Electrographic Characteristics of Lateralized Periodic Discharges Associated With Seizures: A Combined Visual and Quantitative EEG Analysis in 156 Patients
Abstract number :
1.51
Submission category :
3. Neurophysiology / 3B. ICU EEG
Year :
2025
Submission ID :
1264
Source :
www.aesnet.org
Presentation date :
12/6/2025 12:00:00 AM
Published date :
Authors :
Alberto Prat, MD – Clinica Alemana - Universidad del Desarrollo
Alexandre Castro, MD – Hospital Albert Einstein
Nathan Torcida Sedano, MD – Hôpital Universitaire de Bruxelles
Lawrence Hirsch, MD – Yale University School of Medicine
Presenting Author: Nicolas Gaspard, MD, PhD – Hôpital Universitaire de Bruxelles
Rationale: Lateralized periodic discharges (LPDs) are associated with a high risk of acute electrographic seizures in critically ill patients. The risk is influenced by frequency, prevalence, sharpness, and the presence of a plus modifier. Still, as many as 50% of patients with LPDs do not have electrographic seizures. Also, in most cases, the last recorded seizures during a continuous electroencephalogram (CEEG) session are still followed by LPDs. This indicates that not all LPDs carry the same risk of impending seizures. We thus aimed to determine if characteristics of periodicity of LPDs are associated with seizure risk.
Methods: We performed retrospective visual and quantitative assessment of 10-min EEG epochs from 156 patients with LPDs, categorized into having electrographic seizures (n=66) or not (n=90) during CEEG ("seizures" and "no seizures" groups). EEG features examined included symmetry, lobar location, lateralization, frequency, plus modifiers, sharpness, triphasic morphology, fluctuation, lag, polarity, voltage, presence of brief potentially ictal rhythmic discharges (BIRDs), lateralized rhythmic delta activity (LRDA), sporadic epileptiform discharges (SEDs), and degree of encephalopathy. Visual assessment relied on the American Clinical Neurophysiology Society (ACNS) 2021 terminology. For quantitative assessment, the inter-discharge intervals between LPDs were identified in the bipolar derivation in which they were the most prominent using dynamic time warping and the first LPD of the sequence as a template. The inter-discharge interval time series was quantified by its standard deviation (a measure of variability) and with recurrence quantification analysis to extract percentage determinism (a measure of predictability) and entropy (a measure of complexity).
Results: Results are summarized in Table 1. Of all the visually assessed features, only the presence of fluctuation (50% had seizures vs. 27% without fluctuation; p=0.005) and of "plus" modifiers (61% vs. 45%; p=0.016) were associated with the risk of seizures. In quantitative analysis, standard deviation (724+/-187ms vs. 583+/-123; p=0.002), percentage determinism (0.32+/-0.08 vs. 0.23+/0.07; p=0.011), and entropy (0.41+/-0.11 vs. 0.36+/-0.08; p=0.007) of the inter-discharge intervals were all higher in the seizure group.
Conclusions: Our findings confirm that the presence of fluctuation and plus modifiers are associated with a higher probability of seizure occurrence in patients with LPDs. LPDs associated with seizures exhibited a different periodic behavior, characterized by greater variability, predictability and complexity. These features appear to be more relevant than frequency alone when evaluating seizure risk. These insights may improve our understanding of periodic discharges and the ability to predict seizures and guide more effective treatment decisions in critically ill patients.
Funding: No funding was received for this study.
Neurophysiology