Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Background: About one out of three stroke patients are affected by aphasia, an acquired language disorder which impairs the production and/or comprehension of spoken and/or written language. Traditionally, behavioral tests are used during aphasia assessment. However, due to the lack of objective results, questionable psychometric properties, the influence of confounding factors, and the inability to test too severely impaired patients, objective measures might form promising additional assessment tools. Resting-state electroencephalography (rsEEG) is emerging as a convenient method for this, as it does not require overt task performance or sensory stimulation, is inexpensive, has no contra-indications for use and is already present in most hospitals. The present study aims to contribute to filling the gap in the literature concerning functional connectivity (FC) during rsEEG in people with aphasia (PWA). More specifically, it aims to answer the question whether graph metrics characterizing FC networks as measured during rsEEG can predict aphasia severity. Methods: EEG recordings were carried out in 26 PWA during resting state. The weighted phase lag index was used as a measure of FC between the EEG-signals at scalp level. A global and a local graph measure, average characteristic path length and node strength respectively, were calculated for six frequency bands: delta (1–4 Hz), theta (4.5-7 Hz), alpha (7.5-14.5 Hz), beta (15-30 Hz), low-gamma (30.5-49 Hz) and mid-gamma (50-90 Hz). Node strength was averaged across frontal, temporal, parietal, and occipital scalp regions. Two behavioral measures were used to assess aphasia severity, namely the e-ScreeLing and the Nederlandse Benoem Test (“Dutch Naming Test”). Multiple linear regression and Spearman’s rank correlations were implemented to examine whether the graph measures can predict and/or correlate with aphasia severity. Results: The results show that average path length predicts aphasia severity in the delta and theta bands and negatively correlates with aphasia severity in the delta band. Positive correlations with aphasia severity were found for average delta and theta node strength in all four regions. Moreover, in the gamma bands, average node strength in the frontal and occipital lobes was found to negatively correlate with aphasia severity. No significant predictors or correlations were identified in the alpha and beta bands. Conclusion: Based on the current findings, both global network properties in the delta and theta band and local properties in the delta, theta and gamma bands correlate with aphasia severity. While correlations in the gamma bands seem to reflect compensatory mechanisms, it is less clear if changes in the delta and theta bands reflect compensation of language deficits through the engagement of extra cognitive resources during language processing or rather reflect inefficient language processing through overrecruitment or ineffective inhibition of activity irrelevant to language processing. It remains unclear which specific regions are responsible for these findings due to their low spatial resolution, but as correlations with the local measure were found across all four cortical lobes, it is possible that different parts of the language network were involved and correlate with aphasia. Future research is necessary to gain further insight into rsFC in PWA and to establish whether EEG may be a reliable diagnostic tool in the assessment of aphasia.