Photo Credit: YAKOBCHUK VIACHESLAV

The following is a summary of “Electroencephalography based delirium screening in acute supratentorial stroke,” published in the November 2024 issue of Neurology by Hermann, et al.  

Delirium affects up to 25% of patients hospitalized for acute stroke, with increasing susceptibility to advancing age. Generalized slowing in the electroencephalogram (EEG) is a supporting diagnostic marker for delirium.  

Researchers conducted a prospective study to examine the efficacy of EEG-based delirium detection in patients with acute stroke.  

They received both a single-channel EEG using DeltaScan^® and a routine 21-channel EEG within 5 days post-stroke. DeltaScan^® recorded right-sided fronto-parietal EEG and analyzed polymorphic delta activity (PDA) via a proprietary algorithm, while routine EEG assessed power spectral density (PSD) across frequency bands from 2-minute, eyes-closed resting segments. EEG analysis was performed in Magnetoencephalography/EEG software (MNE, v1.3.1) in Python (3.10) and RStudio (v4.2.1).  

The results showed that delirium was diagnosed in 9 of 53 patients (52–90 years) based on the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-V) criteria. DeltaScan^® sensitivity was 44% (95% CI = 15.3–77.3%) and specificity 71% (95% CI = 57–83%). Patients with right hemispheric stroke had a higher likelihood of false positives in DeltaScan^® (P=0.01). Analysis of 21-channel EEG power revealed significant differences in frontal delta and theta power between patients with and without delirium (P<0.05).  

They concluded that bihemispheric EEG may be preferable over unilateral recordings to accurately support delirium diagnosis in patients with stroke, as single-channel EEG over the stroke site may yield false-positive results due to delta or theta slowing.  

Source: bmcneurol.biomedcentral.com/articles/10.1186/s12883-024-03942-3