Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Poster Slams
Left superior temporal sulcus lesions reduce semantic access during oral reading
Poster A19 in Poster Session A, Thursday, October 6, 10:15 am - 12:00 pm EDT, Millennium Hall
Also presenting in Poster Slam A, Thursday, October 6, 10:00 - 10:15 am EDT, Regency Ballroom
Sara M. Dyslin1, J. Vivian Dickens1, Candace M. van der Stelt1, Alycia B. Laks1, Sarah F. Snider1, Andrew T. DeMarco1, Peter E. Turkeltaub1,2; 1Center for Brain Plasticity and Recovery, Georgetown University, 2MedStar National Rehabilitation Hospital
Accurate oral reading relies on the interaction between orthographic, phonological, and semantic knowledge. The reading network in the brain is generally divided into the sublexical, dorsal route and lexical, ventral route. The degree to which semantic knowledge contributes to the lexical route is debated. The importance of semantics in reading is evidenced by the imageability effect, in which better reading performance is observed for words with high imageability (e.g., “horse”) compared to words with low imageability (e.g., “trust”). Most prior research on the role of semantics in reading has focused on individuals with semantic dementia. However, lesion studies of individuals with post-stroke alexia provide a model for isolating regions of the brain that are critical for different aspects of reading. We hypothesized that semantic contributions to reading rely on temporal lobe regions of the ventral stream; thus, lesions to these regions would reduce the imageability effect, particularly for words that are the least supported by lexical knowledge, i.e., low frequency, regular words. Participants included 53 adults with history of a left-hemisphere stroke at least six months prior to testing (Age = 62.1 (14.4); Gender = 22F, 31M; Months since stroke = 48.8(56.5). All participants completed an MRI scan and an oral word reading task with 200 words crossed on frequency, regularity, and imageability, resulting in 8 categories of words with 25 items each. Lesions were manually traced from FLAIR and T1-weighted scans and warped to MNI space. Support Vector Regression Lesion-Symptom Mapping (SVR-LSM) was applied. Analyses tested for lesions that reduced accuracy on high imageability words while controlling for the effect on low imageability words. Four analyses examined this effect for words within each combination of frequency and regularity. Age, education level, and lesion volume were regressed out of behavioral and lesion data, and 10,000 permutations were used to control the cluster-level family-wise error rate. Significant clusters were identified for the analyses of low frequency irregular words and low frequency regular words. Both clusters were localized in the temporal lobe, overlapping in the superior temporal sulcus and posterior middle temporal gyrus. There were no significant lesion locations identified for a reduced imageability effect in high frequency irregular or high frequency regular words. This study clearly demonstrates that loss of semantic support for reading does occur in stroke participants with lesions to the reading network. The findings indicate that the superior temporal sulcus and middle temporal gyrus are critical for the use of semantic knowledge in reading, and that semantic support is most evident in low frequency words. The different location of the lesions in this study relative to the primary area of atrophy in semantic dementia suggests that a different mechanism of action is disrupting semantic contributions to reading in these populations. Further research is needed to determine whether this difference is due to disconnections between semantics and orthography rather than degradation of semantic representations themselves as in semantic dementia.
Topic Areas: Reading, Disorders: Acquired