Search

• Machine learning was used to predict seizure recurrence with electrical status epilepticus (ESES) in idiopathic focal epilepsy. • The partial directed coherence is a reliable feature for predicting seizure recurrence with ESES via machine learning. • Among 6 machine learning algorithms, the Xgboost Classifier model achieved a classification accuracy up to 0.85. Some individuals with idiopathic focal epilepsy (IFE) experience recurring seizures accompanied by the evolution of electrical status epilepticus during sleep (ESES). Here, we aimed to develop a predictor for the early detection of seizure recurrence with ESES in children with IFE using resting state electroencephalogram (EEG) data. The study group included 15 IFE patients who developed seizure recurrence with ESES. There were 17 children in the control group who did not experience seizure recurrence with ESES during at least 2-year follow-up. We used the degree value of the partial directed coherence (PDC) from the EEG data to predict seizure recurrence with ESES via 6 machine learning (ML) algorithms. Among the models, the Xgboost Classifier (XGBC) model achieved the highest specificity of 0.90, and a remarkable sensitivity and accuracy of 0.80 and 0.85, respectively. The CATC showed balanced performance with a specificity of 0.85, sensitivity of 0.73, and an accuracy of 0.80, with an AUC equal to 0.78. For both of these models, F4, Fz and T4 were the overlaps of the top 4 features. Considering its high classification accuracy, the XGBC model is an effective and quantitative tool for predicting seizure recurrence with ESES evolution in IFE patients. We developed an ML-based tool for predicting the development of IFE using resting state EEG data. This could facilitate the diagnosis and treatment of patients with IFE. [ABSTRACT FROM AUTHOR]

• Children with resolved Rolandic epilepsy have increased median nerve somatosensory conduction time compared to controls. • Ventral thalamic volume positively correlated with N20 conduction time. • Focally decreased Rolandic thalamocortical connectivity may support symptom resolution in Rolandic epilepsy. Median nerve somatosensory evoked fields (SEFs) conduction times reflect the integrity of neural transmission across the thalamocortical circuit. We hypothesized median nerve SEF conduction time would be abnormal in children with Rolandic epilepsy (RE). 22 children with RE (10 active; 12 resolved) and 13 age-matched controls underwent structural and diffusion MRI and median nerve and visual stimulation during magnetoencephalography (MEG). N20 SEF responses were identified in contralateral somatosensory cortices. P100 were identified in contralateral occipital cortices as controls. Conduction times were compared between groups in linear models controlling for height. N20 conduction time was also compared to thalamic volume and Rolandic thalamocortical structural connectivity inferred using probabilistic tractography. The RE group had slower N20 conduction compared to controls (p = 0.042, effect size 0.6 ms) and this difference was driven by the resolved RE group (p = 0.046). There was no difference in P100 conduction time between groups (p = 0.83). Ventral thalamic volume positively correlated with N20 conduction time (p = 0.014). Children with resolved RE have focally decreased Rolandic thalamocortical connectivity. These results identify a persistent focal thalamocortical circuit abnormality in resolved RE and suggest that decreased Rolandic thalamocortical connectivity may support symptom resolution in this self-limited epilepsy. [ABSTRACT FROM AUTHOR]

Abstract: The role of neuronal acetylcholine receptors (nAChRs) in epilepsy has been clearly established by the finding of mutations in a subset of genes coding for subunits of the nAChRs in a form of sleep-related epilepsy with familial occurrence in about 30% of probands and dominant inheritance, named autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE). Sporadic and familial forms have similar clinical and EEG features. Seizures begin in middle childhood as clusters of sleep-related attacks with prominent motor activity, and sustained dystonic posturing. In addition to nocturnal seizures, psychosis or schizophrenia, behavioral disorders, memory deficits and mental retardation were described in some individuals. Although over hundred families are on record, only a minority of them have been linked to mutations in the genes coding for the α4, α2 and β2 (CHRNA4, CHRNA2, and CHRNB2) subunits of the nAChRs, indicating that ADNFLE is genetically heterogeneous despite a relatively homogeneous clinical picture. Functional characterization of some mutations suggests that gain of the receptor function might be the basis for epileptogenesis. In vitro and in vivo studies have shown high density of nAChRs in the thalamus, over activated brainstem ascending cholinergic pathway and enhanced GABAergic function, reinforcing the hypothesis that cortico-subcortical networks, regulating arousal from sleep, play a central role in seizure precipitation in ADNFLE. [Copyright &y& Elsevier]