Monitoring the Burden of Seizures and Highly Epileptiform Patterns in Critical Care with a Novel Machine Learning Method.

INTRODUCTION: Current electroencephalography (EEG) practice relies on interpretation by expert neurologists, which introduces diagnostic and therapeutic delays that can impact patients' clinical outcomes. As EEG practice expands, these experts are becoming increasingly limited resources. A highly sensitive and specific automated seizure detection system would streamline practice and expedite appropriate management for patients with possible nonconvulsive seizures. We aimed to test the performance of a recently FDA-cleared machine learning method (Claritγ, Ceribell Inc.) that measures the burden of seizure activity in real time and generates bedside alerts for possible status epilepticus (SE). METHODS: We retrospectively identified adult patients (n = 353) who underwent evaluation of possible seizures with Rapid Response EEG system (Rapid-EEG, Ceribell Inc.). Automated detection of seizure activity and seizure burden throughout a recording (calculated as the percentage of ten-second epochs with seizure activity in any 5-min EEG segment) was performed with Claritγ, and various thresholds of seizure burden were tested (≥ 10% indicating ≥ 30 s of seizure activity in the last 5 min, ≥ 50% indicating ≥ 2.5 min of seizure activity, and ≥ 90% indicating ≥ 4.5 min of seizure activity and triggering a SE alert). The sensitivity and specificity of Claritγ's real-time seizure burden measurements and SE alerts were compared to the majority consensus of at least two expert neurologists. RESULTS: Majority consensus of neurologists labeled the 353 EEGs as normal or slow activity (n = 249), highly epileptiform patterns (HEP, n = 87), or seizures [n = 17, nine longer than 5 min (e.g., SE), and eight shorter than 5 min]. The algorithm generated a SE alert (≥ 90% seizure burden) with 100% sensitivity and 93% specificity. The sensitivity and specificity of various thresholds for seizure burden during EEG recordings for detecting patients with seizures were 100% and 82% for ≥ 50% seizure burden and 88% and 60% for ≥ 10% seizure burden. Of the 179 EEG recordings in which the algorithm detected no seizures, seizures were identified by the expert reviewers in only two cases, indicating a negative predictive value of 99%. DISCUSSION: Claritγ detected SE events with high sensitivity and specificity, and it demonstrated a high negative predictive value for distinguishing nonepileptiform activity from seizure and highly epileptiform activity. CONCLUSIONS: Ruling out seizures accurately in a large proportion of cases can help prevent unnecessary or aggressive over-treatment in critical care settings, where empiric treatment with antiseizure medications is currently prevalent. Claritγ's high sensitivity for SE and high negative predictive value for cases without epileptiform activity make it a useful tool for triaging treatment and the need for urgent neurological consultation.

A longitudinal footprint of genetic epilepsies using automated electronic medical record interpretation.

PURPOSE: Childhood epilepsies have a strong genetic contribution, but the disease trajectory for many genetic etiologies remains unknown. Electronic medical record (EMR) data potentially allow for the analysis of longitudinal clinical information but this has not yet been explored. METHODS: We analyzed provider-entered neurological diagnoses made at 62,104 patient encounters from 658 individuals with known or presumed genetic epilepsies. To harmonize clinical terminology, we mapped clinical descriptors to Human Phenotype Ontology (HPO) terms and inferred higher-level phenotypic concepts. We then binned the resulting 286,085 HPO terms to 100 3-month time intervals and assessed gene-phenotype associations at each interval. RESULTS: We analyzed a median follow-up of 6.9 years per patient and a cumulative 3251 patient years. Correcting for multiple testing, we identified significant associations between "Status epilepticus" with SCN1A at 1.0 years, "Severe intellectual disability" with PURA at 9.75 years, and "Infantile spasms" and "Epileptic spasms" with STXBP1 at 0.5 years. The identified associations reflect known clinical features of these conditions, and manual chart review excluded provider bias. CONCLUSION: Some aspects of the longitudinal disease histories can be reconstructed through EMR data and reveal significant gene-phenotype associations, even within closely related conditions. Gene-specific EMR footprints may enable outcome studies and clinical decision support.

Correction: A longitudinal footprint of genetic epilepsies using automated electronic medical record interpretation.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Harnessing electronic medical records to advance research on multiple sclerosis.

BACKGROUND: Electronic medical records (EMR) data are increasingly used in research, but no studies have yet evaluated similarity between EMR and research-quality data and between characteristics of an EMR multiple sclerosis (MS) population and known natural MS history. OBJECTIVES: To (1) identify MS patients in an EMR system and extract clinical data, (2) compare EMR-extracted data with gold-standard research data, and (3) compare EMR MS population characteristics to expected MS natural history. METHODS: Algorithms were implemented to identify MS patients from the University of California San Francisco EMR, de-identify the data and extract clinical variables. EMR-extracted data were compared to research cohort data in a subset of patients. RESULTS: We identified 4142 MS patients via search of the EMR and extracted their clinical data with good accuracy. EMR and research values showed good concordance for Expanded Disability Status Scale (EDSS), timed-25-foot walk, and subtype. We replicated several expected MS epidemiological features from MS natural history including higher EDSS for progressive versus relapsing-remitting patients and for male versus female patients and increased EDSS with age at examination and disease duration. CONCLUSION: Large real-world cohorts algorithmically extracted from the EMR can expand opportunities for MS clinical research.

Big data in status epilepticus.

Status epilepticus care and treatment are already being touched by the revolution in data science. New approaches designed to leverage the tremendous potential of "big data" in the clinical sphere are enabling researchers and clinicians to extract information from sources such as administrative claims data, the electronic medical health record, and continuous physiologic monitoring data streams. Algorithmic methods of data extraction also offer potential to fuse multimodal data (including text-based documentation, imaging data, and time-series data) to improve patient assessment and stratification beyond the manual capabilities of individual physicians. Still, the potential of data science to impact the diagnosis, treatment, and minute-to-minute care of patients with status epilepticus is only starting to be appreciated. In this brief review, we discuss how data science is impacting the field and draw examples from the following three main areas: (1) analysis of insurance claims from large administrative datasets to evaluate the impact of continuous electroencephalogram (EEG) monitoring on clinical outcomes; (2) natural language processing of the electronic health record to find, classify, and stratify patients for prognostication and treatment; and (3) real-time systems for data analysis, data reduction, and multimodal data fusion to guide therapy in real time. While early, it is our hope that these examples will stimulate investigators to leverage data science, computer science, and engineering methods to improve the care and outcome of patients with status epilepticus and other neurological disorders. This article is part of the Special Issue "Proceedings of the 7th London-Innsbruck Colloquium on Status Epilepticus and Acute Seizures".

An ImmunoChip study of multiple sclerosis risk in African Americans.

The aims of this study were: (i) to determine to what degree multiple sclerosis-associated loci discovered in European populations also influence susceptibility in African Americans; (ii) to assess the extent to which the unique linkage disequilibrium patterns in African Americans can contribute to localizing the functionally relevant regions or genes; and (iii) to search for novel African American multiple sclerosis-associated loci. Using the ImmunoChip custom array we genotyped 803 African American cases with multiple sclerosis and 1516 African American control subjects at 130 135 autosomal single nucleotide polymorphisms. We conducted association analysis with rigorous adjustments for population stratification and admixture. Of the 110 non-major histocompatibility complex multiple sclerosis-associated variants identified in Europeans, 96 passed stringent quality control in our African American data set and of these, >70% (69) showed over-representation of the same allele amongst cases, including 21 with nominally significant evidence for association (one-tailed test P < 0.05). At a further eight loci we found nominally significant association with an alternate correlated risk-tagging single nucleotide polymorphism from the same region. Outside the regions known to be associated in Europeans, we found seven potentially associated novel candidate multiple sclerosis variants (P < 10(-4)), one of which (rs2702180) also showed nominally significant evidence for association (one-tailed test P = 0.034) in an independent second cohort of 620 African American cases and 1565 control subjects. However, none of these novel associations reached genome-wide significance (combined P = 6.3 × 10(-5)). Our data demonstrate substantial overlap between African American and European multiple sclerosis variants, indicating common genetic contributions to multiple sclerosis risk.

Genome, epigenome and RNA sequences of monozygotic twins discordant for multiple sclerosis.

Monozygotic or 'identical' twins have been widely studied to dissect the relative contributions of genetics and environment in human diseases. In multiple sclerosis (MS), an autoimmune demyelinating disease and common cause of neurodegeneration and disability in young adults, disease discordance in monozygotic twins has been interpreted to indicate environmental importance in its pathogenesis. However, genetic and epigenetic differences between monozygotic twins have been described, challenging the accepted experimental model in disambiguating the effects of nature and nurture. Here we report the genome sequences of one MS-discordant monozygotic twin pair, and messenger RNA transcriptome and epigenome sequences of CD4(+) lymphocytes from three MS-discordant, monozygotic twin pairs. No reproducible differences were detected between co-twins among approximately 3.6 million single nucleotide polymorphisms (SNPs) or approximately 0.2 million insertion-deletion polymorphisms. Nor were any reproducible differences observed between siblings of the three twin pairs in HLA haplotypes, confirmed MS-susceptibility SNPs, copy number variations, mRNA and genomic SNP and insertion-deletion genotypes, or the expression of approximately 19,000 genes in CD4(+) T cells. Only 2 to 176 differences in the methylation of approximately 2 million CpG dinucleotides were detected between siblings of the three twin pairs, in contrast to approximately 800 methylation differences between T cells of unrelated individuals and several thousand differences between tissues or between normal and cancerous tissues. In the first systematic effort to estimate sequence variation among monozygotic co-twins, we did not find evidence for genetic, epigenetic or transcriptome differences that explained disease discordance. These are the first, to our knowledge, female, twin and autoimmune disease individual genome sequences reported.

Haplotype-based approach to known MS-associated regions increases the amount of explained risk.

Genome-wide association studies (GWAS), using single nucleotide polymorphisms (SNPs), have yielded 110 non-human leucocyte antigen genomic regions that are associated with multiple sclerosis (MS). Despite this large number of associations, however, only 28% of MS-heritability can currently be explained. Here we compare the use of multi-SNP-haplotypes to the use of single-SNPs as alternative methods to describe MS genetic risk. SNP-haplotypes (of various lengths from 1 up to 15 contiguous SNPs) were constructed at each of the 110 previously identified, MS-associated, genomic regions. Even after correcting for the larger number of statistical comparisons made when using the haplotype-method, in 32 of the regions, the SNP-haplotype based model was markedly more significant than the single-SNP based model. By contrast, in no region was the single-SNP based model similarly more significant than the SNP-haplotype based model. Moreover, when we included the 932 MS-associated SNP-haplotypes (that we identified from 102 regions) as independent variables into a logistic linear model, the amount of MS-heritability, as assessed by Nagelkerke's R-squared, was 38%, which was considerably better than 29%, which was obtained by using only single-SNPs. This study demonstrates that SNP-haplotypes can be used to fine-map the genetic associations within regions of interest previously identified by single-SNP GWAS. Moreover, the amount of the MS genetic risk explained by the SNP-haplotype associations in the 110 MS-associated genomic regions was considerably greater when using SNP-haplotypes than when using single-SNPs. Also, the use of SNP-haplotypes can lead to the discovery of new regions of interest, which have not been identified by a single-SNP GWAS.

Blood RNA profiling in a large cohort of multiple sclerosis patients and healthy controls.

Multiple sclerosis (MS) is the most common autoimmune disease of the central nervous system (CNS). It is characterized by the infiltration of autoreactive immune cells into the CNS, which target the myelin sheath, leading to the loss of neuronal function. Although it is accepted that MS is a multifactorial disorder with both genetic and environmental factors influencing its development and course, the molecular pathogenesis of MS has not yet been fully elucidated. Here, we studied the longitudinal gene expression profiles of whole-blood RNA from a cohort of 195 MS patients and 66 healthy controls. We analyzed these transcriptomes at both the individual transcript and the biological pathway level. We found 62 transcripts to be significantly up-regulated in MS patients; the expression of 11 of these genes was counter-regulated by interferon treatment, suggesting partial restoration of a 'healthy' gene expression profile. Global pathway analyses linked the proteasome and Wnt signaling to MS disease processes. Since genotypes from a subset of individuals were available, we were able to identify expression quantitative trait loci (eQTL), a number of which involved two genes of the MS gene signature. However, all these eQTL were also present in healthy controls. This study highlights the challenge posed by analyzing transcripts from whole blood and how these can be mitigated by using large, well-characterized cohorts of patients with longitudinal follow-up and multi-modality measurements.

Genetic contribution to multiple sclerosis risk among Ashkenazi Jews.

BACKGROUND: Multiple sclerosis (MS) is an autoimmune disease of the central nervous system, with a strong genetic component. Over 100 genetic loci have been implicated in susceptibility to MS in European populations, the most prominent being the 15:01 allele of the HLA-DRB1 gene. The prevalence of MS is high in European populations including those of Ashkenazi origin, and low in African and Asian populations including those of Jewish origin. METHODS: Here we identified and extracted a total of 213 Ashkenazi MS cases and 546 ethnically matched healthy control individuals from two previous genome-wide case-control association analyses, and 72 trios (affected proband and two unaffected parents) from a previous genome-wide transmission disequilibrium association study, using genetic data to define Ashkenazi. We compared the pattern of genetic risk between Ashkenazi and non-Ashkenazi Europeans. We also sought to identify novel Ashkenazi-specific risk loci by performing association tests on the subset of Ashkenazi cases, controls, probands, and parents from each study. RESULTS: The HLA-DRB1*15:01 allele and the non-HLA risk alleles were present at relatively low frequencies among Ashkenazi and explained a smaller fraction of the population-level risk when compared to non-Ashkenazi Europeans. Alternative HLA susceptibility alleles were identified in an Ashkenazi-only association study, including HLA-A*68:02 and one or both genes in the HLA-B*38:01-HLA-C*12:03 haplotype. The genome-wide screen in Ashkenazi did not reveal any loci associated with MS risk. CONCLUSION: These results suggest that genetic susceptibility to MS in Ashkenazi Jews has not been as well established as that of non-Ashkenazi Europeans. This implies value in studying large well-characterized Ashkenazi populations to accelerate gene discovery in complex genetic diseases.

A genome-wide association study of brain lesion distribution in multiple sclerosis.

Brain magnetic resonance imaging is widely used as a diagnostic and monitoring tool in multiple sclerosis and provides a non-invasive, sensitive and reproducible way to track the disease. Topological characteristics relating to the distribution and shape of lesions are recognized as important neuroradiological markers in the diagnosis of multiple sclerosis, although these have been much less well characterized quantitatively than have traditional measures such as T2 hyperintense or T1 hypointense lesion volumes. Here, we used voxel-level 3 T magnetic resonance imaging T1-weighted scans to reconstruct the 3D topology of lesions in 284 subjects with multiple sclerosis and tested whether this is a heritable phenotype. To this end, we extracted the genotypes from a published genome-wide association study on these same individuals and searched for genetic associations with lesion load, shape and topological distribution. Lesion probability maps were created to identify frequently affected areas and to assess the overall distribution of T1 lesions in the subject population as a whole. We then developed an original algorithm to cluster adjacent lesional voxels (cluxels) in each subject and tested whether cluxel topology was significantly associated with any single-nucleotide polymorphism in our data set. To focus on patterns of lesion distribution, we computed the first 10 principal components. Although principal component 1 correlated with lesion load, none of the remaining orthogonal components correlated with any other known variable. We then conducted genome-wide association studies on each of these and found 31 significant associations (false discovery rate <0.01) with principal component 8, which represents a mode of variation of lesion topology in the population. The majority of the loci can be linked to genes related to immune cell function and to myelin and neural growth; some (SYK, MYT1L, TRAPPC9, SLITKR6 and RIC3) have been previously associated with the distribution of white matter lesions in multiple sclerosis. Finally, we used a bioinformatics approach to identify a network of 48 interacting proteins showing genetic associations (P < 0.01) with cluxel topology in multiple sclerosis. This network also contains proteins expressed in immune cells and is enriched in molecules expressed in the central nervous system that contribute to neural development and regeneration. Our results show how quantitative traits derived from brain magnetic resonance images of patients with multiple sclerosis can be used as dependent variables in a genome-wide association study. With the widespread availability of powerful computing and the availability of genotyped populations, integration of imaging and genetic data sets is likely to become a mainstream tool for understanding the complex biological processes of multiple sclerosis and other brain disorders.

Patients recovering from abdominal surgery who walked with volunteers had improved postoperative recovery profiles during their hospitalization.

BACKGROUND: Early walking as part of a perioperative care program benefits patients who have had surgery. However, the impact of early walking by itself on the mental and physical recovery of postoperative patients has not been examined. METHODS: We established a program called walking to recovery (WTR) in which college volunteers provided walking assistance to patients recovering after abdominal surgery. Patients who participated in the program were compared with patients who did not. The postoperative recovery profile survey (PRP-17) was administered on day of discharge to 15 participants and 15 non-participants. Medical records were reviewed to obtain indication for surgery, type of surgery, length of hospital stay, and postoperative complications. At 1 month post-discharge, a short form (SF)-12v2 questionnaire was administered by telephone to assess postoperative quality of life as defined by mental and physical level of function and measured with the mental component score (MCS) and the physical component score (PCS). RESULTS: The average age of participants and non-participants was similar (48.9 ± 9.8 vs. 51.4 ± 8.7 years; p = 0.28). When the two groups were approximately matched by type and severity of surgery, participants had lower PRP-17 composite scores (9.9 vs. 12.5, p = 0.003) and higher indicator sums (9.8 vs. 8.4, p = 0.04) than non-participants, both of which indicate better postoperative recovery in participants. The mean immobilization score was significantly lower in participants (0.3 vs. 0.8, p = 0.04). Postoperative length of stay and MCS did not differ between the two groups, but in participants there was a trend for higher scores in the PCS. CONCLUSIONS: Walking with volunteers was associated with a better PRP during the hospitalization period but not at 1 month follow-up. The WTR program is a sustainable, cost-effective model program for other hospitals to emulate as part of the standard of care of postoperative patients.

Sequencing of the IL6 gene in a case-control study of cerebral palsy in children.

BACKGROUND: Cerebral palsy (CP) is a group of nonprogressive disorders of movement and posture caused by abnormal development of, or damage to, motor control centers of the brain. A single nucleotide polymorphism (SNP), rs1800795, in the promoter region of the interleukin-6 (IL6) gene has been implicated in the pathogenesis of CP by mediating IL-6 protein levels in amniotic fluid and cord plasma and within brain lesions. This SNP has been associated with other neurological, vascular, and malignant processes as well, often as part of a haplotype block. METHODS: To refine the regional genetic association with CP, we sequenced (Sanger) the IL6 gene and part of the promoter region in 250 infants with CP and 305 controls. RESULTS: We identified a haplotype of 7 SNPs that includes rs1800795. In a recessive model of inheritance, the variant haplotype conferred greater risk (OR = 4.3, CI = [2.0-10.1], p = 0.00007) than did the lone variant at rs1800795 (OR = 2.5, CI = [1.4-4.6], p = 0.002). The risk haplotype contains one SNP (rs2069845, CI = [1.2-4.3], OR = 2.3, p = 0.009) that disrupts a methylation site. CONCLUSIONS: The risk haplotype identified in this study overlaps with previously identified haplotypes that include additional promoter SNPs. A risk haplotype at the IL6 gene likely confers risk to CP, and perhaps other diseases, via a multi-factorial mechanism.

Multiple sclerosis: Exploring the limits and implications of genetic and environmental susceptibility.

OBJECTIVE: To explore and describe the basis and implications of genetic and environmental susceptibility to multiple sclerosis (MS) using the Canadian population-based data. BACKGROUND: Certain parameters of MS-epidemiology are directly observable (e.g., the recurrence-risk of MS in siblings and twins, the proportion of women among MS patients, the population-prevalence of MS, and the time-dependent changes in the sex-ratio). By contrast, other parameters can only be inferred from the observed parameters (e.g., the proportion of the population that is "genetically susceptible", the proportion of women among susceptible individuals, the probability that a susceptible individual will experience an environment "sufficient" to cause MS, and if they do, the probability that they will develop the disease). DESIGN/METHODS: The "genetically susceptible" subset (G) of the population (Z) is defined to include everyone with any non-zero life-time chance of developing MS under some environmental conditions. The value for each observed and non-observed epidemiological parameter is assigned a "plausible" range. Using both a Cross-sectional Model and a Longitudinal Model, together with established parameter relationships, we explore, iteratively, trillions of potential parameter combinations and determine those combinations (i.e., solutions) that fall within the acceptable range for both the observed and non-observed parameters. RESULTS: Both Models and all analyses intersect and converge to demonstrate that probability of genetic-susceptibitly, P(G), is limited to only a fraction of the population {i.e., P(G) ≤ 0.52)} and an even smaller fraction of women {i.e., P(G│F) < 0.32)}. Consequently, most individuals (particularly women) have no chance whatsoever of developing MS, regardless of their environmental exposure. However, for any susceptible individual to develop MS, requires that they also experience a "sufficient" environment. We use the Canadian data to derive, separately, the exponential response-curves for men and women that relate the increasing likelihood of developing MS to an increasing probability that a susceptible individual experiences an environment "sufficient" to cause MS. As the probability of a "sufficient" exposure increases, we define, separately, the limiting probability of developing MS in men (c) and women (d). These Canadian data strongly suggest that: (c < d ≤ 1). If so, this observation establishes both that there must be a "truly" random factor involved in MS pathogenesis and that it is this difference, rather than any difference in genetic or environmental factors, which primarily accounts for the penetrance difference between women and men. CONCLUSIONS: The development of MS (in an individual) requires both that they have an appropriate genotype (which is uncommon in the population) and that they have an environmental exposure "sufficient" to cause MS given their genotype. Nevertheless, the two principal findings of this study are that: P(G) ≤ 0.52)} and: (c < d ≤ 1). Threfore, even when the necessary genetic and environmental factors, "sufficient" for MS pathogenesis, co-occur for an individual, they still may or may not develop MS. Consequently, disease pathogenesis, even in this circumstance, seems to involve an important element of chance. Moreover, the conclusion that the macroscopic process of disease development for MS includes a "truly" random element, if replicated (either for MS or for other complex diseases), provides empiric evidence that our universe is non-deterministic.

In depth comparison of an individual's DNA and its lymphoblastoid cell line using whole genome sequencing.

BACKGROUND: A detailed analysis of whole genomes can be now achieved with next generation sequencing. Epstein Barr Virus (EBV) transformation is a widely used strategy in clinical research to obtain an unlimited source of a subject's DNA. Although the mechanism of transformation and immortalization by EBV is relatively well known at the transcriptional and proteomic level, the genetic consequences of EBV transformation are less well understood. A detailed analysis of the genetic alterations introduced by EBV transformation is highly relevant, as it will inform on the usefulness and limitations of this approach. RESULTS: We used whole genome sequencing to assess the genomic signature of a low-passage lymphoblastoid cell line (LCL). Specifically, we sequenced the full genome (40X) of an individual using DNA purified from fresh whole blood as well as DNA from his LCL. A total of 217.33 Gb of sequence were generated from the cell line and 238.95 Gb from the normal genomic DNA. We determined with high confidence that 99.2% of the genomes were identical, with no reproducible changes in structural variation (chromosomal rearrangements and copy number variations) or insertion/deletion polymorphisms (indels). CONCLUSIONS: Our results suggest that, at this level of resolution, the LCL is genetically indistinguishable from its genomic counterpart and therefore their use in clinical research is not likely to introduce a significant bias.

Combined VNS-RNS Neuromodulation for Epilepsy.

SUMMARY: The vagus nerve stimulator (VNS) and responsive nerve stimulator (RNS) are nonpharmacological devices approved for drug-resistant epilepsy. Vagus nerve stimulator was removed before placing an RNS in clinical trials. Two cases of bilateral mesial temporal epilepsy treated concurrently with VNS and bilateral mesial temporal RNS devices were reported. In each case, the VNS device was turned off temporarily, which allowed for a direct comparison of RNS recordings and efficacy with and without simultaneous VNS stimulation. Temporary VNS cessation lead to increased clinical and electrocorticographic seizures despite continued anti-seizure drugs and RNS stimulation. In one case, VNS eliminated seizures from one epileptogenic area, whereas VNS and RNS were required to treat seizures from the contralateral mesial temporal structure. In another case, VNS effectively decreased seizure spread to the symptomatogenic zone. These cases demonstrate synergistic neuromodulation with concurrent use of VNS and RNS in intractable bitemporal epilepsy.

Adverse Childhood Experiences in Patients With Neurologic Disease.

Background and Objectives: To describe the prevalence of high adverse childhood experiences (ACEs) among neurology outpatients and determine their association with health care utilization rates and comorbid medical and psychiatric disease. Methods: This was a cross-sectional study of adults seen for outpatient neurology follow-up at the University of Pennsylvania. Participants completed the ACE questionnaire and depression/anxiety screenings. Health care utilization metrics (emergency department [ED] visits, hospitalizations, and outpatient calls) were obtained for all participants. High ACE scores were defined as a score of ≥4. The prevalence of high ACE scores in our cohort was compared with US historical controls. Statistical associations were adjusted for age, sex, and race/ethnicity. Results: One hundred ninety-eight patients were enrolled in the study. Neurology patients were more likely to have elevated ACE scores compared with US population estimates (23.7% vs 12.6%, p < 0.01). High ACE scores were associated with increased ED utilization (odds ratio [OR] = 21, 95% CI [5.8-76.0], p < 0.01), hospitalizations (OR = 5.2, 95% CI [1.7-15.0], p < 0.01), and telephone encounters (OR 3, 95% CI [1.1-8.2], p < 0.05). High ACEs were also associated with medical and psychiatric comorbidities (OR 5.8, 95% CI [2.0-17.0], p < 0.01 and OR 4.5, 95% CI [2.1-9.6], p < 0.01) and high depression and anxiety scores (OR = 6.9, 95% CI [2.8-17.0], p < 0.01, and OR = 4.3, [95% CI 1.7-11.0], p < 0.01). Discussion: Patients with neurologic conditions are more likely to have high ACEs than the US population, which was associated with higher rates of health care utilization, increased number of medical and psychiatric comorbidities, and higher anxiety and depression scores. Addressing ACEs may be a way to improve the health outcomes of patients with neurologic conditions.

Development of a natural language processing algorithm to extract seizure types and frequencies from the electronic health record.

OBJECTIVE: To develop a natural language processing (NLP) algorithm to abstract seizure types and frequencies from electronic health records (EHR). BACKGROUND: Seizure frequency measurement is an epilepsy quality metric. Yet, abstraction of seizure frequency from the EHR is laborious. We present an NLP algorithm to extract seizure data from unstructured text of clinic notes. Algorithm performance was assessed at two epilepsy centers. METHODS: We developed a rules-based NLP algorithm to recognize terms related to seizures and frequency within the text of an outpatient encounter. Algorithm output (e.g. number of seizures of a particular type within a time interval) was compared to seizure data manually annotated by two expert reviewers ("gold standard"). The algorithm was developed from 150 clinic notes from institution #1 (development set), then tested on a separate set of 219 notes from institution #1 (internal test set) with 248 unique seizure frequency elements. The algorithm was separately applied to 100 notes from institution #2 (external test set) with 124 unique seizure frequency elements. Algorithm performance was measured by recall (sensitivity), precision (positive predictive value), and F1 score (geometric mean of precision and recall). RESULTS: In the internal test set, the algorithm demonstrated 70% recall (173/248), 95% precision (173/182), and 0.82 F1 score compared to manual review. Algorithm performance in the external test set was lower with 22% recall (27/124), 73% precision (27/37), and 0.40 F1 score. CONCLUSIONS: These results suggest NLP extraction of seizure types and frequencies is feasible, though not without challenges in generalizability for large-scale implementation.

iCTNet: a Cytoscape plugin to produce and analyze integrative complex traits networks.

BACKGROUND: The speed at which biological datasets are being accumulated stands in contrast to our ability to integrate them meaningfully. Large-scale biological databases containing datasets of genes, proteins, cells, organs, and diseases are being created but they are not connected. Integration of these vast but heterogeneous sources of information will allow the systematic and comprehensive analysis of molecular and clinical datasets, spanning hundreds of dimensions and thousands of individuals. This integration is essential to capitalize on the value of current and future molecular- and cellular-level data on humans to gain novel insights about health and disease. RESULTS: We describe a new open-source Cytoscape plugin named iCTNet (integrated Complex Traits Networks). iCTNet integrates several data sources to allow automated and systematic creation of networks with up to five layers of omics information: phenotype-SNP association, protein-protein interaction, disease-tissue, tissue-gene, and drug-gene relationships. It facilitates the generation of general or specific network views with diverse options for more than 200 diseases. Built-in tools are provided to prioritize candidate genes and create modules of specific phenotypes. CONCLUSIONS: iCTNet provides a user-friendly interface to search, integrate, visualize, and analyze genome-scale biological networks for human complex traits. We argue this tool is a key instrument that facilitates systematic integration of disparate large-scale data through network visualization, ultimately allowing the identification of disease similarities and the design of novel therapeutic approaches.The online database and Cytoscape plugin are freely available for academic use at: http://www.cs.queensu.ca/ictnet.

Pathway and network-based analysis of genome-wide association studies in multiple sclerosis.

Genome-wide association studies (GWAS) testing several hundred thousand SNPs have been performed in multiple sclerosis (MS) and other complex diseases. Typically, the number of markers in which the evidence for association exceeds the genome-wide significance threshold is very small, and markers that do not exceed this threshold are generally neglected. Classical statistical analysis of these datasets in MS revealed genes with known immunological functions. However, many of the markers showing modest association may represent false negatives. We hypothesize that certain combinations of genes flagged by these markers can be identified if they belong to a common biological pathway. Here we conduct a pathway-oriented analysis of two GWAS in MS that takes into account all SNPs with nominal evidence of association (P < 0.05). Gene-wise P-values were superimposed on a human protein interaction network and searches were conducted to identify sub-networks containing a higher proportion of genes associated with MS than expected by chance. These sub-networks, and others generated at random as a control, were categorized for membership of biological pathways. GWAS from eight other diseases were analyzed to assess the specificity of the pathways identified. In the MS datasets, we identified sub-networks of genes from several immunological pathways including cell adhesion, communication and signaling. Remarkably, neural pathways, namely axon-guidance and synaptic potentiation, were also over-represented in MS. In addition to the immunological pathways previously identified, we report here for the first time the potential involvement of neural pathways in MS susceptibility.