Poster Abstracts

Abstract

Lesion-symptom maps identify lesion locations significantly associated with performance on cognitive measures. Importantly, lesion-symptom maps can be utilized to predict cognitive outcomes in independent cohorts of patients. However, it remains unknown whether adult-derived lesion-symptom maps are also useful for predicting outcomes for individuals with developmental-onset lesions. We evaluated this possibility using lesion-symptom maps of the Wide Range Achievement Test Word Reading and Spelling subtests derived from participants with adult-onset lesions. These lesion-symptom maps were used to predict cognitive performance from lesion location in participants with developmental-onset lesions. All lesion-symptom maps predicted a statistically significant amount of variance for outcomes of the developmental-onset sample, with small effect sizes (i.e., reading r = .21, p = .046; spelling r = .27, p = 0.03). These effect sizes were similar to the within-sample cross-validation effect sizes. In sum, lesion-symptom maps derived from adult-onset subjects may have utility for predicting language outcomes for patients with developmental-onset lesions. However, the small effect sizes raise the question as to whether critical neuroanatomical regions for language differ as a function of age of lesion onset. Future work could emphasize lesion-symptom mapping of developmental-onset brain lesions to aid in predicting outcomes after early brain injury.

First AuthorAlyssa W Sullivan
Second AuthorJax D Skye
Third AuthorMark D Bowren
Fourth AuthorDaniel Tranel
Fifth AuthorAaron D Boes