traumatic brain injury. The Journal of head trauma rehabilitation, 22(5), 278–287. Driscoll, D. M., Dal Monte, O., & Grafman, J. (2011). A need for improved training interventions for the remediation of impairments in social functioning following brain injury. Journal of neurotrauma, 28(2), 319–326. Eames, P. E., & Wood, R. L. (2003). Episodic disorders of behaviour and affect after acquired brain injury. Neuropsychological Rehabilitation, 13(1), 241–258. Fakra, E., Salgado-Pineda, P., Delaveau, P., Hariri, A. R., & Blin, O. (2008). Neural bases of different cognitive strategies for facial affect processing in schizophrenia. Schizophrenia Research, 100(1–3), 191–205. Foland-Ross, L. C., Altshuler, L. L., Bookheimer, S. Y., Lieberman, M. D., Townsend, J., Penfold, C., & Madsen, S. K. (2010). Amygdala reactivity in healthy adults is correlated with prefrontal cortical thickness. The Journal of neuroscience, 30(49), 16673–16678. Fusar-Poli, P., Placentino, A., Carletti, F., Landi, P., Allen, P., Surguladze, S., & Barale, F. (2009). Functional atlas of emotional faces processing: a voxel-based meta-analysis of 105 functional magnetic resonance imaging studies. Journal of Psychiatry & Neuroscience, 34(6), 418. Gainotti, G. (1993). Emotional and psychosocial problems after brain injury. Neuropsychological Rehabilitation, 3(3), 259–277. Green, R. E. A., Turner, G. R., & Thompson, W. F. (2004). Deficits in facial emotion perception in adults with recent traumatic brain injury. Neuropsychologia, 42(2), 133–141. Hariri, A. R., Bookheimer, S. Y., & Mazziotta, J. C. (2000). Modulating emotional responses: effects of a neocortical network on the limbic system. Neuroreport, 11(1), 43. Hariri, A. R., Tessitore, A., Mattay, V. S., Fera, F., & Weinberger, D. R. (2002). The amygdala response to emotional stimuli: a comparison of faces and scenes. NeuroImage, 17(1), 317–323. Haxby, J. V., Hoffman, E. A., & Gobbini, M. I. (2000). The distributed human neural system for face perception. Trends in Cognitive Science, 4(6), 223–233. Hoffman, E. A., & Haxby, J. V. (2000). Distinct representations of eye gaze and identity in the distributed human neural system for face perception. Nature, 3(1), 80–84. Hoofien, D., Gilboa, A., Vakil, E., & Donovick, P. J. (2001). Traumatic brain injury (TBI) 10–20 years later: a comprehensive outcome study of psychiatric symptomatology, cognitive abilities and psychosocial functioning. Brain Injury, 15(3), 189–209.

Brain Imaging and Behavior Hopkins, M. J., Dywan, J., & Segalowitz, S. J. (2002). Altered electrodermal response to facial expression after closed head injury. Brain Injury, 16(3), 245–257. Hornak, J., Bramham, J., Rolls, E. T., Morris, R. G., O’Doherty, J., Bullock, P. R., & Polkey, C. E. (2003). Changes in emotion after circumscribed surgical lesions of the orbitofrontal and cingulate cortices. Brain, 126(Pt 7), 1691–1712. Hornak, J., Rolls, E. T., & Wade, D. (1996). Face and voice expression identification in patients with emotional and behavioural changes following ventral frontal lobe damage. Neuropsychologia, 34(4), 247–261. Hubl, D., Bolte, S., Feineis-Matthews, S., Lanfermann, H., Federspiel, A., Strik, W., & Dierks, T. (2003). Functional imbalance of visual pathways indicates alternative face processing strategies in autism. Neurology, 61(9), 1232–1237. Iacoboni, M., & Mazziotta, J. C. (2007). Mirror neuron system: basic findings and clinical applications. Annals of Neurology, 62(3), 213–218. Ibarretxe-Bilbao, N., Junque, C., Tolosa, E., Marti, M. J., Valldeoriola, F., Bargallo, N., & Zarei, M. (2009). Neuroanatomical correlates of impaired decision‐making and facial emotion recognition in early Parkinson’s disease. European Journal of Neuroscience, 30(6), 1162–1171. Ietswaart, M., Milders, M., Crawford, J. R., Currie, D., & Scott, C. L. (2008). Longitudinal aspects of emotion recognition in patients with traumatic brain injury. Neuropsychologia, 46(148–159). Iidaka, T., Omori, M., Murata, T., Kosaka, H., Yonekura, Y., Okada, T., & Sadato, N. (2001). Neural interaction of the amygdala with the prefrontal and temporal cortices in the processing of facial expressions as revealed by fMRI. Journal of Cognitive Neuroscience, 13(8), 1035–1047. Jackson, H. F., & Moffat, N. J. (1987). Impaired emotional recognition following severe head injury. Cortex, 23, 293–300. Knox, L., & Douglas, J. (2009). Long-term ability to interpret facial expression after traumatic brain injuryand its relation to social integration. Brain and Cognition, 69, 442–449. Leppanen, J. M., Milders, M., Bell, J. S., Terriere, E., & Hietanen, J. K. (2004). Depression biases the recognition of emotionally neutral faces. Psychiatry Research, 128(2), 123–133. Leslie, K. R., Johnson-Frey, S. H., & Grafton, S. T. (2004). Functional imaging of face and hand imitation: towards a motor theory of empathy. NeuroImage, 21(2), 601–607. Lezak, M. D., & O’Brien, K. P. (1988). Longitudinal study of emotional, social, and physical changes after traumatic brain injury. Journal of Learning Disabilities, 21(8), 456. Lieberman, M. D., Eisenberger, N. I., Crockett, M. J., Tom, S. M., Pfeifer, J. H., & Way, B. M. (2007). Putting feelings into words. Psychological Science, 18(5), 421. McDonald, B. C., Flashman, L. A., & Saykin, A. J. (2002). Executive dysfunction following traumatic brain injury: neural substrates and treatment strategies. NeuroRehabilitation, 17(4), 333–344. McDonald, S., Bornhofen, C., & Hunt, C. (2009). Addressing deficits in emotion recognition after severe traumatic brain injury: The role of focused attention and mimicry. Neuropsychological rehabilitation, 19(3), 321–339. McDonald, S., & Flanagan, S. (2004). Social perception deficits after traumatic brain injury: Interaction between emotion recognition, mentalizing ability, and social communication. Neuropsychology, 18(3), 572–579. McDonald, S., Tate, R., Togher, L., Bornhofen, C., Long, E., Gertler, P., & Bowen, R. (2008). Social skills treatment for people with severe, chronic acquired brain injuries: a multicenter trial. Archives of physical medicine and rehabilitation, 89(9), 1648–1659. Milders, M., Fuchs, S., & Crawford, J. R. (2003). Neuropsychological impairments and changes in emotional and social behaviour following severe traumatic brain injury. Journal of Clinical and Experimental Neuropsychology, 25(2), 157–172.

Morecraft, R. J., Stilwell-Morecraft, K. S., & Rossing, W. R. (2004). The motor cortex and facial expression: new insights from neuroscience. The neurologist, 10(5), 235. Morton, M. V., & Wehman, P. (1995). Psychosocial and emotional sequelae of individuals with traumatic brain injury: a literature review and recommendations. Brain Injury, 9, 81–92. Mueser, K. T., Doonan, R., Penn, D. L., Blanchard, J. J., Bellack, A. S., Nishith, P., & DeLeon, J. (1996). Emotion recognition and social competence in chronic schizophrenia. Journal of Abnormal Psychology, 105(2), 271–275. Neumann, D., Babbage, D., Zupan, B., & Willer, B. (Under Review, November 2013). A randomized control trial of emotion recognition training after traumatic brain injury. Journal of Head Trauma Rehabilitation. Nowicki, S., & Mitchell, J. (1998). Accuracy in identifying affect in child and adult faces and voices and social competence in preschool children. Genetic, Social, and General Psychology Monographs, 124(1), 39–59. Oddy, M., Coughlan, T., Tyerman, A., & Jenkins, D. (1985). Socialadjustment after closed head-injury—a further follow-up 7 years after injury. Journal of Neurology, Neurosurgery and Psychiatry, 48(6), 564–568. Phillips, M. (2003). Understanding the neurobiology of emotion perception: implications for psychiatry. British Journal of Psychiatry, 182, 190–192. Phillips, M. L., Drevets, W. C., Rauch, S. L., & Lane, R. (2003). Neurobiology of emotion perception I: the neural basis of normal emotion perception. Biological Psychiatry, 54(5), 504–514. Piggot, J., Kwon, H., Mobbs, D., Blasey, C., Lotspeich, L., Menon, V., & Reiss, A. L. (2004). Emotional attribution in high-functioning individuals with autistic spectrum disorder: a functional imaging study. Journal of the American Academy of Child and Adolescent Psychiatry, 43(4), 473–480. Posamentier, M. T., & Abdi, H. (2003). Processing faces and facial expressions. Neuropsychology Review, 13(3), 113–143. Prigatano, G. P. (1992). Personality disturbances associated with traumatic brain injury. Journal of Consulting and Clinical Psychology, 60(3), 360–368. Radice-Neumann, D., Zupan, B., Babbage, D. R., & Willer, B. (2007). Overview of impaired facial affect recognition in persons with traumatic brain injury. Brain Injury, 21(8), 807–816. Radice-Neumann, D., Zupan, B., Tomita, M., & Willer, B. (2009). Training emotional processing in persons with brain injury. Journal of Head Trauma Rehabilitation, 24(5), 313–323. Ryan, N. P., Anderson, V., Godfrey, C., Eren, S., Rosema, S., Taylor, K., & Catroppa, C. (2013). Social communication mediates the relationship between emotion perception and externalizing behaviors in young adult survivors of pediatric traumatic brain injury (TBI). International Journal of Developmental Neuroscience, 31(8), 811– 819. Sabatinelli, D., Fortune, E. E., Li, Q., Siddiqui, A., Krafft, C., Oliver, W. T., & Jeffries, J. (2011). Emotional perception: meta-analyses of face and natural scene processing. NeuroImage, 54(3), 2524–2533. Sato, W., Kochiyama, T., Yoshikawa, S., Naito, E., & Matsumura, M. (2004). Enhanced neural activity in response to dynamic facial expressions of emotion: an fMRI study. Cognitive Brain Research, 20(1), 81–91. Scheid, R., Preul, C., Gruber, O., Wiggins, C., & von Cramon, D. Y. (2003). Diffuse axonal injury associated with chronic traumatic brain injury: evidence from T2*-weighted gradient-echo imaging at 3 T. American Journal of Neuroradiology, 24(6), 1049– 1056. Schneider, F., Grodd, W., Weiss, U., Klose, U., Mayer, K. R., Nägele, T., & Gur, R. C. (1997). Functional MRI reveals left amygdala activation during emotion. Psychiatry Research: Neuroimaging, 76(2–3), 75–82.

Brain Imaging and Behavior Schultz, R. T., Grelotti, D. J., Klin, A., Kleinman, J., Van der Gaag, C., Marois, R., & Skudlarski, P. (2003). The role of the fusiform face area in social cognition: implications for the pathobiology of autism. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 358(1430), 415–427. Singh, M., Jeong, J., Hwang, D., Sungkarat, W., & Gruen, P. (2010). Novel diffusion tensor imaging methodology to detect and quantify injured regions and affected brain pathways in traumatic brain injury. Magnetic resonance imaging, 28(1), 22–40. Spell, L. A., & Frank, E. (2000). Recognition of nonverbal communication of affect following traumatic brain injury. Journal of Nonverbal Behavior, 24(4), 285–300. Stein, J. L., Wiedholz, L. M., Bassett, D. S., Weinberger, D. R., Zink, C. F., Mattay, V. S., & Meyer-Lindenberg, A. (2007). A validated network of effective amygdala connectivity. NeuroImage, 36(3), 736–745. Surguladze, S., Brammer, M. J., Keedwell, P., Giampietro, V., Young, A. W., Travis, M. J., & Phillips, M. L. (2005). A differential pattern of neural response toward sad versus happy facial expressions in major depressive disorder. Biological psychiatry, 57(3), 201–209. Tarr, M. J., & Gauthier, I. (2000). FFA: a flexible fusiform area for subordinate-level visual processing automatized by expertise. Nature Neuroscience, 3(8), 764–769. Tessitore, A., Hariri, A. R., Fera, F., Smith, W. G., Chase, T. N., Hyde, T. M., & Mattay, V. S. (2002). Dopamine modulates the response of the human amygdala: a study in Parkinson's disease. The Journal of neuroscience, 22(20), 9099–9103.

Turkeltaub, P. E., Eden, G. F., Jones, K. M., & Zeffiro, T. A. (2002). Meta-analysis of the functional neuroanatomy of single-word reading: method and validation. NeuroImage, 16(3), 765–780. Vandenberghe, R., Price, C., Wise, R., Josephs, O., & Frackowiak, R. (1996). Functional anatomy of a common semantic system for words and pictures. Nature, 383(6597), 254–256. Wagner, A. D., Desmond, J. E., Demb, J. B., Glover, G. H., & Gabrieli, J. D. E. (1997). Semantic repetition priming for verbal and pictorial knowledge: a functional MRI study of left inferior prefrontal cortex. Journal of Cognitive Neuroscience, 9(6), 714–726. Wang, A. T., Dapretto, M., Hariri, A. R., Sigman, M., & Bookheimer, S. Y. (2004). Neural correlates of facial affect processing in children and adolescents with autism spectrum disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 43(4), 481–490. Winston, J. S., Henson, R., Fine-Goulden, M. R., & Dolan, R. J. (2004). fMRI-adaptation reveals dissociable neural representations of identity and expression in face perception. Journal of neurophysiology, 92(3), 1830–1839. Wright, P., & Liu, Y. (2006). Neutral faces activate the amygdala during identity matching. NeuroImage, 29(2), 628–636. Yim, J., Babbage, D. R., Zupan, B., Neumann, D., & Willer, B. (2013). The relationship between facial affect recognition and cognitive functioning after traumatic brain injury. Brain Injury, 27(10), 1155–1161. Zupan, B., & Neumann, D. (2013). Affect Recognition in Traumatic Brain injury: Responses to Unimodal and Multimodal Media. Journal of Head Trauma Rehabilitation doi:10.1097/HTR. 0b013e31829dded6.