Search

Objective: We evaluated changes in cognitive domains after neurosurgical lesioning of cortical sites with significant high‐gamma power modulations (HGM) during a visual naming task, although these sites were found language‐negative on standard‐of‐care electrical stimulation mapping (ESM). Methods: In drug‐resistant epilepsy patients who underwent resection/ablation after stereo‐electroencephalography (SEEG), we computed reliable change indices (RCIs) from a battery of presurgical and 1‐year postsurgical neuropsychological assessments. We modeled RCIs as a function of lesioning even one HGM language site, number of HGM language sites lesioned, and the magnitude of naming‐related HGM. The analyses were adjusted for 1‐year seizure freedom, operated hemispheres, and the volumes of surgical lesions. Results: In 37 patients with 4455 SEEG electrode contacts (1839 and 2616 contacts in right and left hemispheres, respectively), no ESM language sites were lesioned. Patients with lesioning of even one HGM language site showed significantly lower RCIs for Peabody Picture Vocabulary Test (PPVT), working memory, and verbal learning immediate (VLI) scores. RCI declines with higher number of HGM language sites lesioned were seen in PPVT (slope [β] = −.10), working memory (β = −.10), VLI (β = −.14), and letter–word identification (LWI; β = −.14). No neuropsychological domains improved after lesioning of HGM language sites. Significant effects of the HGM magnitude at lesioned sites were seen on working memory (β = −.31), story memory immediate (β = −.27), verbal learning recognition (β = −.18), LWI (β = −.16), spelling (β = −.49), and passage comprehension (β = −.33). Because working memory was significantly affected in all three analyses, patients with maximal working memory decline were examined post hoc, revealing that all such patients had HGM naming sites lesioned in the posterior quadrants of either hemisphere. Significance: HGM language mapping should be used as an adjunct to ESM in clinical practice and may help counsel patients/families about postsurgical cognitive deficits. [ABSTRACT FROM AUTHOR]

Objective: Delay to resective epilepsy surgery results in avoidable disease burden and increased risk of mortality. The objective was to prospectively validate a natural language processing (NLP) application that uses provider notes to assign epilepsy surgery candidacy scores. Methods: The application was trained on notes from (1) patients with a diagnosis of epilepsy and a history of resective epilepsy surgery and (2) patients who were seizure‐free without surgery. The testing set included all patients with unknown surgical candidacy status and an upcoming neurology visit. Training and testing sets were updated weekly for 1 year. One‐ to three‐word phrases contained in patients' notes were used as features. Patients prospectively identified by the application as candidates for surgery were manually reviewed by two epileptologists. Performance metrics were defined by comparing NLP‐derived surgical candidacy scores with surgical candidacy status from expert chart review. Results: The training set was updated weekly and included notes from a mean of 519 ± 67 patients. The area under the receiver operating characteristic curve (AUC) from 10‐fold cross‐validation was 0.90 ± 0.04 (range = 0.83‐0.96) and improved by 0.002 per week (P <.001) as new patients were added to the training set. Of the 6395 patients who visited the neurology clinic, 4211 (67%) were evaluated by the model. The prospective AUC on this test set was 0.79 (95% confidence interval [CI] = 0.62‐0.96). Using the optimal surgical candidacy score threshold, sensitivity was 0.80 (95% CI = 0.29‐0.99), specificity was 0.77 (95% CI = 0.64‐0.88), positive predictive value was 0.25 (95% CI = 0.07‐0.52), and negative predictive value was 0.98 (95% CI = 0.87‐1.00). The number needed to screen was 5.6. Significance: An electronic health record–integrated NLP application can accurately assign surgical candidacy scores to patients in a clinical setting. [ABSTRACT FROM AUTHOR]

Objective: We describe the development and evaluation of a system that uses machine learning and natural language processing techniques to identify potential candidates for surgical intervention for drug-resistant pediatric epilepsy. The data are comprised of free-text clinical notes extracted from the electronic health record (EHR). Both known clinical outcomes from the EHR and manual chart annotations provide gold standards for the patient’s status. The following hypotheses are then tested: 1) machine learning methods can identify epilepsy surgery candidates as well as physicians do and 2) machine learning methods can identify candidates earlier than physicians do. These hypotheses are tested by systematically evaluating the effects of the data source, amount of training data, class balance, classification algorithm, and feature set on classifier performance. The results support both hypotheses, with F-measures ranging from 0.71 to 0.82. The feature set, classification algorithm, amount of training data, class balance, and gold standard all significantly affected classification performance. It was further observed that classification performance was better than the highest agreement between two annotators, even at one year before documented surgery referral. The results demonstrate that such machine learning methods can contribute to predicting pediatric epilepsy surgery candidates and reducing lag time to surgery referral. [ABSTRACT FROM AUTHOR]

Objective Methods Results Significance The most common medically resistant epilepsy (MRE) involves the temporal lobe (TLE), and children designated as temporal plus epilepsy (TLE+) have a five‐times increased risk of postoperative surgical failure. This retrospective, blinded, cross‐sectional study aimed to correlate visual and computational analyses of magnetoencephalography (MEG) virtual sensor waveforms with surgical outcome and epilepsy classification (TLE and TLE+).Patients with MRE who underwent MEG and iEEG monitoring and had at least 1 year of postsurgical follow‐up were included in this retrospective analysis. User‐defined virtual sensor (UDvs) beamforming was completed with virtual sensors placed manually and symmetrically in the bilateral amygdalohippocampi, inferior/middle/superior temporal gyri, insula, suprasylvian operculum, orbitofrontal cortex, and temporoparieto‐occipital junction. Additionally, MEG effective connectivity was computed and quantified using eigenvector centrality (EC) to identify hub regions. More conventional MEG methods (equivalent current dipole [ECD], standardized low‐resolution brain electromagnetic tomography, synthetic aperture magnetometry beamformer), UDvs beamformer, and EC hubs were compared to iEEG.Eighty patients (38 female, 42 male) with MRE (mean age = 11.3 ± 6.2 years, range = 1.0–31.5) were identified and included. Twenty‐five patients (31.3%) were classified as TLE, whereas 55 (68.8%) were TLE+. When modeling the association between MEG method, iEEG, and postoperative surgical outcome (odds of a worse [International League Against Epilepsy (ILAE) class > 2] outcome), a significant result was seen only for UDvs beamformer (odds ratio [OR] = 1.22, 95% confidence interval [CI] = 1.01–1.48). Likewise, when the relationship between MEG method, iEEG, and classification (TLE and TLE+) was modeled, only UDvs beamformer had a significant association (OR = 1.47, 95% CI = 1.13–1.92). When modeling the association between EC hub location and resection/ablation to postoperative surgical outcome (odds of a good [ILAE 1–2] outcome), a significant association was seen (OR = 1.22, 95% CI = 1.05–1.43).This study demonstrates a concordance between UDvs beamforming and iEEG that is related to both postsurgical seizure outcome and presurgical classification of epilepsy (TLE and TLE+). UDvs beamforming could be a complementary approach to the well‐established ECD, improving invasive electrode and surgical resection planning for patients undergoing epilepsy surgery evaluations and treatments. [ABSTRACT FROM AUTHOR]

Objective: The postsynaptic density protein of excitatory neurons PSD-95 is encoded by discs large MAGUK scaffold protein 4 (DLG4), de novo pathogenic variants of which lead to DLG4-related synaptopathy. The major clinical features are developmental delay, intellectual disability (ID), hypotonia, sleep disturbances, movement disorders, and epilepsy. Even though epilepsy is present in 50% of the individuals, it has not been investigated in detail. We describe here the phenotypic spectrum of epilepsy and associated comorbidities in patients with DLG4-related synaptopathy., Methods: We included 35 individuals with a DLG4 variant and epilepsy as part of a multicenter study. The DLG4 variants were detected by the referring laboratories. The degree of ID, hypotonia, developmental delay, and motor disturbances were evaluated by the referring clinician. Data on awake and sleep electroencephalography (EEG) and/or video-polygraphy and brain magnetic resonance imaging were collected. Antiseizure medication response was retrospectively assessed by the referring clinician., Results: A large variety of seizure types was reported, although focal seizures were the most common. Encephalopathy related to status epilepticus during slow-wave sleep (ESES)/developmental epileptic encephalopathy with spike-wave activation during sleep (DEE-SWAS) was diagnosed in >25% of the individuals. All but one individual presented with neurodevelopmental delay. Regression in verbal and/or motor domains was observed in all individuals who suffered from ESES/DEE-SWAS, as well as some who did not. We could not identify a clear genotype-phenotype relationship even between individuals with the same DLG4 variants., Significance: Our study shows that a subgroup of individuals with DLG4-related synaptopathy have DEE, and approximately one fourth of them have ESES/DEE-SWAS. Our study confirms DEE as part of the DLG4-related phenotypic spectrum. Occurrence of ESES/DEE-SWAS in DLG4-related synaptopathy requires proper investigation with sleep EEG., (© 2023 International League Against Epilepsy.)