Presentation

Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Poster Slams

Neural Correlates of Impaired Recognition of Emotion in Speech in Primary Progressive Aphasia (PPA)

Poster B51 in Poster Session B and Reception, Thursday, October 6, 6:30 - 8:30 pm EDT, Millennium Hall

Shannon Sheppard1,2, Jennifer Shea2,3, K. Alex Pacl2, Naydene Valencia1, Emilia Vitti2, Kristina Ruch2, Andreia Faria2, Argye Hillis2,3; 1Chapman University, 2Johns Hopkins University School of Medicine, 3Johns Hopkins University

Introduction Primary Progressive Aphasia (PPA) is a clinical syndrome caused by neurodegenerative disease. The most prominent clinical feature is decline in language skills in the earlier stages. Emerging research suggests some people with dementia experience aprosodia, which is impaired recognition/expression of prosodic features (e.g., pitch, volume, rate) used to convey emotions in speech. Based on our work in right hemisphere (RH) stroke we developed a three-stage model of receptive emotional prosody consisting of three processing stages (Stage 1: acoustic analysis; Stage 2: analyzing abstract representations of acoustic characteristics that convey emotion; Stage 3: semantic processing), which also interact with domain-general emotion recognition. Impaired emotional prosody in patients with dementia causes communication breakdowns with severe consequences including depression, and increased disruptive behaviors. However, few studies have investigated emotional prosody deficits in PPA. We aimed to: 1) to investigate receptive emotional prosody in PPA, 2) to examine the processes underlying receptive aprosodia in PPA, and 3) to investigate how atrophy contributed to specific impairments. Methods Participants: We enrolled 33 participants with PPA (mean age = 69.2 years). Procedure: MPRAGE scans were acquired in a subset of 23 patients. The whole brain was automatically segmented into regions of interest (ROIs). ROI volumes were calculated using MRICloud. Prosody Testing: Participants were given six behavioral assessments. Emotional prosody recognition was assessed by asking participants to choose the emotion of the speaker based on their tone of voice in producing 25 pseudoword sentences. Participants were also tested on each of the three stages of prosodic processing, as well as recognition of emotional facial expressions to assess presence of multi-modal emotion recognition deficits. Analysis: K-medoids analysis was used to identify different performance profiles by entering accuracy on each of the behavioral tests. The relationship between ROIs involved in receptive emotional prosody, and each stage of the three-stage receptive prosody model, was modelled using Least Absolute Shrinkage and Selection Operator (LASSO) regression for each behavioral test. Volumes of seven RH ROIs, and seven LH homologues, all normalized by cerebral volume, were entered into the LASSO models, along with the ratio of cerebral to intracranial volume (to control for inter-individual differences in brain size), and age. Results The k-medoids analysis identified three patient clusters. Cluster 2 had the most severely impaired emotion recognition in speech and facial expressions, and semantic processing of emotional words (stage 3). Cluster 1 also had impaired emotion recognition in speech and faces, but experienced stage 2 deficits. Most participants in Cluster 3 did not have impaired emotion recognition of speech or faces. LASSO analyses indicated that several LH and RH regions were important for emotion recognition, and each prosodic stage. Important regions for receptive prosody included right inferior frontal gyrus and basal ganglia. Discussion Many participants with PPA had impaired recognition of emotion in speech and faces. Impaired emotion recognition was characterized by different patterns of deficits of the three-stage model of receptive prosody. Although patients in clusters 1 and 2 both had impaired emotion recognition, they would likely benefit from different treatment approaches.

Topic Areas: Disorders: Acquired, Prosody