High-frequency oscillations and interictal epileptiform discharges predict infantile spasms
Hautala, S.; La Grassa, S.; Lauronen, L.; Peltola, M.; Palomäki, M.; Metsähonkala, E.-L.; Metsäranta, M.; Jonsson, H.; Gaily, E.; Harju, M.; Al-Sa'd, M.; Mikkonen, K.; Nevalainen, P.
Show abstract
Early acquired brain injury is a major risk factor for infantile epileptic spasms syndrome (IESS), which may impair cognitive development, especially if diagnosis and treatment are delayed. However, individual-level prediction of which infants will develop IESS is currently not possible. We assessed whether high-frequency oscillations (HFOs) in scalp EEG or recurrent interictal epileptiform discharges (IED) during the first months of life could predict forthcoming IESS. Our population-based cohort included 36 infants with cortical injury due to infarction, haemorrhage, infection or trauma involving a large cortical area ([≥] anterior/posterior cerebral artery territory or [≥] half of the middle cerebral artery territory), or hypoxic-ischaemic encephalopathy with cortical and deep grey matter involvement. The infants underwent repeated EEGs during the first year of life until 12 months of age or until IESS diagnosis. HFOs during sleep were scored both visually and automatically, whereas IEDs were assessed visually only. We tested whether HFO rate increased during the first year of life using a mixed-effects model with within- and between-subject random effects. Using only EEGs recorded prior to IESS diagnosis, we evaluated whether HFO rate or recurrent IEDs could predict IESS development by training a ridge-regularized logistic regression model with exhaustive leave-2-subjects-out cross-validation. Eleven infants (31%) developed IESS. HFO rate increased with age in both groups but more steeply in the IESS group [within person slope {beta} = 2.43 (IESS) vs. 0.06 (no-IESS) units/month, P < 0.001]. The logistic regression model showed that both HFO rate [AUC 0.801 (95% CI 0.668, 0.936)] and recurrent IEDs [AUC 0.826 (95% CI 0.720, 0.932)] were able to predict forthcoming IESS. However, in a multivariable model, only recurrent IEDs remained independently associated with IESS, and HFO rate did not add predictive value. The marked increase in HFO rate toward IESS diagnosis supports their role as a biomarker of epileptogenesis. During the first months of life, HFOs and recurrent IEDs performed equally well in predicting subsequent IESS. However, IEDs are easier to apply to clinical practice.
Matching journals
The top 3 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- Low frequency interictal EEG biomarker for localizing seizures 96%
- Blinded study: prospectively defined high frequency oscillations predict seizure outcome in individual patients 95%
- Excitation/Inhibition balance relates to cognitive function and gene expression in Temporal Lobe Epilepsy: an hdEEG assessment with aperiodic exponent 95%
Similar papers in this journal
- Source-sink connectivity: A novel interictal EEG marker for seizure localization 94%
- Genotype-phenotype correlations in SCN8A -related disorders reveal prognostic and therapeutic implications 94%
- Predictive models for starting antiseizure medication withdrawal following epilepsy surgery in adults 93%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.