Elsevier

NeuroImage

Volume 30, Issue 2, 1 April 2006, Pages 580-587
NeuroImage

Cerebral pathways in processing of affective prosody: A dynamic causal modeling study

https://doi.org/10.1016/j.neuroimage.2005.09.059Get rights and content

Abstract

This study was conducted to investigate the connectivity architecture of neural structures involved in processing of emotional speech melody (prosody). 24 subjects underwent event-related functional magnetic resonance imaging (fMRI) while rating the emotional valence of either prosody or semantics of binaurally presented adjectives. Conventional analysis of fMRI data revealed activation within the right posterior middle temporal gyrus and bilateral inferior frontal cortex during evaluation of affective prosody and left temporal pole, orbitofrontal, and medial superior frontal cortex during judgment of affective semantics. Dynamic causal modeling (DCM) in combination with Bayes factors was used to compare competing neurophysiological models with different intrinsic connectivity structures and input regions within the network of brain regions underlying comprehension of affective prosody. Comparison on group level revealed superiority of a model in which the right temporal cortex serves as input region as compared to models in which one of the frontal areas is assumed to receive external inputs. Moreover, models with parallel information conductance from the right temporal cortex were superior to models in which the two frontal lobes accomplish serial processing steps. In conclusion, connectivity analysis supports the view that evaluation of affective prosody requires prior analysis of acoustic features within the temporal and that transfer of information from the temporal cortex to the frontal lobes occurs via parallel pathways.

Introduction

In spoken language, information about the emotional state of the speaker can be expressed via propositional cues at the verbal level and via non-verbal means of communication by modulation of the speech melody (affective prosody). Affective prosody is characterized by variations of suprasegmental language features, such as pitch, syllable duration, and voice quality (Banse and Scherer, 1996). Evidence obtained from lesion studies indicates a right-hemispheric superiority in processing of these features (Blonder et al., 1991, Bowers et al., 1987, Borod et al., 2002). In line with this suggestion, functional imaging studies have consistently demonstrated right-lateralized activations at the level of the auditory cortex during evaluation of affective prosody (Mitchell et al., 2003, Wildgruber et al., 2005). However, it has been shown that left-hemispheric lesions can also compromise comprehension of affective speech melody (Adolphs et al., 2002, Hornak et al., 1996, Hornak et al., 2003, Kucharska-Pietura et al., 2003, Pell, 1998, van Lancker and Sidtis, 1992) challenging the hypothesis that processing of affective prosody is exclusively subserved by the right hemisphere. Specifically, unilateral lesions within the inferior frontal cortex of both hemispheres as well as deep white matter lesions of the mid-rostral part of the corpus callosum can result in severe deficits in comprehension of affective prosody (Hornak et al., 1996, Hornak et al., 2003, Ross et al., 1997). Accordingly, functional neuroimaging and event-related electrophysiological studies on the neural correlates underlying the comprehension of affective prosody demonstrated bilateral activations in the inferior frontal lobe (Imaizumi et al., 1997, Pihan et al., 2000, Wildgruber et al., 2002, Wildgruber et al., 2004). These converging results from lesion and neuroimaging studies suggest that the frontal lobes of both hemispheres cooperate in decoding of non-verbal emotional information in the voice and that intact transcallosal communication of information is necessary for comprehension of affective prosody. However, neither lesion studies nor conventional analysis of functional imaging data can clarify whether this cooperation of both hemispheres is accomplished in a serial way via sequential processing steps or if both frontal lobes receive their information independently from each other from the right temporal cortex. It was the aim of this study to investigate the connectivity pattern subserving communication between the right temporal cortex and the frontal lobes during decoding of affective prosody. To this end, we used event-related functional magnetic resonance imaging (fMRI) combined with the novel technique of dynamic causal modeling (Friston et al., 2003). Dynamic causal modeling enables inferences on (1) the parameters representing influences of experimentally designed inputs, (2) the intrinsic coupling of different brain regions, and (3) how this coupling is modulated by an experimental factor. Given the lack of knowledge on the connectivity between neural areas implicated in processing of affective prosody, we precluded modulating factors and focused on the investigation of input regions and the intrinsic connectivity pattern within this network. Therefore, we compared models in which the right secondary auditory cortex serves as input region with models in which direct inputs are assumed to enter the network via one of the frontal areas. Furthermore, to investigate the architecture of the interregional connections, we compared dynamic causal models corresponding to serial and parallel processing within the frontal lobes.

Section snippets

Subjects

24 right-handed German native speakers (11 males, 13 females, mean age 24.4 years) with no history of neurological or psychiatric illness participated in an fMRI experiment. Handedness was determined using the Edinburgh Inventory (Oldfield, 1971). The Ethical Committee of the University of Tuebingen had approved the investigation. Informed consent was obtained according to the Declaration of Helsinki.

Stimuli

Six professional actors (3 females/3 males) pronounced 162 German adjectives in either happy,

Behavioral data

Mean group ratings of acoustic stimuli in the prestudy and mean group ratings during fMRI were strongly correlated for both emotional word content (r = 0.93) and affective prosody (r = 0.93) indicating that the participants of the fMRI experiment did comprehend verbal and non-verbal affective information in presence of scanner noise with sufficient accuracy (see Fig. 2).

Conventional fMRI analysis

To identify brain regions specifically contributing to the processing of affective prosody and emotional word content, blood

Discussion

In the present study, conventional analysis of fMRI data based on a general linear model was employed to identify brain regions underlying understanding of affective prosody and emotional word content. Subsequently, dynamic causal modeling (Friston et al., 2003) was used to investigate input regions and architecture of interregional connections within the network involved in comprehension of affective prosody.

Acknowledgments

This study was supported by the Deutsche Forschungsgemeinschaft (DFG grant WI 2101/1-1) and by the Junior Science Program of the Heidelberger Academy of Sciences and Humanities.

References (48)

  • R.C. Oldfield

    Assessment and analysis of handedness: the Edinburgh inventory

    Neuropsychologia

    (1971)
  • M. Pell

    Recognition of prosody following unilateral brain lesion: influence of functional and structural attributes of prosodic contours

    Neuropsychologia

    (1998)
  • W. Penny et al.

    Comparing dynamic causal models

    NeuroImage

    (2004)
  • D. Poeppel et al.

    Auditory lexical decision, categorial perception, and FM direction discrimination differentially engage left and right auditory cortex

    Neuropsychologia

    (2004)
  • E.D. Ross et al.

    Lateralization of affective prosody in brain and the callosal integration of hemispheric language functions

    Brain Lang.

    (1997)
  • S.K. Scott et al.

    The neuroanatomical and functional organization of speech perception

    Trends Neurosci.

    (2003)
  • S.K. Scott et al.

    Going beyond the information given: a neural system supporting semantic interpretation

    NeuroImage

    (2003)
  • N. Tzourio-Mazoyer et al.

    Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain

    NeuroImage

    (2002)
  • D. Wildgruber et al.

    Dynamic brain activation during processing of emotional intonation: influence of acoustic parameters, emotional valence, and sex

    NeuroImage

    (2002)
  • D. Wildgruber et al.

    Identification of emotional intonation evaluated by fMRI

    NeuroImage

    (2005)
  • R. Adolphs et al.

    Neural systems for recognition of emotional prosody: a 3-D lesion study

    Emotion

    (2002)
  • R. Banse et al.

    Acoustic profiles in vocal emotion expression

    J. Pers. Soc. Psychol.

    (1996)
  • L.X. Blonder et al.

    The role of the right hemisphere in emotional communication

    Brain

    (1991)
  • J.R. Binder et al.

    Neural correlates of sensory and decision processes in auditory object identification

    Nat. Neurosci.

    (2004)
  • Cited by (193)

    • Aberrant Emotional Prosody Circuitry Predicts Social Communication Impairments in Children With Autism

      2023, Biological Psychiatry: Cognitive Neuroscience and Neuroimaging
    • Brain mapping of emotional prosody in patients with drug-resistant temporal epilepsy: An indicator of plasticity

      2022, Cortex
      Citation Excerpt :

      The ability to decode prosody was intact also in the RTLE group. The neuroimaging results showed an essential role of the RSTG for emotional prosody perception, in line with previous findings from the literature (Alba-Ferrara et al., 2011, 2012; Ethofer et al., 2006, 2012; Mitchell & Ross, 2008; Ross et al., 1988; Wildgruber et al., 2006). Our results offer additional support to Ross and Mesulam's hypothesis (1979), as we found a predominant role of the right hemisphere in emotional prosody, which connects and interacts with the functional anatomical organization of propositional language in the left hemisphere.

    View all citing articles on Scopus
    View full text