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

Poster B8, Wednesday, November 8, 3:00 – 4:15 pm, Harborview and Loch Raven Ballrooms

Spatiotemporal neuronal activation patterns during verbal fluency tasks

Shawniqua T. Williams1, Preya A. Shah1, Vitória Piai2, Heather Gatens1, Abba Krieger1, Timothy H. Lucas, II1, Brian Litt1;1University of Pennsylvania, 2Radboud University

-----BACKGROUND----- Tests of verbal fluency are frequently used to evaluate linguistic and cognitive function in a variety of neurocognitive disorders including epilepsy, Alzheimer’s disease and ADHD (Troyer et al. 1997). Impaired performance reflects deficits in lexical access ability and/or executive control (Shao et al. 2014). Functional studies show that verbal fluency tasks activate a broad network of bifrontal and bitemporal regions, most notably the left inferior frontal, precentral and fusiform gyri (Parks et al. 1988; Birn et al. 2010). However, the timing of activation of these regions has yet to be elucidated. Intracranial electroencephalography (iEEG) provides a unique opportunity to investigate the neuronal activation patterns associated with verbal fluency task performance with high spatial and temporal precision. -----METHODS----- We used subdural recordings to study the neuronal activation patterns in patients undergoing presurgical evaluation for intractable epilepsy. Silastic-embedded platinum-iridium electrodes (AdTech Medical) were placed in the subdural spaces or inserted into mesial temporal regions based on clinical necessity. Recordings were obtained at rest and during semantic and phonemic verbal fluency tasks using a Natus or Neuralynx data acquisition system (sampling frequency >500Hz). Epochs were extracted from 1250ms prior to 750ms after onset of each utterance. After removal of spikes and other artifact, high gamma (70-110Hz) spectral power was determined in 200ms moving windows with 10ms step sizes using the Thomson multitaper method (Thomson 1982), and log-normalized by the resting high gamma power at each contact. For one subject, high gamma estimates were limited to 85-110Hz due to artifact in the 70-80Hz range. -----RESULTS----- Eight implants were recorded, 5 of which were predominantly left hemispheric. All subjects were left-dominant for language and had epilepsy onset after language development. Left temporal lobe epilepsy (TLE) patients tended to have worse performance than right TLE patients. In all but one subject we observed robust increases in high gamma activity in sensorimotor cortex starting 300ms prior to utterance onset. There were also high gamma increases in posterior temporal regions starting 250ms after utterance onset. Differences between semantic and phonemic conditions varied in timing and location across subjects. Among subjects in whom mesial temporal depth electrodes were present, a significant correlation was observed between maximum (across electrode contacts) task-related hippocampal high gamma activity and production rate in semantic (p<0.01) but not phonemic (p=0.72) verbal fluency tasks. This relationship was not seen when the sampled contacts were restricted to hippocampi involved in the epileptic onset zone. -----DISCUSSION----- To our knowledge, this is the first study using iEEG to characterize spatiotemporal activation patterns in a letter- and category-motivated free recall format. Our results are consistent with studies of stimulus-locked word generation (Edwards et al. 2010; Korzeniewska et al. 2008; Indefrey 2011). Variability in the differences between semantic and phonemic verbal fluency conditions may reflect inter-individual variability in search strategies and conceptual representations (Hirshorn & Thompson-Schill 2006; Wang et al. 2011). The correlation between hippocampal activity and performance appeared to be driven by the healthy (nonepileptic) hippocampus. Further analyses will investigate the roles of lower frequency oscillations and category imageability.

Topic Area: Control, Selection, and Executive Processes

Back to Poster Schedule