Event Abstract

Back to Event

Structural fragmentation of linguistic brain networks predicts aphasia severity, but not response to treatment.

Alexander M. Swiderski1, 2*, Haley Dresang2, William Hula1, 2, Michael W. Dickey1, 2, Fang-Cheng Yeh2, Juan Fernandez-Miranda3 and Patrick J. Doyle1
  • 1 VA Pittsburgh Healthcare System, United States
  • 2 University of Pittsburgh, United States
  • 3 Stanford University, United States

Introduction. Network measures [1]–[3] can characterize the complex relationship between brain regions and their connectivity. Research has shown that (dis)organization of language-related neural networks predicts both aphasia severity and treatment response. For example, individuals with more severe neural-network disorganization post-stroke have more severe aphasia than individuals with relatively preserved network organization [4] and measures of temporal-lobe network integration predict patient response to phonological/semantic cueing treatment [5]. Semantic Feature Analysis (SFA; [6]), one of the most well-studied aphasia treatments, improves both naming and overall aphasia severity [7]–[9]. However, it remains unknown how residual neural networks of stroke survivors support response to this treatment. The purpose of this study was to replicate the finding that global network integration predicts aphasia severity [4]-[5] and assess whether temporal-lobe network organization is predictive of response to SFA. Methods. Eighteen participants with aphasia secondary to left-hemisphere stroke > 6 MPO participated. The Comprehensive Aphasia Test (CAT[10]) was administered before treatment to estimate aphasia severity. Diffusion spectrum images were collected with a Siemens 3T Tim Trio Scanner using a 2D EPI diffusion sequence and were reconstructed by q-space diffeomorphic reconstruction [11]. A connectivity matrix was then generated from language specific ROIs (Figure 1) to estimate normalized small worldness (NSW[2]) and average left-temporal betweenness centrality (BC[1]). Estimates of NSW and BC were binary, denoting the presence or absence of connections, but not information regarding connection strengths [3]. Treatment: Participants received intensive SFA treatment (3-3.5 hours/day, 4-5days/week, for 4 weeks) as part of a clinical trial at the VA Pittsburgh Healthcare System. Analysis: CAT Modality Mean T-score served as the aphasia severity outcome variable. The difference in empirical logit-transformed proportion-correct scores between post- and pre-treatment of treated items was the SFA treatment-response outcome variable. Bayesian linear regression and Bayes Factors (BF) were used to analyze the relationship between network measures and behavioral measures. Each regression model was validated with leave one out (LOO) cross-validation [12]. Results. Fit statistics and LOO estimates for each model were excellent (Table 1). NSW was a moderately strong predictor of aphasia severity (r = .48, 95%HDI: 0.01,0.95), with a BF of .99 not favoring the naïve model over the predictor model. The relationships between SFA treatment response and NSW (r = 0.31, 95%HDI: -0.21, 0.84) and left-temporal lobe BC (r = -0.17, 95%HDI: -0.70, 0.37) were not reliably different from zero. Discussion. Community structure of preserved neural networks was predictive of overall aphasia severity [4]-[5], [8]. However, the lack of relationship between the network measures and response to treatment is not consistent with prevailing findings that global and peri-Svlvian network architecture are critical in the promotion of response to treatment for aphasia [5],[8]. A potential cause for these unexpected findings may be related to the variable connection densities of ROIs [3]. Further investigations of response to SFA varying as a function of network measures may benefit from controlling for the size of ROIs and examining ROI-specific estimates of BC.

Figure 1
Image
Figure 2
Image

References

[1] S. Kintali, “Betweenness Centrality : Algorithms and Lower Bounds,” ArXiv08091906 Cs, Sep. 2008. [2] D. S. Bassett and E. T. Bullmore, “Small-World Brain Networks Revisited,” The Neuroscientist, vol. 23, no. 5, pp. 499–516, Oct. 2017. [3] E. Bullmore and O. Sporns, Complex brain networks: Graph theoretical analysis of structural and functional systems, vol. 10. 2009. [4] B. K. Marebwa, J. Fridriksson, G. Yourganov, L. Feenaughty, C. Rorden, and L. Bonilha, “Chronic post-stroke aphasia severity is determined by fragmentation of residual white matter networks,” Sci. Rep., vol. 7, no. 1, p. 8188, Aug. 2017. [5] L. Bonilha, E. Gleichgerrcht, T. Nesland, C. Rorden, and J. Fridriksson, “Success of Anomia Treatment in Aphasia Is Associated With Preserved Architecture of Global and Left Temporal Lobe Structural Networks,” Neurorehabil. Neural Repair, p. 1545968315593808, Jul. 2015. [6] M. Boyle and C. A. Coelho, “Application of Semantic Feature Analysis as a Treatment for Aphasic Dysnomia,” Am. J. Speech Lang. Pathol., vol. 4, no. 4, p. 94, Nov. 1995. [7] E. A. Efstratiadou, I. Papathanasiou, R. Holland, A. Archonti, and K. Hilari, “A Systematic Review of Semantic Feature Analysis Therapy Studies for Aphasia,” J. Speech Lang. Hear. Res. JSLHR, vol. 61, no. 5, pp. 1261–1278, 17 2018. [8] W. D. Hula et al., “Left Ventral Stream White Matter Connectivity Predicts Response to Semantic Feature Analysis Treatment in Chronic Aphasia,” Front. Hum. Neurosci., 2017. [9] S. J. Oh et al., “Treatment Efficacy of Semantic Feature Analyses for Persons with Aphasia: Evidence from Meta-Analyses,” Commun. Sci. Disord., vol. 21, no. 2, pp. 310–323, Jun. 2016. [10] K. Swinburn, G. Porter, and D. Howard, Comprehensive aphasia test. Hove, UK: Psychology Press, 2004. [11] F. Yeh, V. J. Wedeen, and W. I. Tseng, “Generalized q-Sampling Imaging,” IEEE Trans. Med. Imaging, vol. 29, no. 9, pp. 1626–1635, Sep. 2010. [12] A. Vehtari, A. Gelman, and J. Gabry, “Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC,” Jul. 2015.

Keywords: Diffusion Spectrum Imaging (DSI), Aphasia, semantic feature analysis (SFA), Bayesian inference, graph theory

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

Presentation Type: Poster presentation

Topic: Eligible for student award

Citation: Swiderski AM, Dresang H, Hula W, Dickey MW, Yeh F, Fernandez-Miranda J and Doyle PJ (2019). Structural fragmentation of linguistic brain networks predicts aphasia severity, but not response to treatment.. Front. Hum. Neurosci. Conference Abstract: Academy of Aphasia 57th Annual Meeting. doi: 10.3389/conf.fnhum.2019.01.00116

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: 07 May 2019; Published Online: 09 Oct 2019.

* Correspondence: Mr. Alexander M Swiderski, VA Pittsburgh Healthcare System, Pittsburgh, Pennsylvania, 15240, United States, Aswiderski@pitt.edu