Cortical network topological modifications underlie clinical evolution in the acute phase of ischemic stroke

Iacovelli C.;Reale G.;Baldazzi G.;Pani D.;Zauli A.;Moci M.;Manganotti P.;Marinelli L.;Sacco S.;Furlanis G.;Ajcevic M.;Giovannini S.

;Crosetti S.;Grazzini M.;Garbuglia M.;Caliandro P.

2025-01-01

---

Abstract

Background: Following stroke, brain networks can be described by strength of local connections (clustering coefficient [Cw]) and strength of global interconnections (path length [Lw]) between nodes, and their balance (Small-worldness [Sw]). Objective. To identify electroencephalography (EEG) networks predicting clinical evolution in stroke through a multicenter cross-sectional study. Methods: We consecutively recruited 87 anterior circulation ischemic stroke patients. We obtained resting-state EEG (31 electrodes, 10-10 system) within 24 hours from stroke (T0) and at discharge from stroke unit (4–10 days after stroke; T1). EEG data were elaborated with EEGLAB and Lagged Linear Coherence among cortical sources of EEG signals was analyzed using eLORETA. We performed a multiple linear regression with National Institutes of Health Stroke Scale (NIHSS) at T0 and T1 as dependent variables and Cw, Lw, and Sw of delta, theta, and alpha networks as independent variables. Results: We found a negative association between alpha1 Sw and NIHSS at T0 (β = −.232, P = .04) meaning that the lower is alpha efficiency the higher is clinical severity and a positive association between delta Sw and NIHSS at T1 (β = .423, P < .001) meaning that the higher is delta efficiency the higher is clinical severity. We found positive association between delta Sw at T0 and NIHSS at T1 (β = .259, P = .02), meaning that the higher is delta efficiency in the hyperacute phase the higher is clinical severity at T1. Conclusions: A higher delta Sw within 24 hours after stroke is associated to higher NIHSS within 10 days. Delta brain network rearrangement in the hyperacute phase is a potential neurophysiological measure to be integrated in multi-modal prognostic models.