Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Subprocesses in speech planning: a voxel-based lesion-symptom mapping study of Apraxia of Speech
Poster Session C, Saturday, September 13, 11:00 am - 12:30 pm, Field House
Rahulkrishna G. Thimmugari1, Alycia B. Laks1,2, Peter E. Turkeltaub1,2, Andrew T. DeMarco1; 1Center for Brain Plasticity and Recovery, Georgetown University, 2Research Division, MedStar National Rehabilitation Hospital
Speech planning is typically characterized as an intermediate phase in the speech control network (Van Der Merwe, 2021). It involves the ventral premotor cortex (vPMC), with input from subcortical, frontal, and parietal regions. Computational models differ in how they depict the architecture of this network (Hickok et al., 2023; Miller & Guenther, 2021). Apraxia of Speech (AOS), a disorder affecting speech planning, provides a disease model for understanding system breakdown. Prior voxel-based lesion-symptom mapping (VLSM) work supports vPMC’s role in speech planning (Basilakos et al., 2015; Conterno et al., 2021). However, a key prediction of computational models and clinical observations remains untested: that AOS can be behaviorally and neurally disambiguated into phonetic and prosodic subtypes. This study investigates AOS subtypes by examining prevalence of speech error features (prosodic, phonetic) and associated neural correlates using VLSM. Participants included 87 left-hemisphere chronic stroke survivors (39 female/48 male; mean age 60.83; mean lesion volume 97.17cm²; mean chronicity 43.87mos). Participants underwent neuroimaging with high-resolution MPRAGE/FLAIR scans; lesions were manually traced. Behavioral testing included the ASRS-3 and WAB-R. Participants with AOS were subtyped as phonetic/prosodic/mixed based on a phonetic-to-prosodic feature ratio ((phonetic-prosodic)/(phonetic+prosodic)) derived from the ASRS-3. We interpreted ratios ≤-0.3 as prosodic; >-0.3 and ≤0.3 as mixed; and >0.3 as phonetic. VLSM (DeMarco & Turkeltaub, 2018) outcome variables included WAB-AQ; ASRS-3 Apraxia Severity, Aphasia Severity, Phonetic Features, and Prosodic Features; as well as the subtype ratio. Significance was set at p<0.005 and cluster-corrected at p<0.05, 10,000 permutations. 43.68% of our sample received an AOS diagnosis as indicated by ASRS-3. Of these participants, 47.37% showed Prosodic subtype; 28.95% showed Mixed subtype; and 18.42% showed Phonetic subtype. VLSM examining ASRS-3 Apraxia Severity revealed a significant cluster in ventral pre- and postcentral gyrus, also overlapping with white matter. VLSM examining ASRS-3 Aphasia Severity revealed a significant cluster encompassing superior temporal gyrus (STG) and posterior parietal cortex (PPC), while VLSM examining WAB-AQ revealed a significant, overlapping cluster in STG. VLSM examining ASRS-3 Phonetic Features with ASRS-3 Prosodic Features as covariates (and vice versa) revealed no significant clusters. VLSM examining subtype ratio with ASRS-3 Apraxia Severity as a covariate revealed only a subthreshold cluster in subcortical white matter, potentially implicating SLF III in phonetic AOS. The overall prevalence rate indicates that almost half of our sample has AOS, replicating prior work (Ziegler et al., 2022). We found strong behavioral support for distinct AOS subtypes, with rates of prosodic subtype double that of phonetic subtype. Our lesion results replicate prior VLSM work, showing that AOS is related to lesions in vPMC and the surrounding white matter, and aphasia to lesions in STG and PPC (Basilakos et al., 2015). We did not find lesion evidence supporting the neural disambiguation of phonetic and prosodic subtypes, i.e. that speech planning can be split into phonetic and prosodic subprocesses. However, the subthreshold cluster in subcortical white matter relating to subtype ratio points to future work using connectome-based lesion-symptom mapping.
Topic Areas: Speech Motor Control, Disorders: Acquired