**1.2 Measurement of human brain activity by NIRS**

In the field of brain science, fMRI and positron emission tomography (PET) have mainly been used to measure human brain activity since the 1990s. Coupled with the development of cognitive neuroscience, brain imaging techniques using NIRS have spread rapidly since the early 2000s.

When neural activity occurs in the human brain, regional cerebral blood flow increases in specific brain regions associated with the performance of a particular task. Therefore, hemoglobin concentration of blood increases in the regions. Capturing the change of hemoglobin concentrations enables us to identify the active regions in the brain. NIRS noninvasively monitors the hemodynamic change mediated by the change in hemoglobin concentration (advantages and disadvantages of NIRS in comparison with other devices are described in Section 3).

The NIRS device emits a near-infrared light from the surface of the head through optical fiber and detects the scattered and reflected light in the brain. The light of the near-infrared range, from 700 to 1,000 nm, has relatively high permeability in living tissue. The measurement uses the different absorbance characteristics between oxygenated hemoglobin (oxy-Hb) and deoxygenated hemoglobin (deoxy-Hb). Because the light path's length from the emitting position to the detecting position cannot be measured, absolute concentration changes of hemoglobin in the brain tissue cannot be determined. The relative concentration changes in oxy-Hb, deoxy-Hb, and total-Hb are calculated by using three wavelengths in the current NIRS devices according to the modified Beer-Lambert law (e.g., Hoshi & Tamura, 1993; Villinger & Chance, 1997).

Previous studies using brain imaging techniques of multi-channel NIRS clarified the neural mechanisms involved in various cognitive functions, such as motion perception of the human body (Shimada, Hiraki, Matsuda, & Oda, 2004), language processing (Herrmann, Ehlis, & Fallgatter, 2003; Noguchi, Takeuchi, & Sakai, 2000), emotion (Suzuki, Gyoba, & Sakuta, 2005), Stroop effect (Schroeter, Zysset, Kruggel, & Yves von Cramon, 2003), and Go-Nogo task (Herrmann, Plichta, Ehlis, & Fallgatter, 2005).

#### **1.3 Neuroscience perspectives on empathy**

#### **1.3.1 Functional components of empathy**

"Empathy" is understanding another person's internal state, including their thoughts and feelings, imaging the viewpoint of the other, and responding with compassion to the other's distress (e.g., Decety & Ickes, 2009; Preston & de Waal, 2002). The psychological construct of empathy that motivates prosocial behaviors is regulated by both a basic emotional contagion system (affective components) and a more advanced cognitive perspective taking system (cognitive components).

Emotion contagion occurs by the influence of others' emotions automatically without selfawareness (Hatfield, Rapson, & Le, 2009). On the other hand, cognitive perspective taking is the ability to understand the other's thoughts and feelings by imaging his or her viewpoint. In order to understand the other intentionally and consciously, cognitive empathy may modulate and control emotions depending on executive resources (i.e., higher cognitive functions including working memory, attention control, and memory retrieval).

In the field of brain science, fMRI and positron emission tomography (PET) have mainly been used to measure human brain activity since the 1990s. Coupled with the development of cognitive neuroscience, brain imaging techniques using NIRS have spread rapidly since

When neural activity occurs in the human brain, regional cerebral blood flow increases in specific brain regions associated with the performance of a particular task. Therefore, hemoglobin concentration of blood increases in the regions. Capturing the change of hemoglobin concentrations enables us to identify the active regions in the brain. NIRS noninvasively monitors the hemodynamic change mediated by the change in hemoglobin concentration (advantages and disadvantages of NIRS in comparison with other devices are

The NIRS device emits a near-infrared light from the surface of the head through optical fiber and detects the scattered and reflected light in the brain. The light of the near-infrared range, from 700 to 1,000 nm, has relatively high permeability in living tissue. The measurement uses the different absorbance characteristics between oxygenated hemoglobin (oxy-Hb) and deoxygenated hemoglobin (deoxy-Hb). Because the light path's length from the emitting position to the detecting position cannot be measured, absolute concentration changes of hemoglobin in the brain tissue cannot be determined. The relative concentration changes in oxy-Hb, deoxy-Hb, and total-Hb are calculated by using three wavelengths in the current NIRS devices according to the modified Beer-Lambert law (e.g., Hoshi & Tamura,

Previous studies using brain imaging techniques of multi-channel NIRS clarified the neural mechanisms involved in various cognitive functions, such as motion perception of the human body (Shimada, Hiraki, Matsuda, & Oda, 2004), language processing (Herrmann, Ehlis, & Fallgatter, 2003; Noguchi, Takeuchi, & Sakai, 2000), emotion (Suzuki, Gyoba, & Sakuta, 2005), Stroop effect (Schroeter, Zysset, Kruggel, & Yves von Cramon, 2003), and Go-

"Empathy" is understanding another person's internal state, including their thoughts and feelings, imaging the viewpoint of the other, and responding with compassion to the other's distress (e.g., Decety & Ickes, 2009; Preston & de Waal, 2002). The psychological construct of empathy that motivates prosocial behaviors is regulated by both a basic emotional contagion system (affective components) and a more advanced cognitive perspective taking

Emotion contagion occurs by the influence of others' emotions automatically without selfawareness (Hatfield, Rapson, & Le, 2009). On the other hand, cognitive perspective taking is the ability to understand the other's thoughts and feelings by imaging his or her viewpoint. In order to understand the other intentionally and consciously, cognitive empathy may modulate and control emotions depending on executive resources (i.e., higher cognitive

functions including working memory, attention control, and memory retrieval).

**1.2 Measurement of human brain activity by NIRS** 

the early 2000s.

described in Section 3).

1993; Villinger & Chance, 1997).

system (cognitive components).

Nogo task (Herrmann, Plichta, Ehlis, & Fallgatter, 2005).

**1.3 Neuroscience perspectives on empathy 1.3.1 Functional components of empathy** 

Decety (2006) proposed a neuroscientific model corresponding to the conceptual model of empathy. This model consists of four major functional components: shared representation between the self and the other, mental flexibility to take the other's perspective, selfawareness, and emotion regulation. It is assumed that these four components dynamically interact to produce empathy.

The four components are related to brain functions according to Decety's model. Shared representation is related to fronto-parietal networks based on the shared circuits between perception-action, and self-awareness is related to the inferior parietal lobule and the anterior insula on the right side. Mental flexibility is related to the prefrontal cortex. Emotion regulation is involved in the interaction between prefrontal and anterior cingulated systems and subcortical emotion-generation systems.

## **1.3.2 Empathy and ventrolateral prefrontal cortex**

According to the model of empathy (Decety, 2006) described above, it is hypothesized that the interaction between bottom-up processing and top-down processing produces empathy. Bottom-up processing begins by an input of perceived data (information from the outside world) and interprets the perceived data under the influence of the physical characteristics of the stimuli. When we meet an other person, we are resonant to the movements and emotions of that person by the perceptual input automatically and unconsciously. In other words, emotions are contagious, and this processing proceeds in a bottom-up fashion automatically. On the other hand, top-down processing is influenced by the context of the present situation and by knowledge from past experience that individuals use as stimuli. Consequently, in order to understand others from the viewpoint of the others, intentional and conscious mental efforts are required. These cognitive empathic processes modulate emotion regulation and perspective taking, depending on the executive functions for higher controlled processing of working memory, attention control, and memory retrieval.

Such higher cognitive-controlled functions are involved in the prefrontal cortex, including the ventrolateral prefrontal cortex (VLPFC). The right VLPFC is well known as a critical region for general inhibition and for regulating affective responses. The VLPFC also modulates the activity of the amygdala, which plays a key role in emotional appraisal and is related to detections of fear expression and eye-direction (Adolphs, Tranel, Damasio, & Damasio, 1994, 1995; Wicker, Michel, Henaff, & Decety, 1998). Therefore, the VLPFC is a critical area for the processing of emotional regulation via cortical-subcortical pathways (Batson, Early, & Salvarini, 1997).

#### **1.3.3 "Pain network" which feels the pain of others**

Previous studies on perspective taking have been reported in the brain imaging studies that deal with empathy for the physical pain of others (Decety & Grezes, 2006; Singer, Seymour, O'Doherty, Kaube, Dolan, & Frith 2004; Singer, Seymour, O'Doherty, Stephan, Dolan, & Frith, 2006). These studies found the existence of a "*pain network*" including the anterior cingulate cortex (ACC) and insula, which are involved in understanding the pain of others like one's own pain.

Singer et al. (2004) used fMRI to compare brain activities between two conditions: when participants felt pain in oneself, and when participants observed that their beloved partner,

Using NIRS to Investigate Social Relationship in Empathic Process 71

the empathic process and investigated the individual differences in perspective taking and the neural responses evoked by the inhibition of emotions; this was done by analyzing the

**2. A NIRS study on perspective taking associated with social relationships by** 

Following the paradigm of Singer et al. (2006), after playing the Prisoner's Dilemma game to form a good or a bad impression of the opponents, the participants observed the opponents' facial expressions (happy, neutral, and angry) and evaluated the valence of each facial expression on a 7-point scale ranging from "pleasantness" to "unpleasantness." The participants observed the facial expressions under two perspective-taking conditions: selfperspective vs. other-perspective. The brain activity was measured by using NIRS while the participants were evaluating the facial expressions. The participants were divided into two groups according to the points of the Interpersonal Reactivity Index (IRI) that they completed after all experiments. The IRI measures the components of empathy, from which we took particular note of perspective taking (e.g., 'I sometimes try to understand my

The prediction was that a higher unpleasant emotion would be produced in the selfperspective condition rather than in the other-perspective condition during observation of an unfair opponent, since negative emotions should arise automatically when the participants observe the facial expression of the unfair opponent. On the other hand, in the other-perspective condition with the unfair opponent, the participants should rate the facial expression from the viewpoint of the unfair opponent while inhibiting the negative emotions for the disliked opponent, especially for a happy expression. Therefore, it was expected that the activation would increase in the right VLPFC in the other-perspective

Moreover, the right VLPFC might play an important role for individuals who have a high ability of perspective taking. The right VLPFC would likely have a greater impact on taking the other-perspective, since top-down control from the prefrontal cortex functions very well for individuals with a high ability of perspective taking than for individuals with a low

Thirty-seven healthy volunteers (18 females; mean age ± SD: 19.5 ± 3.4) were divided into two groups according to the perspective taking scale of a Japanese version of the IRI (Davis, 1983): 19 participants with a high perspective-taking ability (mean score 24.79 ± 1.76) and 18

Pictures of happy and anger facial expressions were made by a morphing technique, based on digitized grayscale images of 6 Japanese faces (3 men and 3 women) showing a neutral

participants with a low perspective-taking ability (mean score 18.83 ± 1.89).

friends better by imagining how things look from their perspective.').

activation of VLPFC.

**Nomura et al. (2010)** 

**2.1 Purpose** 

condition.

**2.2 Method** 

**2.2.2 Stimuli** 

**2.2.1 Participants** 

ability of perspective taking.

who came to the laboratory with them, felt pain. In addition, the subjective empathic abilities of the participants were measured by questionnaires. The results show that the activated brain areas in common between the self condition (i.e., the participants themselves feel a pain) and the other condition (i.e., the participants observe their partners feeling a pain) were the bilateral anterior insula (AI), rostral anterior cingulate cortex (ACC), brainstem, and cerebellum. The activation levels of AI and ACC were significantly correlated positively with the empathic ability of individuals. These findings indicate that AI and ACC form the neural basis of understanding the emotions of one's own and others' pain and that the areas are related to emotion processing to evoke the empathic response to the pain of others.

Furthermore, Singer et al. (2006) reported evidence that the empathic response to the pain of others is affected by the social relationship with the other. Namely, a person shows strong empathy for the pain of a favorite person, whereas a person does not show empathy for the pain of non-favorite persons. More interestingly, the results indicate that males appear to feel pleasure in the pain of the non-favorite others.

In the experimental paradigm presented by Singer et al. (2006), participants played the Prisoner's Dilemma game in order to form a good or a bad impression of the opponent players (confederates) before measuring the brain activity by fMRI. In the game, one of two opponents made a cooperative and fair play toward the participants, whereas the other opponent made an uncooperative and unfair play. As a result, the participants came to like the fair opponent but came to dislike the unfair opponent. After the game, the brain activity was measured while the participants were observing the opponent receive a pain to the hand by electrical stimulation.

The results showed that activation in the pain network encompassing the AI and ACC was observed for fair opponents in both male and female participants. However, this activation was significantly reduced in males for pain given to the unfair opponents. At the same time, it was reported that the activity of the nucleus accumbens, known as a reward-related area, increased depending on the degree that the participant strongly desired revenge. This means that for males, the pain felt by the opponents who show unfair behavior brings them satisfaction in their revenge.

Nevertheless, even if they are disliked or unknown others, we can give them a helping hand when we encounter a situation in which the disliked or unknown other feels pain or distress, and we feel the need to help them from an ethical viewpoint. In order to clarify the neural mechanisms underlying such prosocial behaviors that suppress our own feelings, we must examine how the individual difference in perspective-taking ability influences not only the social relationships with the others but also influences the ability to evaluate the emotional states based on the social relationship.

Previous studies have demonstrated that emotion regulation processing is involved in the right VLPFC in relation to regulating or suppressing negative emotions caused by pain (Lieberman, Eisenberger, Crockett, Tom, Pfeifer, & Way, 2007; Wager, Davidson, Hughes, Lindquist, & Ochsner, 2008). In addition, it has also been reported that both self- and otherperspective taking are related to the activation of the pain network (Jackson, Brunet, Meltzoff, & Decety, 2006), and the activation of these areas increases depending on the degree of subjective empathic abilities as measured by questionnaires (Singer et al., 2004, 2006). Therefore, Nomura et al. (2010) focused on the mechanism of perspective taking in the empathic process and investigated the individual differences in perspective taking and the neural responses evoked by the inhibition of emotions; this was done by analyzing the activation of VLPFC.
