Poster E45, Saturday, August 18, 3:00 – 4:45 pm, Room 2000AB
Neural Correlates of Impaired Emotional Prosody Recognition in Acute Right Hemisphere Stroke
Shannon M. Sheppard1, Lynsey M. Keator1, Bonnie L. Breining1, Kevin Kim1, Sadhvi Saxena1, Donna Tippett1,2,3, Amy Wright1, Argye E. Hillis1,2,3;1Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD 21287, 2Department of Physical Medicine and Rehabilitation, Johns Hopkins University School of Medicine, Baltimore, MD 21287, 3Department of Cognitive Science, Krieger School of Arts and Sciences, Johns Hopkins University, Baltimore, MD 21218
Fully comprehending language requires understanding semantics and syntactic structure and perceiving subtle prosodic cues. Damage to right hemisphere (RH) regions is associated with difficulty understanding emotional prosody – the expression of emotion in speech (Dara et al., 2014). Some studies have implicated damage to RH temporal regions (Ross & Monnot, 2008, Wright et al. 2018) in affective prosody recognition impairments. Yet, other studies, often with heterogenous participant pools, have not found a clear link (Breitenstein et al. 1998). Schirmer and Kotz (2006) proposed a RH ventral stream (sound to meaning) for processing emotional prosody recognition projecting from the superior temporal gyrus (STG) to the anterior superior temporal sulcus (STS). Evaluative judgments are subsequently made in frontal regions. This proposed stream is analogous to the LH ventral stream for speech processing. The aim of the current study was to investigate the relationship between recognition of emotion and lesion location in individuals with acute right hemisphere strokes using lesion mapping. Participants with RH damage following ischemic stroke were identified. The acute time point (within five days of stroke) included 20 participants (nine women) aged 23-85 years (M = 59.74; SD = 13.95). The chronic time point (six – twelve months post-stroke) included five participants (three women) (mean age at acute time point = 47.60; SD = 14.98). All participants received an acute MRI scan, including diffusion-weighted imaging (DWI). Areas of acute ischemia were identified and traced on DWI images. An emotion recognition test was given at both time points; participants listened to semantically neutral sentences spoken with emotional prosody and selected the corresponding emotion from six choices (happy, sad, angry, afraid, surprised, bored). DWI Trace images and lesion tracings were normalized, and proportion of damaged tissue in each of the parcels in the JHU atlas (Mori et al., 2008) were calculated for each participant. Associations between emotion recognition errors and the percentage of voxels damaged were identified. Z-scores were corrected for multiple comparisons; resulting statistical maps were thresholded using 2000 permutations (one-tailed p < 0.05). Acutely, emotion recognition was below that of controls (mean accuracy of 54.58% +/-SD 18.48% vs. 82.50% +/- SD 5.12%). Impaired emotional prosody recognition accuracy at the acute time point was associated with greater damage to the right posterior STG (Z = -2.69). Longitudinally, the three (of five) participants who demonstrated acute impaired emotional prosody recognition (> 1 SD below the mean from a group of healthy age-matched controls) all three showed improvement at the chronic time point. Only one improved to normal (83.3% accuracy). The lesion mapping results indicate the right posterior STG plays a critical role in the comprehension of emotional prosody, consistent with the RH ventral stream hypothesis of emotional prosody recognition. Longitudinal results demonstrate emotional prosody can improve from acute to chronic stages of recovery, but not all participants will achieve recovery to normal limits. Future research will further examine neural regions critical for specific components of emotional prosody, and the neural correlates and patterns of recovery in RH stroke.
Topic Area: Language Disorders