Using the local heterogeneity of neural responses to index the integrity of representations and track recovery of function
-
1
Johns Hopkins University, Department of Cognitive Science, United States
Introduction: Little is known about changes in neural representations following post-stroke recovery. Recent work suggests that the local heterogeneity of neural responses may index the degree to which representations are more or less broadly distributed within a region (Jiang et al., 2013), with greater heterogeneity indicating more highly-tuned, compact representations. We introduce a novel technique – Local-Heterogeneity Regression (Hreg) Analysis - to quantify local neural heterogeneity. We apply this approach to acquired dysgraphia focusing on left ventral occipitotemporal cortex (vOTC) that has been associated with learning orthographic representations, testing the following hypotheses: 1) Pre-training local-Hreg values in vOTC will correlate with spelling accuracy; 2) Recovery of spelling will be indexed within vOTC by increases in heterogeneity.
Methods: Participants were 6 individuals with post-stroke, chronic dysgraphia. Individualized word sets were developed: 40 TRAINING words (25 - 80% letter accuracy) and 30 KNOWN words (100% letter accuracy). CART-based treatment (Beeson, 1999) was administered to TRAINING items. FMRI was carried out at pre- and post-training which included: spelling with KNOWN and TRAINING words and a non-spelling CONTROL.
Whole brain, Local-Hreg search-light analysis was performed. For each searchlight, it employed a general psychophysiological interaction analysis (gPPI; McLaren et al., 2012) that uses the neural response of the center voxel as the basis for comparison with surrounding voxels. Based on the similarity/dissimilarity of the responses, local-Hreg indexes voxel-to-voxel interactions within each searchlight: the lower the average interaction, the higher the local heterogeneity.
Hypothesis 1: Regions of interest (ROI) within left vOTC were identified via two whole-brain approaches applied to pre-training data: 1) univariate contrast of mean BOLD response for KNOWN > CONTROL (corrected 0.05); 2) local-Hreg of the TRAINING > KNOWN (p<0.1 uncorrected). Within these clusters, local-Hreg values for TRAINING > KNOWN were extracted for each participant, and were correlated with pre-treatment behavioral accuracy for TRAINING items.
Hypothesis 2: A whole-brain local-Hreg analysis was carried out comparing the neural responses to the TRAINING condition at pre-training versus post-training time-points.
Results: 1) The correlation between spelling accuracy and local-Hreg values for the vOTC cluster identified via the univariate approach was r=0.897 (p=0.016); for the cluster identified via the local-Hreg approach it was r=0.98 (p=0.0006) (see Figure 1 A, B). Both results reveal that Hreg values were strongly correlated with pre-training spelling performance, such that low heterogeneity within the left vOTC indexed low representational integrity.
2) In a whole-brain comparison of local-Hreg values at pre versus post training for the TRAINING items, a left anterior vOTC area (Peak MNI y= -46) was associated with increases in the heterogeneity (compactness and selectivity) of neural responses (see Figure 1 C).
Discussion: First, we established that the heterogeneity of orthographic neural representations prior to training is strongly associated with spelling accuracy. Second, we identified training-related increases in neural heterogeneity for trained items within left vOTC. This work provides a novel approach for tracking neural responses to training by quantifying changes in the neural heterogeneity and, therefore, can improve our understanding of how the damaged brain responds to treatment.
Acknowledgements
Acknowledgements: Jennifer Shea, Chloe Haviland, and the multi-site, NIDCD-supported project examining the neurobiology of language recovery in aphasia (DC006740).
References
Beeson, P. M., Hirsch, F. M., & Rewega, M. A. (2002). Successful single-word writing treatment: Experimental analyses of four cases. Aphasiology, 16(4), 473-491.
Jiang, X., Bollich, A., Cox, P., Hyder, E., James, J., Gowani, S. A., Riesenhuber, M. (2013). A quantitative link between face discrimination deficits and neuronal selectivity for faces in autism. NeuroImage : Clinical, 2, 320–331.
McLaren, D. G., Ries, M. L., Xu, G., & Johnson, S. C. (2012). A generalized form of context-dependent psychophysiological interactions (gPPI): a comparison to standard approaches. NeuroImage, 61(4), 1277–1286.
Keywords:
spelling,
orthography,
training,
dysgraphia,
fMRI
Conference:
54th Annual Academy of Aphasia Meeting, Llandudno, United Kingdom, 16 Oct - 18 Oct, 2016.
Presentation Type:
Platform Sessions
Topic:
Academy of Aphasia
Citation:
Purcell
J,
Wiley
R and
Rapp
B
(2016). Using the local heterogeneity of neural responses to index the integrity of representations and track recovery of function.
Front. Psychol.
Conference Abstract:
54th Annual Academy of Aphasia Meeting.
doi: 10.3389/conf.fpsyg.2016.68.00123
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:
01 May 2016;
Published Online:
15 Aug 2016.
*
Correspondence:
PhD. Jeremy Purcell, Johns Hopkins University, Department of Cognitive Science, Baltimore, MD, 21218, United States, purcel14@gmail.com