Large-scale neural slowing measured with electroencephalography indexes secondary thalamic degeneration in stroke.

A central priority in stroke recovery research is the development of useful biomarkers that can reveal underlying disease states to aid diagnosis, prognosis, stratification, and treatment, ideally using tools that are readily available in clinical settings. Large-scale neural slowing has the potential to be such a biomarker, but its usefulness is currently limited by uncertainty about its nature and underlying causes. In this work, we sought to address these gaps by parameterizing abnormal resting state electroencephalography (EEG) spectral features across the scalp, and investigating their relationship to thalamic atrophy and dysfunction, using structural magnetic resonance imaging (MRI) and a computational model of corticothalamic circuit dynamics, respectively. As predicted, stroke patients (n=25) exhibited widespread spectral abnormalities, including significantly increased aperiodic exponent and offset, lower alpha frequency, and reduced beta power, which together can account for the frequently observed shift in power towards low frequencies after stroke. Furthermore, corticothalamic models fit to power spectra across the scalp inferred broad thalamic disinhibition. Crucially, these abnormalities (except beta reductions) were strongly predicted by ipsilesional thalamic atrophy measured with MRI, despite the lack of direct thalamus damage in this sample. Together, these findings highlight secondary thalamic injury as an important consequence of stroke and potential cause of widespread neural dysfunction, helping to clarify the underlying causes of post-stroke neural slowing, and demonstrating its feasibility as a clinically-accessible biomarker.