Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Word-Form learning and re-learning: Shared and/or unique neurocognitive processes?
Poster Session F, Friday, October 2, 2:45 - 4:45 pm, Wangari Maathai
Brenda Rapp1, Yuan Tao1, Sartaj Singh2, Jennifer Shea1, Robert Wiley3; 1Johns Hopkins University, 2University of South Carolina, 3University of North Carolina, Greensboro
INTRODUCTION The neural mechanisms that support the “relearning” of word-forms following neural injury are poorly understood. Two key questions: To what extent does word re-learning in the injured brain share the mechanisms of new word learning in the intact brain? Also, are there neurocognitive mechanisms that are unique to re-learning? The significant methodological challenge is to distinguish between shared and unique mechanisms in re-learning. To address this, we analyzed BOLD and behavioral data from two studies: one involving new word-form learning in older adults, the other involving behavioral treatment of word forms in post-stroke aphasia. We examined if the brain areas in which BOLD responsiveness was associated with word learning or relearning are shared or unique in these two groups. METHODS New word-form learning. 23 healthy older adults (mean age = 67.1) were tasked with learning 36 novel word spellings over ~6 weeks. Word-form re-learning in post-stroke dysgraphia. 20 individuals (mean age = 63) were administered dysgraphia word-form treatment for 40 words over ~10 weeks. All were administered an fMRI language-based learning localizer, consisting of brief stories and reversed speech, with instructions that comprehension questions would follow. First, for each group, whole-brain voxel-based analyses were used to identify ROIs in which extent of behavioral improvement in word learning or relearning was associated with BOLD response in the localizer task. This yielded two sets of ROIs: word learning and word re-learning ROIs. Second, for every ROI in each set, we used LMEM to evaluate if individual differences in learning for the other participant group was predicted by each individual’s BOLD response. (i.e., word learning ROIs were evaluated for an association between BOLD and word re-learning performance, and vice versa). RESULTS Word learning substrates shared with re-learning. Five bilateral ROIs (n=10) exhibited significant relationships between word learning and BOLD response: medial temporal lobe (MTL)), prefrontal lobe, insula, superior temporal gyrus (STG) and middle temporal gyrus (MTG). Of these, 6 also exhibited significant relationships between word re-learning performance and BOLD response: right MTL, right insula, and bilateral prefrontal ROI and bilateral MTG. Further analysis revealed that, overall, word-relearning was associated with right-lateralized neural recruitment, while word learning did not show significant lateralization. Unique word re-learning substrates. Four ROIs, not identified in the previous analysis, exhibited significant relationships between word re-learning and BOLD response: right dorso-lateral frontoparietal cortex, supplementary motor area and bilateral lingual gyrus. Of these, only the right dorsal lateral frontoparietal ROI showed a significant relationship between word learning performance and BOLD, supporting the inference that the other three ROIs support mechanisms uniquely associated with re-learning. CONCLUSIONS We found evidence that word-form relearning after stroke is associated with many of the same mechanisms supporting new word-form learning in healthy adults, while relearning also recruits unique mechanisms. Shared areas include both subcortical (MTL) and cortical areas (prefrontal cortex) memory areas, as well as classic language areas (insula and MTG). Unique to relearning was the overall right-lateralized recruitment of the shared areas and the contributions of right supplementary motor cortex and bilateral lingual gyri.
Topic Areas: Speech-Language Treatment, Disorders: Acquired