Objective: This study was undertaken to investigate the regional burden of enlarged perivascular spaces (EPVSs) in patients with temporal lobe epilepsy with hippocampal sclerosis (TLE‐HS) and explore its prognostic relevance. Methods: In this retrospective observational study, EPVSs in the temporal lobe (T‐EPVS), centrum semiovale (CS‐EPVS), basal ganglia (BG‐EPVS), midbrain, and hippocampus were visually rated in 68 treatment‐naïve patients with TLE‐HS. Regional EPVS burden was dichotomized into high and low degrees (cutoff: >10 for BG‐EPVS/T‐EPVS; >20 for CS‐EPVS). Cox proportional hazards models were used to determine the potential predictors of seizure freedom (SF; no seizure for >1 year) and delayed SF (SF achieved >6 months after initiating antiseizure medication [ASM]). Multivariate logistic regression using stepwise variable selection based on the Akaike information criterion was performed to investigate whether EPVS burden was associated with medical refractoriness (never achieving SF). Results: Of the 68 patients, 20 were classified into the refractory group (29.4%). The high T‐EPVS group had an older epilepsy onset (37.3 ± 12.3 vs. 26.5 ± 13.0 years, p = .005), higher pretreatment seizure density (median = 12.0, interquartile range [IQR] = 5.0–20.0 vs. 4.0, IQR = 2.0–10.5, p = .008), and lower focal to bilateral tonic–clonic seizure prevalence (13.3% vs. 73.6%, p < .001) than the low T‐EPVS group. High T‐EPVS burden (odds ratio [OR] = 10.908, 95% confidence interval [CI] = 1.895–62.789) was an independent predictor of medial refractoriness, along with female sex (OR = 12.906, 95% CI = 2.214–75.220) and ASM treatment duration (OR = .985, 95% CI = .971–.999). The low T‐EPVS group had higher probability of achieving delayed SF than the high T‐EPVS group ( p _Log‐rank = .030, p _{Cox regression} = .038), whereas the probability of achieving SF was comparable between the two groups ( p _Log‐rank = .053, p _{Cox regression} = .146). Significance: Increased T‐EPVS burden may serve as an imaging marker of unfavorable prognosis in patients with TLE‐HS, underscoring the potential role of perivascular dysfunction in diminished ASM response.

Recently, the most bothersome symptom has been recommended as a co-primary endpoint in clinical trials on the acute treatment of migraine. Probable migraine is a subtype of migraine that fulfills all but one criterion for migraine diagnosis. We aimed to compare the most bothersome symptom between probable migraine and migraine. This study analyzed data from a nationwide study conducted in Korea, and the most bothersome symptom was assessed by requesting the participants to select one of the four typical accompanying symptoms of migraine. Responses to acute treatment were evaluated using the migraine Treatment Optimization Questionnaire-6. Nausea was the most bothersome symptom, followed by phonophobia and vomiting in the migraine group (nausea, 61.8%; phonophobia, 25.3%; vomiting, 10.0%; and photophobia, 2.9%) and the probable migraine group (nausea, 82.2%; phonophobia, 9.5%; vomiting, 5.6%; and photophobia, 2.7%). In participants with migraine, vomiting (adjusted odds ratio = 6.513; 95% confidence interval, 1.763–24.057) and phonophobia (adjusted odds ratio = 0.437; 95% confidence interval, 0.206–0.929) were significantly associated with severe headache intensity and nausea was significantly associated with >3 headache days per 30 days (adjusted odds ratio = 0.441; 95% confidence, 0.210–0.927). Different patterns of associations were observed in probable migraine.

Objective: Lennox–Gastaut syndrome (LGS) is typically characterized by drug‐resistant epilepsy and subsequent cognitive deterioration. Surgery is a rare but viable option for the control of seizures in a subset of patients with LGS. This study aimed to describe the organization of the epileptogenic zone network (EZN) in patients with LGS using stereoelectroencephalography (SEEG) and to report the outcome of post‐SEEG treatment. Methods: A quantitative SEEG signal analysis was conducted in 14 consecutive patients with LGS, in whom a potentially localized EZN was suggested based on a comprehensive noninvasive evaluation. The EZN and the irritative zone network were identified using relevant biomarkers during ictal (epileptogenicity index and connectivity epileptogenicity index) and interictal (spikes and high‐frequency oscillations) recordings. The applied post‐SEEG treatments were assessed, including SEEG‐guided radiofrequency thermocoagulation (RF‐TC), surgery, and neurostimulation. Results: The seizure onset patterns showed some specificity by seizure type, with 84% of tonic seizures involving low‐voltage fast activity. The EZN of patients with LGS was often, but not always, complex and extensive, involving two or more lobes (79%) and both hemispheres (64%). The lateral neocortical structures, particularly the lateral premotor and dorsolateral prefrontal cortices, were identified as being most frequently involved in the EZN. Among the explored subcortical structures, only the pulvinar, central–lateral thalamic nucleus, and hypothalamic hamartoma belonged to the EZN. Twelve patients (86%) underwent SEEG‐guided RF‐TC, with 50% experiencing a >50% reduction in baseline seizure frequency. Four patients (29%) underwent curative surgery for significant involvement of a lesion in the EZN, and one case achieved an Engel class I outcome. Significance: This is the first quantitative SEEG study in patients with LGS to demonstrate the utility of SEEG in identifying patients who may benefit from surgery and to perform SEEG‐guided RF‐TC. Nevertheless, the indications for SEEG should be carefully assessed, as localized EZN is uncommon in LGS.

Background: Delirium affects up to 50% of patients following high-risk surgeries and is associated with poor long-term prognosis. This study employed machine learning to predict delirium using polysomnography (PSG) and sleep-disorder questionnaire data, and aimed to identify key sleep-related factors for improved interventions and patient outcomes. Methods: We studied 912 adults who underwent surgery under general anesthesia at a tertiary hospital (2013–2024) and had PSG within 5 years of surgery. Delirium was assessed via clinical diagnoses, antipsychotic prescriptions, and psychiatric consultations within 14 days postoperatively. Sleep-related data were collected using PSG and questionnaires. Machine learning predictions were performed to identify postoperative delirium, focusing on model accuracy and feature importance. Results: This study divided the 912 patients into an internal training set (700) and an external test set (212). Univariate analysis identified significant delirium risk factors: midazolam use, prolonged surgery duration, and hypoalbuminemia. Sleep-related variables such as fewer rapid eye movement (REM) episodes and higher daytime sleepiness were also linked to delirium. An extreme gradient-boosting-based classification task achieved an AUC of 0.81 with clinical variables, 0.60 with PSG data alone, and 0.84 with both, demonstrating the added value of PSG data. Analysis of Shapley additive explanations values highlighted important predictors: surgery duration, age, midazolam use, PSG-derived oxygen saturation nadir, periodic limb movement index, and REM episodes, demonstrating the relationship between sleep patterns and the risk of delirium. Conclusions: The artificial intelligence model integrates clinical and sleep variables and reliably identifies postoperative delirium, demonstrating that sleep-related factors contribute to its identification. Predicting patients at high risk of developing postoperative delirium and closely monitoring them could reduce the costs and complications associated with delirium.

Kim SJ, et al. J Korean Med Sci. 2023 Jul;38(31):e237. https://doi.org/10.3346/jkms.2023.38.e237