Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Feature-driven characterization of motor speech disorders in nonfluent variant primary progressive aphasia
Poster Session C, Thursday, October 1, 10:45 am - 12:45 pm, Wangari Maathai
This poster is part of the Sandbox Series.
G. Lynn Kurteff1, Willa Keegan-Rodewald1, Zoe Ezzes1,2,3, Lisa D. Wauters1,4, Maria Luisa Mandelli1, Diana A. Rodriguez1, Ryan L. Newbury1, Zachary A. Miller1, Howard J. Rosen1, William W. Seeley1, Maria Luisa Gorno-Tempini1; 1University of California, San Francisco, 2San Diego State University, 3University of California, San Diego, 4The University of Texas at Austin
INTRODUCTION – The international criteria for nonfluent variant primary progressive aphasia (nfvPPA) state that the most prominent clinical feature at symptom onset and initial phases must be difficulty with speech/language in the domains of agrammatism and/or apraxia of speech (AoS). However, the phenotype of specific motor speech impairments is highly variable across the nfvPPA spectrum of disorders. This heterogeneity is complicated by debate surrounding AoS (and to a lesser extent, dysarthria and dysphonia): prosodic and phonetic variants of AoS have been proposed in nfvPPA as well as a separate clinical syndrome called primary progressive AoS, but agreement among neurologists and speech-language pathologists concerning these distinctions remains unsteady. Additionally, many individuals with nfvPPA exhibit phenotypes of motor speech impairment that fall outside the conventional characterizations of AOS and dysarthria, which were originally described in stroke. This project aims to characterize data-driven phenotypes of neurodegenerative motor speech impairment using an agnostic, feature-based approach. METHODS – Fifty motor speech features spanning the range of motor speech impairments observed in nfvPPA will be retrospectively rated as absent or present in participants from the UCSF Fein Memory and Aging Center historical cohort. Trained speech-language pathologists will systematically rate motor speech features using recordings of a standardized motor speech evaluation battery. Tasks include passage reading, picture description, mono- and multisyllabic word repetition, diadochokinetic rate (e.g., “papapa”, “pataka”), and sustaining /a/. A multidimensional item response theory model will be constructed to reduce the dimensionality of the rated speech features. This framework condenses the original 50-item feature space into a smaller set of latent traits. Within-subject estimations of each latent trait will next be regressed against subject-specific neuroimaging deviations in cortical thickness to construct maps of atrophy associated with each trait. PRELIMINARY RESULTS – Features have been rated in a preliminary cohort of 30 patients. The most commonly observed features were: segmental sound errors, speech sound distortions, impaired sequential motion rates (i.e., “pataka”), increase in speech errors with utterance length/syllabic complexity, and reduced rate of speech. The preliminary results demonstrate that the putative diagnostic labels of AoS and dysarthria do not cleanly map onto patterns of motor speech impairment in nfvPPA. Most participants showed a mixture of impairments that could be attributed to either AoS or dysarthria (severity collinearity of r = 0.31), with many also exhibiting dysphonia. Data-driven feature grouping showed stronger inter-feature correlations than conventional a priori classifications (e.g. phonetic AoS, prosodic AoS), which did not map cleanly onto distinct patterns of atrophy. FUTURE DIRECTIONS – Our preliminary results demonstrate that the stroke-centric diagnostic labels of dysarthria and AoS are not ideal phenotypes of motor speech impairment in neurodegeneration. Motor speech impairment in nfvPPA appears to map less onto the weakness/coordination dichotomy posited by dysarthria/AoS but instead patterns to specific speech effectors (e.g., laryngeal function, respiration, oral articulators) and to distinct neurobiological hubs (e.g., anterior insula, precentral gyrus, supplementary motor area). We plan to present results from a larger retrospective cohort (estimated n of 150) for this Sandbox Series.
Topic Areas: Speech Motor Control, Disorders: Acquired