You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster B36, Wednesday, November 8, 3:00 – 4:15 pm, Harborview and Loch Raven Ballrooms

Changes in neural activity during a semantic verification task as a result of treatment in persons with aphasia

Shreya Chaturvedi1, Jeffrey Johnson1, Yansong Geng1, Erin Meier1, Swathi Kiran1;1Boston University

Introduction Numerous studies have investigated the effects of rehabilitation on neural activation in persons with aphasia (PWA); however, results vary. Some report the recruitment of right language regions (Peck et al., 2004; Thiel et al., 2013) and others report the importance of the left hemisphere and perilesional activation in semantic processing specifically (Fridriksson, 2010; Fridriksson et al., 2012). To build upon these investigations, we examined changes in activation during a semantic feature verification task after a 12-week treatment for anomia. Methods Twenty PWA (age range = 42 – 79 years, M = 61, SD = 10.9; months post onset [MPO] range = 10 – 152; M = 56, SD = 49.3) received language treatment to improve naming for 36 items over two categories. Treatment was provided twice a week for 2 hours per session and involved sorting pictured items by category; multiple attempts at naming pictured items; reading/reviewing semantic features of an item; and verifying whether or not various features applied to pictured items. Treatment was terminated after 12 weeks or when PWA achieved 90% accuracy on two consecutive weekly naming probes. Before and after treatment, PWA were scanned on a semantic feature verification task to decide whether or not written features applied to pictured items. Stimuli were pictures of trained items (experimental condition) and scrambled pictures (control condition). To obtain normative data on task-specific activation, 19 healthy control participants were scanned on the task as well. Individual and group analyses based on the General Linear Model (GLM) were performed for each time point (pre- and post-treatment) for the contrast of: trained pictures – scrambled pictures. Additionally, treatment-induced changes in activation were identified in SPM post-treatment relative to pre-treatment. We also accounted for percent improvement (average percent change across two trained categories) and lesion volume in whole brain analyses. Results 18 out of 20 patients improved after treatment, as indicated by a paired t-test comparing their average pre-treatment and post-treatment naming accuracy (t17 = 4.9, p< 0.001) (M = 25%, range = -6 to 65%, SD = 0.2). Healthy participants showed significant activation for semantic feature verification in Left Inferior Frontal Gyrus (LIFG), and) LPosterior Medial Temporal Gyrus (LPMTG). In patients, increased activation that emerged as a function of treatment was found in LSuperior MG, LHippocampus, LS Occipital Gyrus, LCuneus, RAnterior Cingulate Cortex, RParacentral Lobule, and RAngular Gyrus. Bilateral activation was demonstrated in IFG, MCC, MFG, MOG, Pre and Postcentral Gyrus, and Precuneus. Accounting for percent improvement in rehabilitation, LSFG, LMFG, RSFG, RMFG, RIFG, LPostcentral Gyrus, LSOG, LPrecuneus, RAngular Gyrus, and RSOG showed increased activation. Conclusion Our results indicate patterns of activation consistent with previous studies that PWA employ traditional language specific regions like IFG (Turkeltaub, Messing, Norise, & Hamilton, 2011), However, we also found that patients showed increased activation in domain general regions like MFG, ACC, and Precuneus (Fedorenko, Duncan, & Kanwisher, 2013). Cumulatively, this suggests that the semantic verification tasks engage both task specific and domain general systems as a result of semantic feature based treatment.

Topic Area: Language Therapy

Back to Poster Schedule