Event Abstract

Back to Event

Multivariate Lesion-Symptom Mapping of Spontaneous Speech: Evidence from Acute Stroke

Junhua Ding1, Randi Martin2, Cris Hamilton3 and Tatiana Schnur1*
  • 1 Department of Neurosurgery, Baylor College of Medicine, United States
  • 2 Department of Psychology, Rice University, United States
  • 3 Institutional Reporting, Research, and Information Systems, The University of Texas at Austin, United States

Introduction. We quantified language at the word and sentence level during spontaneous telling of the “Cinderella” story, using a detailed analysis (quantitative production analysis, QPA; Rochon, Saffran, Berndt, & Schwartz, 2000) and a data-driven approach to identify the structure of relations among variables (principal components analysis, PCA) . To identify the brain regions necessary for spontaneous speech, we used a multivariate lesion-symptom mapping (LSM) approach and tested an unbiased, relatively large subject sample during the acute/early subacute stage of a left-hemisphere stroke, avoiding the reorganization of brain function. Methods. Following Rochon et al., (2000), we calculated 13 QPA measures for inclusion in the PCA (see Figure) from 65 subjects (61 ± 14 years; 37 males) with first-time symptomatic left-hemisphere stroke (median 3 days from symptom onset). We quantified the extent of patients’ lesions manually on diffusion-weighted/T2 FLAIR imaging (CT = 5). To determine the association between language measures and lesion location, we performed support-vector regressions on voxels lesioned in at least 3 (5%) of 52 subjects, controlling for lesion volume (excluding 13 cases for either subarachnoid hemorrhage (n=3), unavailability of lesion demarcation (n=8), or behavioral outliers (n=2)). The model with the best prediction ability and voxels with the highest weights in the corresponding model were determined by 1000 permutation tests (p < 0.05). Results and Discussion. PCA revealed four principal factors (varimax rotation; eigenvalues > 1; see Figure) with some consistencies compared to previous characterizations of quantitative production in chronic stroke (Gordon, 2006). Sentence structure (Factor 1) had high loadings on variables such as sentence length and embedding index where poor performance was mainly associated with damage to the posterior temporal lobe (pMTG). Previous work examining picture description in chronic stroke found in contrast that a single measure (mean utterance length) was associated with the left inferior frontal lobe (IFG; Borovsky, Saygin, Bates, & Dronkers, 2007). Lexical selection (Factor 2) loaded highly on the proportions of verbs, pronouns and closed-class words produced, where lower and higher proportions were related to damage to the inferior frontal gyrus and pMTG, respectively. The lower proportion result is broadly consistent with chronic stroke deficits in producing a range of different words during picture description (Halai, Woollams, & Lambon Ralph, 2017). Grammatical accuracy (Factor 3) combined high loadings on the proportion of well-formed sentences and the correct use of determiners and was associated with damage to the IFG. This result is consistent with a similar analysis of connected speech in a neurodegenerative patient population (Wilson et al., 2010). Lastly, production fluency (Factor 4) included words per minute where lower fluency was associated with damage to widespread cortical and subcortical areas. Our result not only included a similar area found by Halai et al. (2017), but also another area (thalamus). In future work, we will use a longitudinal analysis in the same subject cohort to identify whether these acute brain-behavior associations change during recovery. This work will help understand the processes in spontaneous speech susceptible to brain-behavior reorganization.

Figure 1
Image

Acknowledgements

This work was supported by an R01DC014976 award to the Baylor College of Medicine from the National Institute on Deafness and Other Communication Disorders.

References

Borovsky, Arielle, Saygin, Ayse Pinar, Bates, Elizabeth, & Dronkers, Nina. (2007). Lesion correlates of conversational speech production deficits. Neuropsychologia, 45(11), 2525-2533. doi: https://doi.org/10.1016/j.neuropsychologia.2007.03.023 Gordon, Jean K. (2006). A quantitative production analysis of picture description. Aphasiology, 20(2-4), 188-204. doi: 10.1080/02687030500472777 Halai, Ajay D., Woollams, Anna M., & Lambon Ralph, Matthew A. (2017). Using principal component analysis to capture individual differences within a unified neuropsychological model of chronic post-stroke aphasia: Revealing the unique neural correlates of speech fluency, phonology and semantics. Cortex, 86, 275-289. doi: https://doi.org/10.1016/j.cortex.2016.04.016 Rochon, Elizabeth, Saffran, Eleanor M., Berndt, Rita Sloan, & Schwartz, Myrna F. (2000). Quantitative Analysis of Aphasic Sentence Production: Further Development and New Data. Brain and Language, 72(3), 193-218. doi: https://doi.org/10.1006/brln.1999.2285 Wilson, Stephen M., Henry, Maya L., Besbris, Max, Ogar, Jennifer M., Dronkers, Nina F., Jarrold, William, . . . Gorno-Tempini, Maria Luisa. (2010). Connected speech production in three variants of primary progressive aphasia. Brain, 133(7), 2069-2088. doi: 10.1093/brain/awq129

Keywords: spontaneous speech, acute stroke, multivariate lesion-symptom mapping, quantitative production analysis, principle component analysis

Conference: Academy of Aphasia 57th Annual Meeting, Macau, Macao, SAR China, 27 Oct - 29 Oct, 2019.

Presentation Type: Poster presentation

Topic: Not eligible for student award

Citation: Ding J, Martin R, Hamilton C and Schnur T (2019). Multivariate Lesion-Symptom Mapping of Spontaneous Speech: Evidence from Acute Stroke. Front. Hum. Neurosci. Conference Abstract: Academy of Aphasia 57th Annual Meeting. doi: 10.3389/conf.fnhum.2019.01.00105

Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.

The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.

Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.

For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.

Received: 03 May 2019; Published Online: 09 Oct 2019.

* Correspondence: Prof. Tatiana Schnur, Department of Neurosurgery, Baylor College of Medicine, Houston, Texas, 77030, United States, tatiana.schnur@bcm.edu