Estimating the likelihood of epilepsy from clinically noncontributory electroencephalograms using computational analysis: A retrospective, multisite case-control study.

OBJECTIVE: This study was undertaken to validate a set of candidate biomarkers of seizure susceptibility in a retrospective, multisite case-control study, and to determine the robustness of these biomarkers derived from routinely collected electroencephalography (EEG) within a large cohort (both epilepsy and common alternative conditions such as nonepileptic attack disorder). METHODS: The database consisted of 814 EEG recordings from 648 subjects, collected from eight National Health Service sites across the UK. Clinically noncontributory EEG recordings were identified by an experienced clinical scientist (N = 281; 152 alternative conditions, 129 epilepsy). Eight computational markers (spectral [n = 2], network-based [n = 4], and model-based [n = 2]) were calculated within each recording. Ensemble-based classifiers were developed using a two-tier cross-validation approach. We used standard regression methods to assess whether potential confounding variables (e.g., age, gender, treatment status, comorbidity) impacted model performance. RESULTS: We found levels of balanced accuracy of 68% across the cohort with clinically noncontributory normal EEGs (sensitivity =61%, specificity =75%, positive predictive value =55%, negative predictive value =79%, diagnostic odds ratio =4.64, area under receiver operated characteristics curve =.72). Group level analysis found no evidence suggesting any of the potential confounding variables significantly impacted the overall performance. SIGNIFICANCE: These results provide evidence that the set of biomarkers could provide additional value to clinical decision-making, providing the foundation for a decision support tool that could reduce diagnostic delay and misdiagnosis rates. Future work should therefore assess the change in diagnostic yield and time to diagnosis when utilizing these biomarkers in carefully designed prospective studies.

Late Diagnosis of Narcolepsy With Cataplexy: A Novel Case of Cataplectic Facies Presenting in an Elderly Woman.

None: Cataplectic facies is an unusual feature described in children with narcolepsy and cataplexy. The typical manifestations of cataplectic facies consist of repetitive mouth opening, tongue protrusion, and ptosis. An interesting observation is that the usual emotional triggers associated with cataplexy such as laughter and joking are not always present, thus hampering diagnosis of the underlying syndrome. Cataplectic facies is thought to be a phenomenon observed in the early stages of narcolepsy type 1 and is thought to disappear by the time the patient reaches puberty. We present a unique case of an elderly woman with narcolepsy type 1 demonstrating cataplectic facies. The novel circumstances of this case highlight that facial cataplexy can present later in life, in contrast with previous descriptions that report resolution of cataplectic facies before puberty. Wider recognition of these features throughout the life course may aid in accurate diagnosis and thereby ensure swift access to appropriate treatment.

Habitual sleep durations and subjective sleep quality predict white matter differences in the human brain.

Self-imposed short sleep durations are increasingly commonplace in society, and have considerable health and performance implications for individuals. Reduced sleep duration over multiple nights has similar behavioural effects to those observed following acute total sleep deprivation, suggesting that lack of sleep affects brain function cumulatively. A link between habitual sleep patterns and functional connectivity has previously been observed, and the effect of sleep duration on the brain's intrinsic functional architecture may provide a link between sleep status and cognition. However, it is currently not known whether differences in habitual sleep patterns across individuals are related to changes in the brain's white matter, which underlies structural connectivity. In the present study we use diffusion-weighted imaging and a group comparison application of tract based spatial statistics (TBSS) to investigate changes to fractional anisotropy (FA) and mean diffusivity (MD) in relation to sleep duration and quality, hypothesising that white matter metrics would be positively associated with sleep duration and quality. Diffusion weighted imaging data was acquired from a final cohort of 33 (23-29 years, 10 female, mean 25.4 years) participants. Sleep patterns were assessed for a 14 day period using wrist actigraphs and sleep diaries, and subjective sleep quality with the Pittsburgh Sleep Quality Index (PSQI). Median splits based on total sleep time and PSQI were used to create groups of shorter/longer and poorer/better sleepers, whose imaging data was compared using TBSS followed by post-hoc correlation analysis in regions identified as significantly different between the groups. There were significant positive correlations between sleep duration and FA in the left orbito-frontal region and the right superior corona radiata, and significant negative correlations between sleep duration and MD in right orbito-frontal white matter and the right inferior longitudinal fasciculus. Improved sleep quality was positively correlated with FA in left caudate nucleus, white matter tracts to the left orbito-frontal region, the left anterior cingulum bundle and the white matter tracts associated with the right operculum and insula, and negatively correlated with MD in left orbito-frontal white matter and the left anterior cingulum bundle. Our findings suggest that reduced cumulative total sleep time (cTST) and poorer subjective sleep quality are associated with subtle white matter micro-architectural changes. The regions we identified as being related to habitual sleep patterns were restricted to the frontal and temporal lobes, and the functions they support are consistent with those which have previously been demonstrated as being affected by short sleep durations (e.g., attention, cognitive control, memory). Examining how inter-individual differences in brain structure are related to habitual sleep patterns could help to shed light on the mechanisms by which sleep habits are associated with brain function, behaviour and cognition, as well as potentially the networks and systems responsible for variations in sleep patterns themselves.

Variability in Cumulative Habitual Sleep Duration Predicts Waking Functional Connectivity.

STUDY OBJECTIVES: We examined whether interindividual differences in habitual sleep patterns, quantified as the cumulative habitual total sleep time (cTST) over a 2-w period, were reflected in waking measurements of intranetwork and internetwork functional connectivity (FC) between major nodes of three intrinsically connected networks (ICNs): default mode network (DMN), salience network (SN), and central executive network (CEN). METHODS: Resting state functional magnetic resonance imaging (fMRI) study using seed-based FC analysis combined with 14-d wrist actigraphy, sleep diaries, and subjective questionnaires (N = 33 healthy adults, mean age 34.3, standard deviation ± 11.6 y). Data were statistically analyzed using multiple linear regression. Fourteen consecutive days of wrist actigraphy in participant's home environment and fMRI scanning on day 14 at the Birmingham University Imaging Centre. Seed-based FC analysis on ICNs from resting-state fMRI data and multiple linear regression analysis performed for each ICN seed and target. cTST was used to predict FC (controlling for age). RESULTS: cTST was specific predictor of intranetwork FC when the mesial prefrontal cortex (MPFC) region of the DMN was used as a seed for FC, with a positive correlation between FC and cTST observed. No significant relationship between FC and cTST was seen for any pair of nodes not including the MPFC. Internetwork FC between the DMN (MPFC) and SN (right anterior insula) was also predicted by cTST, with a negative correlation observed between FC and cTST. CONCLUSIONS: This study improves understanding of the relationship between intranetwork and internetwork functional connectivity of intrinsically connected networks (ICNs) in relation to habitual sleep quality and duration. The cumulative amount of sleep that participants achieved over a 14-d period was significantly predictive of intranetwork and inter-network functional connectivity of ICNs, an observation that may underlie the link between sleep status and cognitive performance.

Multimodal neuroimaging investigations of alterations to consciousness: the relationship between absence epilepsy and sleep.

The link between epilepsy and sleep is well established on many levels. The focus of the current review is on recent neuroimaging investigations into the alterations of consciousness that are observed during absence seizures and the descent into sleep. Functional neuroimaging provides simultaneous cortical and subcortical recording of activity throughout the brain, allowing a detailed definition and characterization of large-scale brain networks and the interactions between them. This has led to the identification of a set of regions which collectively form the consciousness system, which includes contributions from the default mode network (DMN), ascending arousal systems, and the thalamus. Electrophysiological and neuroimaging investigations have also clearly demonstrated the importance of thalamocortical and corticothalamic networks in the evolution of sleep and absence epilepsy, two phenomena in which the subject experiences an alteration to the conscious state and a disconnection from external input. However, the precise relationship between the consciousness system, thalamocortical networks, and consciousness itself remains to be clarified. One of the fundamental challenges is to understand how distributed brain networks coordinate their activity in order to maintain and implement complex behaviors such as consciousness and how modifications to this network activity lead to alterations in consciousness. By taking into account not only the level of activation of individual brain regions but also their connectivity within specific networks and the activity and connectivity of other relevant networks, a more specific quantification of brain states can be achieved. This, in turn, may provide a more fundamental understanding of the alterations to consciousness experienced in sleep and epilepsy.

The structural and functional connectivity of the posterior cingulate cortex: comparison between deterministic and probabilistic tractography for the investigation of structure-function relationships.

The default mode network (DMN) is one of the most studied resting-state networks, and is thought to be involved in the maintenance of consciousness within the alert human brain. Although many studies have examined the functional connectivity (FC) of the DMN, few have investigated its underlying structural connectivity (SC), or the relationship between the two. We investigated this question in fifteen healthy subjects, concentrating on connections to the precuneus/posterior cingulate cortex (PCC), commonly considered as the central node of the DMN. We used group independent component analysis (GICA) and seed-based correlation analysis of fMRI data to quantify FC, and streamline and probabilistic tractography to identify structural tracts from diffusion tensor imaging (DTI) data. We first assessed the presence of structural connections between the DMN regions identified with GICA. Of the 15 subjects, when using the probabilistic approach 15 (15) demonstrated connections between the PCC and mesial prefrontal cortex (mPFC), 11 (15) showed connections from the PCC to the right inferior parietal cortex (rIPC) and 8 (15) to the left IPC. Next, we assessed the strength of FC (magnitude of temporal correlation) and SC (mean fractional anisotropy of reconstructed tracts (streamline), number of super-threshold voxels within the mask region (probabilistic)). The lIPC had significantly reduced FC to the PCC compared to the mPFC and rIPC. No difference in SC strength between connections was found using the streamline approach. For the probabilistic approach, mPFC had significantly lower SC than both IPCs. The two measures of SC strength were significantly correlated, but not for all paired connections. Finally, we observed a significant correlation between SC and FC for both tractography approaches when data were pooled across PCC-lIPL, PCC-rIPL and PCC-mPFC connections, and for some individual paired connections. Our results suggest that the streamline approach is advantageous for characterising the connectivity of long white matter tracts (PCC-mPFC), whilst the probabilistic approach was more reliable at identifying PCC-IPC connections. The direct comparison of FC and SC indicated that pairs of nodes with stronger structural connections also had stronger functional connectivity, and that this was maintained with both tractography approaches. Whilst the definition of SC strength remains controversial, our results could be considered to provide some degree of validation for the measures of SC strength that we have used. Direct comparisons of SC and FC are necessary in order to understand the structural basis of functional connectivity, and to characterise and quantify the changes in the brain's functional architecture that occur as a result of normal physiology or pathology.