Differential sex-independent amygdala response to infant crying and laughing in parents versus nonparents.

BACKGROUND: Animal and human studies implicate forebrain neural circuits in maternal behavior. Here, we hypothesized that human brain response to emotional stimuli relevant for social interactions between infants and adults are modulated by sex- and experience-dependent factors. METHODS: We used functional magnetic resonance imaging and examined brain response to infant crying and laughing in mothers and fathers of young children and in women and men without children. RESULTS: Women but not men, independent of their parental status, showed neural deactivation in the anterior cingulate cortex, as indexed by decreased blood oxygenation level-dependent signal, in response to both infant crying and laughing. The response pattern changed fundamentally with parental experience: in the amygdala and interconnected limbic regions, parents (independent of sex) showed stronger activation from crying, whereas nonparents showed stronger activation from laughing. CONCLUSIONS: Our data show sex- and experience-dependent modulation of brain response to infant vocalizations. Successful recognition and evaluation of infant vocalizations can be critical for bonding mechanisms and for offspring well-being and survival. Thus, the modulation of responses by experience seems to represent an adaptive mechanism that can be related to reproductive fitness.     

Parsing neural mechanisms of social and physical risk identifications

Psychometric studies of risk perception have categorized personal risks into social and physical domains. To investigate whether and how the human brain differentiates social and physical risks, we scanned human adults using functional magnetic resonance imaging when they identified potential risks involved in social and physical behaviors. We found that the identification of risky behaviors in both domains induced increased activations in the anterior medial prefrontal cortex (MPFC, BA9/10)/ventral anterior cingulate (ACC) and posterior cingulate (PCC) relative to identification of safe behaviors. However, social risks induced stronger anterior MPFC activation whereas physical risks were associated with stronger ventral ACC activity. In addition, anterior MPFC activity was negatively correlated with the rating scores of the degree of social risk whereas PCC activity was positively correlated with the rating scores of the degree of physical risk. Relative to an autobiographical control task, the social risk identification task induced stronger sustained activity in the left supplementary motor area/dorsal ACC and increased transient activity in bilateral posterior insula. The physical risk identification task, however, resulted in stronger sustained activity in the right cuneus/precuneus and increased transient activation in bilateral amygdala. Our results indicate the existence of distinct neural mechanisms underlying social and physical risk identifications and provide neural bases for the psychometric categorization of risks into different domains.     

Audition of laughing and crying leads to right amygdala activation in a low-noise fMRI setting

Adequate behavioral responses to socially relevant stimuli are often impaired after lesions of the amygdala. These impaired behavioral responses in particular concern the recognition of facial, and sometimes vocal, expressions of fear. Using low-noise functional magnetic resonance imaging (fMRI) in combination with controlled sound delivery, we investigated how the amygdala, insula and auditory cortex are involved in the processing of affective non-verbal vocalizations (laughing, crying) in healthy humans. The same samples of male and female laughing and crying were presented in two different experimental conditions: self-induction of the corresponding emotions while listening, and detection of artificial pitch shifts in the same stimuli. Both conditions led to bilateral activation of the amygdala, insula and auditory cortex with a right-hemisphere advantage in the amygdala, and larger activation during laughing than crying in the auditory cortex with a slight right-hemisphere advantage for laughing, both likely due to acoustic stimulus features. The results show that amygdala activation by emotionally meaningful sounds like laughing and crying is independent of the emotional involvement, suggesting the pattern recognition aspect of these sounds is crucial for this activation. This aspect was revealed by a low-noise fMRI protocol which presumably minimized confounding effects of stressful high-noise fMRI.     

Brain activation during the perception of stressful word stimuli concerning interpersonal relationships in anorexia nervosa patients with high degrees of alexithymia in an fMRI paradigm

Several studies have reported that anorexia nervosa (AN) patients have high levels of alexithymia. However, relatively little is known about the underlying neurobiological relationships between alexithymia and AN. We used functional magnetic resonance imaging to examine the brain responses in 30 AN patients and 20 healthy women during the processing of negative words concerning interpersonal relationships. We investigated the relationship between alexithymia levels and brain activation in AN. AN patients showed significant activation of the orbitofrontal cortex, dorsolateral prefrontal cortex and medial prefrontal cortex while processing negative words concerning interpersonal relationships, as compared to the processing of neutral words. Moreover, the subjective rating of unpleasantness with negative words and neural activities in the amygdala, posterior cingulate cortex (PCC) and anterior cingulate cortex (ACC) negatively correlated with the level of alexithymia in AN. Our neuroimaging results suggest that AN patients tend to cognitively process negative words concerning interpersonal relationships, resulting in activation of the prefrontal cortex. Lower activation of the amygdala, PCC and ACC in response to these words may contribute to the impairments of emotional processing that are hallmarks of alexithymia. Functional abnormalities associated with alexithymia may be involved in the emotional processing impairments in AN patients.     

Context conditioning and extinction in humans: differential contribution of the hippocampus, amygdala and prefrontal cortex

Functional magnetic resonance imaging was used to investigate the role of the hippocampus, amygdala and medial prefrontal cortex (mPFC) in a contextual conditioning and extinction paradigm provoking anxiety. Twenty-one healthy persons participated in a differential context conditioning procedure with two different background colours as contexts. During acquisition increased activity to the conditioned stimulus (CS+) relative to the CS− was found in the left hippocampus and anterior cingulate cortex (ACC). The amygdala, insula and inferior frontal cortex were differentially active during late acquisition. Extinction was accompanied by enhanced activation to CS+ vs. CS− in the dorsal anterior cingulate cortex (dACC). The results are in accordance with animal studies and provide evidence for the important role of the hippocampus in contextual learning in humans. Connectivity analyses revealed correlated activity between the left posterior hippocampus and dACC (BA32) during early acquisition and the dACC, left posterior hippocampus and right amygdala during extinction. These data are consistent with theoretical models that propose an inhibitory effect of the mPFC on the amygdala. The interaction of the mPFC with the hippocampus may reflect the context-specificity of extinction learning.     

Left auditory cortex and amygdala, but right insula dominance for human laughing and crying

Evidence suggests that in animals their own species-specific communication sounds are processed predominantly in the left hemisphere. In contrast, processing linguistic aspects of human speech involves the left hemisphere, whereas processing some prosodic aspects of speech as well as other not yet well-defined attributes of human voices predominantly involves the right hemisphere. This leaves open the question of hemispheric processing of universal (species-specific) human vocalizations that are more directly comparable to animal vocalizations. The present functional magnetic resonance imaging study addresses this question. Twenty subjects listened to human laughing and crying presented either in an original or time-reversed version while performing a pitch-shift detection task to control attention. Time-reversed presentation of these sounds is a suitable auditory control because it does not change the overall spectral content. The auditory cortex, amygdala, and insula in the left hemisphere were more strongly activated by original than by time-reversed laughing and crying. Thus, similar to speech, these nonspeech vocalizations involve predominantly left-hemisphere auditory processing. Functional data suggest that this lateralization effect is more likely based on acoustical similarities between speech and laughing or crying than on similarities with respect to communicative functions. Both the original and time-reversed laughing and crying activated more strongly the right insula, which may be compatible with its assumed function in emotional self-awareness.     

Abnormal brain response during the auditory emotional processing in schizophrenic patients with chronic auditory hallucinations

OBJECTIVE: Few neuroimaging studies have been conducted regarding clinical associations between auditory hallucinations (AHs) and affective disturbances in patients with schizophrenia. This study aimed to elucidate the neurobiological basis of emotional disturbances in schizophrenic patients with persisting AHs. METHODS: Using functional magnetic resonance imaging (fMRI), the cortical responsiveness during the processing of laughing and crying sounds was measured and compared between 14 hallucinating schizophrenic patients, 14 nonhallucinating schizophrenic patients and 28 normal controls. RESULTS: The hallucinating patients showed differential neural activities in various areas including the amygdala, the hippocampus, the cingulate, the prefrontal cortex, and the parietal cortex, compared with the nonhallucinating patients and the normal controls. In particular, compared with the nonhallucinators, the hallucinators revealed reduced activation in the left amygdala and the bilateral hippocampus during the processing of crying sounds. CONCLUSION: Our findings suggest that the persistence of AHs in schizophrenia may induce functional disturbances of the emotion-related interconnected neural networks, including reduced responsiveness in the amygdala and hippocampus to negative stimuli.     

Anterior cingulate cortex modulates preparatory activation during certain anticipation of negative picture

We studied the neural activation associated with anticipations of emotional pictures using functional magnetic resonance imaging (fMRI) by directly comparing certain with uncertain anticipation conditions. While being scanned with fMRI, healthy participants (n=18) were cued to anticipate and then perceive emotional stimuli having predictable (i.e., certain) emotional valences (i.e., positive and negative), given a preceding cue, as well as cued stimuli of uncertain valence (positive or negative). During anticipation of pictures with certain negative valence, activities of supracallosal anterior cingulate cortex, ventrolateral prefrontal cortex, insula, and amygdala were enhanced relative activity levels that for the uncertain emotional anticipation condition. This result suggests that these brain regions are involved in anticipation of negative images, and that their activity levels may be enhanced by the certainty of anticipation. Furthermore, the supracallosal anterior cingulate cortex showed functional connectivity with the insula, prefrontal cortex, and occipital cortex during the certain negative anticipation. These findings are consistent with an interpretation that top-down modulation, arising from anterior brain regions, is engaged in certain negative anticipation within the occipital cortex. It is thought that the limbic system involving the amygdala, ACC, and insula, engaged emotional processes, and that the input system involving the visual cortex entered an idling state.     

Neural correlates of using distancing to regulate emotional responses to social situations

Cognitive reappraisal is a commonly used and highly adaptive strategy for emotion regulation that has been studied in healthy volunteers. Most studies to date have focused on forms of reappraisal that involve reinterpreting the meaning of stimuli and have intermixed social and non-social emotional stimuli. Here we examined the neural correlates of the regulation of negative emotion elicited by social situations using a less studied form of reappraisal known as distancing. Whole brain fMRI data were obtained as participants viewed aversive and neutral social scenes with instructions to either simply look at and respond naturally to the images or to downregulate their emotional responses by distancing. Three key findings were obtained accompanied with the reduced aversive response behaviorally. First, across both instruction types, aversive social images activated the amygdala. Second, across both image types, distancing activated the precuneus and posterior cingulate cortex (PCC), intraparietal sulci (IPS), and middle/superior temporal gyrus (M/STG). Third, when distancing one's self from aversive images, activity increased in dorsal anterior cingulate (dACC), medial prefrontal cortex (mPFC), lateral prefrontal cortex, precuneus and PCC, IPS, and M/STG, meanwhile, and decreased in the amygdala. These findings demonstrate that distancing from aversive social cues modulates amygdala activity via engagement of networks implicated in social perception, perspective-taking, and attentional allocation.     

Shared neural mechanisms for processing emotions in music and vocalizations

The neural mechanisms involved in the processing of vocalizations and music were compared, in order to observe possible similarities in the encoding of their emotional content. Positive and negative emotional vocalizations (e.g. laughing, crying) and violin musical stimuli digitally extracted from them were used as stimuli. They shared the melodic profile and main pitch/frequency characteristics. Participants listened to vocalizations or music while detecting rare auditory targets (bird tweeting, or piano's arpeggios). EEG was recorded from 128 sites. P2, N400 and Late positivity responses of ERPs were analysed. P2 peak was earlier in response to vocalizations, while P2 amplitude was larger to positive than negative stimuli. N400 was greater to negative than positive stimuli. LP was greater to vocalizations than music and to positive than negative stimuli. Source modelling using swLORETA suggested that, among N400 generators, the left middle temporal gyrus and the right uncus responded to both music and vocalizations, and more to negative than positive stimuli. The right parahippocampal region of the limbic lobe and the right cingulate cortex were active during music listening, while the left superior temporal cortex only responded to human vocalizations. Negative stimuli always activated the right middle temporal gyrus, whereas positively valenced stimuli always activated the inferior frontal cortex. The processing of emotional vocalizations and music seemed to involve common neural mechanisms. Notation obtained from acoustic signals showed how emotionally negative stimuli tended to be in Minor key, and positive stimuli in Major key, thus shedding some lights on the brain ability to understand music.     

Sex and performance level effects on brain activation during a verbal fluency task: A functional magnetic resonance imaging study

Neuroimaging studies investigating the neural correlates of verbal fluency (VF) focused on sex differences without taking into account behavioural variation. Nevertheless, group differences in this verbal ability might account for neurocognitive differences elicited between men and women. The aim of this study was to test sex and performance level effects and the combination of these on cerebral activation. Four samples of 11 healthy students (N = 44) selected on the basis of sex and contrasted VF scores, high fluency (HF) versus low fluency (LF), performed a covert phonological VF task during scans. Within- and between-group analyses were conducted. Consistent with previous studies, for each sample, the whole-group analysis reported activation in the inferior frontal gyrus (IFG), insula, anterior cingulate cortex (ACC), medial frontal gyrus (mFG), superior (SPL) and inferior parietal lobules (IPL), inferior visual areas, cerebellum, thalamus and basal ganglia. Between-group analyses showed an interaction between sexes and performances in the right precuneus, left ACC, right IFG and left dorsolateral prefrontal cortex (dlPFC). HF men showed more activation than LF ones in the right precuneus and left dlPFC. LF men showed more activation in the right IFG than HF ones and LF women elicited more activation in the left ACC than HF ones. A sex main effect was found regardless of performance in the left inferior temporal gyrus (ITG), cerebellum, anterior and posterior cingulate cortexes and in the right superior frontal gyrus (SFG) and dlPFC, lingual gyrus and ACC, with men eliciting significantly greater activations than women. A performance main effect was found for the left ACC and the left cerebellum regardless of sex. LF subjects had stronger activations than HF ones in the ACC whereas HF subjects showed stronger activations in the cerebellum. Activity in three discrete subregions of the ACC is related to sex, performance and their interaction, respectively. Our findings emphasize the need to consider sex and performance level in functional imaging studies of VF.     

Individual differences in amygdala and ventromedial prefrontal cortex activity are associated with evaluation speed and psychological well-being

Using functional magnetic resonance imaging, we examined whether individual differences in amygdala activation in response to negative relative to neutral information are related to differences in the speed with which such information is evaluated, the extent to which such differences are associated with medial prefrontal cortex function, and their relationship with measures of trait anxiety and psychological well-being (PWB). Results indicated that faster judgments of negative relative to neutral information were associated with increased left and right amygdala activation. In the prefrontal cortex, faster judgment time was associated with relative decreased activation in a cluster in the ventral anterior cingulate cortex (ACC, BA 24). Furthermore, people who were slower to evaluate negative versus neutral information reported higher PWB. Importantly, higher PWB was strongly associated with increased activation in the ventral ACC for negative relative to neutral information. Individual differences in trait anxiety did not predict variation in judgment time or in amygdala or ventral ACC activity. These findings suggest that people high in PWB effectively recruit the ventral ACC when confronted with potentially aversive stimuli, manifest reduced activity in subcortical regions such as the amygdala, and appraise such information as less salient as reflected in slower evaluative speed.     

Neural mechanisms of automatic and direct processing of phobogenic stimuli in specific phobia.

BACKGROUND: The study aimed to identify brain activation during direct and automatic processing of phobogenic stimuli in specific phobia. METHODS: Responses to phobia-related and neutral pictures (spiders and mushrooms) were measured by means of event-related functional magnetic resonance imaging during two different tasks. In the identification task, subjects were asked to identify the object (spider or mushroom). In a demanding distraction task, subjects had to match geometric figures displayed in the foreground of the pictures. RESULTS: Phobics showed greater responses to spiders versus mushrooms in the left amygdala, left insula, left anterior cingulate gyrus (ACC), and left dorsomedial prefrontal cortex (DMPFC) during the identification task and in the left and right amygdala during the distraction task. All of these activations were also significantly increased compared to control subjects who did not show stronger brain activation to spiders versus mushrooms under any task condition. CONCLUSIONS: Our findings propose specific neural correlates of automatic versus direct evaluation of phobia-relevant threat. While the amygdala, especially the right amygdala, seems to be crucially involved in automatic stimuli processing, activation of areas such as the insula, ACC and DMPFC is rather associated with direct threat evaluation and requires sufficient attentional resources.     

Localization of pain‐related brain activation: A meta‐analysis of neuroimaging data

A meta‐analysis of 140 neuroimaging studies was performed using the activation‐likelihood‐estimate (ALE) method to explore the location and extent of activation in the brain in response to noxious stimuli in healthy volunteers. The first analysis involved the creation of a likelihood map illustrating brain activation common across studies using noxious stimuli. The left thalamus, right anterior cingulate cortex (ACC), bilateral anterior insulae, and left dorsal posterior insula had the highest likelihood of being activated. The second analysis contrasted noxious cold with noxious heat stimulation and revealed higher likelihood of activation to noxious cold in the subgenual ACC and the amygdala. The third analysis assessed the implications of using either a warm stimulus or a resting baseline as the control condition to reveal activation attributed to noxious heat. Comparing noxious heat to warm stimulation led to peak ALE values that were restricted to cortical regions with known nociceptive input. The fourth analysis tested for a hemispheric dominance in pain processing and showed the importance of the right hemisphere, with the strongest ALE peaks and clusters found in the right insula and ACC. The fifth analysis compared noxious muscle with cutaneous stimuli and the former type was more likely to evoke activation in the posterior and anterior cingulate cortices, precuneus, dorsolateral prefrontal cortex, and cerebellum. In general, results indicate that some brain regions such as the thalamus, insula and ACC have a significant likelihood of activation regardless of the type of noxious stimuli, while other brain regions show a stimulus‐specific likelihood of being activated. Hum Brain Mapp, 2013. © 2011 Wiley Periodicals, Inc.     

Common and distinct networks underlying reward valence and processing stages: a meta-analysis of functional neuroimaging studies

To better understand the reward circuitry in human brain, we conducted activation likelihood estimation (ALE) and parametric voxel-based meta-analyses (PVM) on 142 neuroimaging studies that examined brain activation in reward-related tasks in healthy adults. We observed several core brain areas that participated in reward-related decision making, including the nucleus accumbens (NAcc), caudate, putamen, thalamus, orbitofrontal cortex (OFC), bilateral anterior insula, anterior cingulate cortex (ACC) and posterior cingulate cortex (PCC), as well as cognitive control regions in the inferior parietal lobule and prefrontal cortex (PFC). The NAcc was commonly activated by both positive and negative rewards across various stages of reward processing (e.g., anticipation, outcome, and evaluation). In addition, the medial OFC and PCC preferentially responded to positive rewards, whereas the ACC, bilateral anterior insula, and lateral PFC selectively responded to negative rewards. Reward anticipation activated the ACC, bilateral anterior insula, and brain stem, whereas reward outcome more significantly activated the NAcc, medial OFC, and amygdala. Neurobiological theories of reward-related decision making should therefore take distributed and interrelated representations of reward valuation and valence assessment into account.     

The anterior cingulate cortex and the phonatory control in monkey and man

Electrical stimulation of the anterior cingulate cortex yields vocalization in the monkey. The elicited vocalizations seem to represent primary stimulus responses. Monkeys are not able to perform a vocal conditioning task after ablation of the anterior cingulate cortex. However, they can carry out a lever-pressing conditioning task following destruction of this area. It is hypothesized that the anterior cingulate cortex exerts the volitional control of species-specific vocalizations in monkey. The non-verbal emotional vocal utterances are considered to be the human homologue of monkey's vocalizations. Therefore, bilateral lesion of the anterior cingulate cortex in man should hamper the volitional control of emotional vocal utterances in man as it does in monkeys. One personal observation is reported where after a bilateral infarction of the anterior cingulate cortex the patient's voice showed a permanent lack of emotional expression. The anterior cingulate cortex seems to play the decisive role in the volitional verbalization of emotions.     

Emotion triggers executive attention: Anterior cingulate cortex and amygdala responses to emotional words in a conflict task

Coherent behavior depends on attentional control that detects and resolves conflict between opposing actions. The current functional magnetic resonance imaging study tested the hypothesis that emotion triggers attentional control to speed up conflict processing in particularly salient situations. Therefore, we presented emotionally negative and neutral words in a version of the flanker task. In response to conflict, we found activation of the dorsal anterior cingulate cortex (ACC) and of the amygdala for emotional stimuli. When emotion and conflict coincided, a region in the ventral ACC was activated, which resulted in faster conflict processing in reaction times. Emotion also increased functional connectivity between the ventral ACC and activation of the dorsal ACC and the amygdala in conflict trials. These data suggest that the ventral ACC integrates emotion and conflict and prioritizes the processing of conflict in emotional trials. This adaptive mechanism ensures rapid detection and resolution of conflict in potentially threatening situations signaled by emotional stimuli. Hum Brain Mapp, 2011. © 2010 Wiley‐Liss, Inc.     

Effects of psychological stress on the cerebral processing of visceral stimuli in healthy women

Abstract  The aim of the study was to analyse effects of psychological stress on the neural processing of visceral stimuli in healthy women. The brain functional magnetic resonance imaging blood oxygen level-dependent response to non-painful and painful rectal distensions was recorded from 14 healthy women during acute psychological stress and a control condition. Acute stress was induced with a modified public speaking stress paradigm. State anxiety was assessed with the State-Trait-Anxiety Inventory; chronic stress was measured with the Perceived Stress Questionnaire. During non-painful distensions, activation was observed in the right posterior insular cortex (IC) and right S1. Painful stimuli revealed activation of the bilateral anterior IC, right S1, and right pregenual anterior cingulate cortex. Chronic stress score was correlated with activation of the bilateral amygdala, right posterior IC (post-IC), left periaqueductal grey (PAG), and right dorsal posterior cingulate gyrus (dPCC) during non-painful stimulation, and with activation of the right post-IC, right PAG, left thalamus (THA), and right dPCC during painful distensions. During acute stress, state anxiety was significantly higher and the acute stress – control contrast revealed activation of the right dPCC, left THA and right S1 during painful stimulation. This is the first study to demonstrate effects of acute stress on cerebral activation patterns during visceral pain in healthy women. Together with our finding that chronic stress was correlated wit the neural response to visceral stimuli, these results provide a framework for further studies addressing the role of chronic stress and emotional disturbances in the pathophysiology of visceral hyperalgesia.     

Local glutamate in cingulate cortex subregions differentially correlates with affective network activations during face perception

The cingulate cortex is involved in emotion recognition/perception and regulation. Rostral and caudal subregions belong to different brain networks with distinct roles in affective perception. Despite recent accounts of the relevance of cingulate cortex glutamate (Glu) on blood‐oxygen‐level‐dependent (BOLD) responses, the specificity of the subregional Glu levels during emotional tasks remains unclear. Seventy‐two healthy participants (age = 27.33 ± 6.67, 32 women) performed an affective face‐matching task and underwent magnetic resonance spectroscopy (MRS) at 7 Tesla. Correlations between the BOLD response during emotion perception and Glu concentration in the pregenual anterior cingulate cortex (pgACC) and anterior midcingulate cortex (aMCC) were compared on a whole‐brain level. Post hoc specificity of the association with an affect was assessed. Lower Glu in the pgACC correlated with stronger activation differences between negative and positive faces in the left inferior and superior frontal gyrus (L IFG and L SFG). In contrast, lower Glu in the aMCC correlated with BOLD contrasts in the posterior cingulate cortex (PCC). Furthermore, negative face detection was associated with prolonged response time (RT). Our results demonstrate a subregion‐specific involvement of cingulate cortex Glu in interindividual differences during viewing of affective facial expressions. Glu levels in the pgACC were correlated with frontal area brain activations, whereas Glu in the salience network component aMCC modulated responses in the PCC–precuneus. We show that region‐specific metabolite mapping enables specific activation of different BOLD signals in the brain underlying emotional perception.     

Neural correlates of adolescents’ viewing of parents’ and peers’ emotions: Associations with risk-taking behavior and risky peer affiliations

Social reorientation from parents to same-age peers is normative in adolescence, but the neural correlates of youths’ socioemotional processing of parents and peers have not been explored. In the current study, 22 adolescents (average age 16.98) underwent neuroimaging (functional magnetic resonance imaging) while viewing and rating emotions shown in brief video clips featuring themselves, their parents, or an unfamiliar peer. Viewing self vs. other and parents vs. the peer activated regions in the medial prefrontal cortex, replicating prior findings that this area responds to self-relevant stimuli, including familiar and not just similar others. Viewing the peer compared with parents elicited activation in posterior ‘mentalizing’ structures, the precuneus, posterior cingulate cortex (PCC), bilateral posterior superior temporal sulcus and right temporoparietal junction, as well as the ventral striatum and bilateral amygdala and hippocampus. Relative activations in the PCC and precuneus to the peer vs. the parent were related both to reported risk-taking behavior and to affiliations with more risk-taking peers. The results suggest neural correlates of the adolescent social reorientation toward peers and away from parents that may be associated with adolescents’ real-life risk-taking behaviors and social relationships.