Poster D50, Friday, August 17, 4:45 – 6:30 pm, Room 2000AB

Effects of active and sham tDCS on lexical decision in three persons with chronic aphasia

Rachael M. Harrington1, Simone R. Roberts1,2, Lisa C. Krishnamurthy1,2,3, Venkatagiri Krishnamurthy2,3, Amy D. Rodriguez3, Keith M. McGregor2,3, Marcus Meinzer4, Bruce Crosson1,2,3;1Georgia State University, Atlanta, GA, USA, 2Emory University, Atlanta, GA, USA, 3Atlanta VAMC, Decatur, GA, USA, 4University of Queensland, Herston, QLD, AU

Introduction: Noninvasive brain stimulation can be used to improve language recovery after stroke. It is unclear how tDCS montage should be selected based on task demands and individual patient profiles. Here we present performance differences on a lexical-decision task during sham and active tDCS in three patients with unique lesion profiles. Methods: 3 native-English speaking adults with chronic aphasia were enrolled in the study. MPRAGE were collected from each patient (Siemens Prisma 3T (TR=2530 ms, TE=2.96 ms, 1 mm slice thickness)). In the scanner, patients performed an auditory lexical decision task in which they indicated via button press whether the stimulus was a real word or nonword. Real words were balanced for number of syllables, frequency, category, and number of phonemes, and nonwords were created from this set. In two separate sessions, patients performed the lexical decision task under sham or active tDCS (NeuroConn, DC-Stimulator MC4 4-channel) with the cathode over F8 and the anode over F7. Accuracy and reaction time data were collected in E-Prime. The MPRAGE images were twice warp-transformed into MNI152 space using LINDA and proprietary algorithms. Lesion size was determined in mricron and lesions were overlaid on the Harvard-Oxford cortical and subcortical structural atlases and the 1 mm JHU white-matter tractography atlas to determine lesion location. Results: QED03, a 66-year-old male with mild nonfluent aphasia had a total lesion volume of 39.38 cc and large areas of the lesion occupied pars opercularis (pOp) (31.1%), pars triangularis (pTr) (15.1%), Heschl’s gyrus (12.4%) and anterior superior temporal gyrus (aSTG) (1.7%). On the lexical decision task, QED03’s response time decreased for both nonwords and real words in the active tDCS condition. Accuracy, however, worsened when discriminating nonwords but improved for discriminating real words. QED05, a 59-year-old male with mild fluent aphasia had a smaller lesion with a total lesion volume of 3.73 cc. This lesion occupied only 3.6% of pOp. In the active tDCS condition, response time decreased for nonwords and increased for real words while accuracy improved for nonwords but decreased for real words. QED06, a 55-year-old male with moderate nonfluent aphasia had a lesion volume of 0.7 cc. This lesion occupied 1.4% of aSTG. In the active tDCS condition, accuracy and response time worsened for discriminating nonwords and real words. Discussion: Task demands and lesion profiles may interact in these outcomes. In this montage, excitability of cortex under anode increases in left frontal areas and decreases in right homologue. Patients with lesions in left frontal gyrus improved on some aspect of the task while the patient with more posterior lesion worsened on the task in the active tDCS condition. After increasing IFG excitability in the lesioned hemisphere, performance associated with the lesioned area improved. The selected montage, however, likely did not affect posterior excitability and excitation of left hemisphere pTr may have prevented balanced activation of the nonlesioned area. We hypothesize that task sensitivities and individual aphasia/lesion profiles may influence findings and should be further examined with this and other montages.

Topic Area: Language Disorders