Gut enterochromaffin cells drive visceral pain and anxiety.

Gastrointestinal (GI) discomfort is a hallmark of most gut disorders and represents an important component of chronic visceral pain1. For the growing population afflicted by irritable bowel syndrome, GI hypersensitivity and pain persist long after tissue injury has resolved2. Irritable bowel syndrome also exhibits a strong sex bias, afflicting women three times more than men1. Here, we focus on enterochromaffin (EC) cells, which are rare excitable, serotonergic neuroendocrine cells in the gut epithelium3-5. EC cells detect and transduce noxious stimuli to nearby mucosal nerve endings3,6 but involvement of this signalling pathway in visceral pain and attendant sex differences has not been assessed. By enhancing or suppressing EC cell function in vivo, we show that these cells are sufficient to elicit hypersensitivity to gut distension and necessary for the sensitizing actions of isovalerate, a bacterial short-chain fatty acid associated with GI inflammation7,8. Remarkably, prolonged EC cell activation produced persistent visceral hypersensitivity, even in the absence of an instigating inflammatory episode. Furthermore, perturbing EC cell activity promoted anxiety-like behaviours which normalized after blockade of serotonergic signalling. Sex differences were noted across a range of paradigms, indicating that the EC cell-mucosal afferent circuit is tonically engaged in females. Our findings validate a critical role for EC cell-mucosal afferent signalling in acute and persistent GI pain, in addition to highlighting genetic models for studying visceral hypersensitivity and the sex bias of gut pain.

Differences in functional connectivity distribution after transcranial direct-current stimulation: A connectivity density point of view.

In this manuscript, we consider the problem of relating functional connectivity measurements viewed as statistical distributions to outcomes. We demonstrate the utility of using the distribution of connectivity on a study of resting-state functional magnetic resonance imaging association with an intervention. The method uses the estimated density of connectivity between nodes of interest as a functional covariate. Moreover, we demonstrate the utility of the procedure in an instance where connectivity is naturally considered an outcome by reversing the predictor/response relationship using case/control methodology. The method utilizes the density quantile, the density evaluated at empirical quantiles, instead of the empirical density directly. This improved the performance of the method by highlighting tail behavior, though we emphasize that by being flexible and non-parametric, the technique can detect effects related to the central portion of the density. To demonstrate the method in an application, we consider 47 primary progressive aphasia patients with various levels of language abilities. These patients were randomly assigned to two treatment arms, transcranial direct-current stimulation and language therapy versus sham (language therapy only), in a clinical trial. We use the method to analyze the effect of direct stimulation on functional connectivity. As such, we estimate the density of correlations among the regions of interest and study the difference in the density post-intervention between treatment arms. We discover that it is the tail of the density, rather than the mean or lower order moments of the distribution, that demonstrates a significant impact in the classification. The new approach has several benefits. Among them, it drastically reduces the number of multiple comparisons compared with edge-wise analysis. In addition, it allows for the investigation of the impact of functional connectivity on the outcomes where the connectivity is not geometrically localized.

Mechanosensory Stimulation via Nanchung Expressing Neurons Can Induce Daytime Sleep in Drosophila.

The neuronal and genetic bases of sleep, a phenomenon considered crucial for well-being of organisms, has been under investigation using the model organism Drosophila melanogaster Although sleep is a state where sensory threshold for arousal is greater, it is known that certain kinds of repetitive sensory stimuli, such as rocking, can indeed promote sleep in humans. Here we report that orbital motion-aided mechanosensory stimulation promotes sleep of male and female Drosophila, independent of the circadian clock, but controlled by the homeostatic system. Mechanosensory receptor nanchung (Nan)-expressing neurons in the chordotonal organs mediate this sleep induction: flies in which these neurons are either silenced or ablated display significantly reduced sleep induction on mechanosensory stimulation. Transient activation of the Nan-expressing neurons also enhances sleep levels, confirming the role of these neurons in sleep induction. We also reveal that certain regions of the antennal mechanosensory and motor center in the brain are involved in conveying information from the mechanosensory structures to the sleep centers. Thus, we show, for the first time, that a circadian clock-independent pathway originating from peripherally distributed mechanosensors can promote daytime sleep of flies Drosophila melanogasterSIGNIFICANCE STATEMENT Our tendency to fall asleep in moving vehicles or the practice of rocking infants to sleep suggests that slow rhythmic movement can induce sleep, although we do not understand the mechanistic basis of this phenomenon. We find that gentle orbital motion can induce behavioral quiescence even in flies, a highly genetically tractable system for sleep studies. We demonstrate that this is indeed true sleep based on its rapid reversibility by sensory stimulation, enhanced arousal threshold, and homeostatic control. Furthermore, we demonstrate that mechanosensory neurons expressing a TRPV channel nanchung, located in the antennae and chordotonal organs, mediate orbital motion-induced sleep by communicating with antennal mechanosensory motor centers, which in turn may project to sleep centers in the brain.

Modulation of fear generalization by the zona incerta.

Fear expressed toward threat-associated stimuli is an adaptive behavioral response. In contrast, the generalization of fear responses toward nonthreatening cues is a maladaptive and debilitating dimension of trauma- and anxiety-related disorders. Expressing fear to appropriate stimuli and suppressing fear generalization require integration of relevant sensory information and motor output. While thalamic and subthalamic brain regions play important roles in sensorimotor integration, very little is known about the contribution of these regions to the phenomenon of fear generalization. In this study, we sought to determine whether fear generalization could be modulated by the zona incerta (ZI), a subthalamic brain region that influences sensory discrimination, defensive responses, and retrieval of fear memories. To do so, we combined differential intensity-based auditory fear conditioning protocols in mice with C-FOS immunohistochemistry and designer receptors exclusively activated by designer drugs (DREADDs)-based manipulation of neuronal activity in the ZI. C-FOS immunohistochemistry revealed an inverse relationship between ZI activation and fear generalization: The ZI was less active in animals that generalized fear. In agreement with this relationship, chemogenetic inhibition of the ZI resulted in fear generalization, while chemogenetic activation of the ZI suppressed fear generalization. Furthermore, targeted stimulation of GABAergic cells in the ZI reduced fear generalization. To conclude, our data suggest that stimulation of the ZI could be used to treat fear generalization in the context of trauma- and anxiety-related disorders.

A generative-discriminative framework that integrates imaging, genetic, and diagnosis into coupled low dimensional space.

We propose a novel optimization framework that integrates imaging and genetics data for simultaneous biomarker identification and disease classification. The generative component of our model uses a dictionary learning framework to project the imaging and genetic data into a shared low dimensional space. We have coupled both the data modalities by tying the linear projection coefficients to the same latent space. The discriminative component of our model uses logistic regression on the projection vectors for disease diagnosis. This prediction task implicitly guides our framework to find interpretable biomarkers that are substantially different between a healthy and disease population. We exploit the interconnectedness of different brain regions by incorporating a graph regularization penalty into the joint objective function. We also use a group sparsity penalty to find a representative set of genetic basis vectors that span a low dimensional space where subjects are easily separable between patients and controls. We have evaluated our model on a population study of schizophrenia that includes two task fMRI paradigms and single nucleotide polymorphism (SNP) data. Using ten-fold cross validation, we compare our generative-discriminative framework with canonical correlation analysis (CCA) of imaging and genetics data, parallel independent component analysis (pICA) of imaging and genetics data, random forest (RF) classification, and a linear support vector machine (SVM). We also quantify the reproducibility of the imaging and genetics biomarkers via subsampling. Our framework achieves higher class prediction accuracy and identifies robust biomarkers. Moreover, the implicated brain regions and genetic variants underlie the well documented deficits in schizophrenia.

Determining Thresholds for Three Indices of Autoregulation to Identify the Lower Limit of Autoregulation During Cardiac Surgery.

OBJECTIVES: Monitoring cerebral autoregulation may help identify the lower limit of autoregulation in individual patients. Mean arterial blood pressure below lower limit of autoregulation appears to be a risk factor for postoperative acute kidney injury. Cerebral autoregulation can be monitored in real time using correlation approaches. However, the precise thresholds for different cerebral autoregulation indexes that identify the lower limit of autoregulation are unknown. We identified thresholds for intact autoregulation in patients during cardiopulmonary bypass surgery and examined the relevance of these thresholds to postoperative acute kidney injury. DESIGN: A single-center retrospective analysis. SETTING: Tertiary academic medical center. PATIENTS: Data from 59 patients was used to determine precise cerebral autoregulation thresholds for identification of the lower limit of autoregulation. These thresholds were validated in a larger cohort of 226 patients. METHODS AND MAIN RESULTS: Invasive mean arterial blood pressure, cerebral blood flow velocities, regional cortical oxygen saturation, and total hemoglobin were recorded simultaneously. Three cerebral autoregulation indices were calculated, including mean flow index, cerebral oximetry index, and hemoglobin volume index. Cerebral autoregulation curves for the three indices were plotted, and thresholds for each index were used to generate threshold- and index-specific lower limit of autoregulations. A reference lower limit of autoregulation could be identified in 59 patients by plotting cerebral blood flow velocity against mean arterial blood pressure to generate gold-standard Lassen curves. The lower limit of autoregulations defined at each threshold were compared with the gold-standard lower limit of autoregulation determined from Lassen curves. The results identified the following thresholds: mean flow index (0.45), cerebral oximetry index (0.35), and hemoglobin volume index (0.3). We then calculated the product of magnitude and duration of mean arterial blood pressure less than lower limit of autoregulation in a larger cohort of 226 patients. When using the lower limit of autoregulations identified by the optimal thresholds above, mean arterial blood pressure less than lower limit of autoregulation was greater in patients with acute kidney injury than in those without acute kidney injury. CONCLUSIONS: This study identified thresholds of intact and impaired cerebral autoregulation for three indices and showed that mean arterial blood pressure below lower limit of autoregulation is a risk factor for acute kidney injury after cardiac surgery.

PACAP increases Arc/Arg 3.1 expression within the extended amygdala after fear conditioning in rats.

The stress-related neuropeptide pituitary adenylate cyclase-activating polypeptide (PACAP) is implicated in neuromodulation of learning and memory. PACAP can alter synaptic plasticity and has direct actions on neurons in the amygdala and hippocampus that could contribute to its acute and persistent effects on the consolidation and expression of conditioned fear. We recently demonstrated that intracerebroventricular (ICV) infusion of PACAP prior to fear conditioning (FC) results in initial amnestic-like effects followed by hyper-expression of conditioned freezing with repeated testing, and analyses of immediate-early gene c-Fos expression suggested that the central nucleus of the amygdala (CeA), but not the lateral/basolateral amygdala (LA/BLA) or hippocampus, are involved in these PACAP effects. Here, we extend that work by examining the expression of the synaptic plasticity marker activity-regulated cytoskeleton-associated protein (Arc/Arg 3.1) after PACAP administration and FC. Male Sprague-Dawley rats were implanted with cannula for ICV infusion of PACAP-38 (1.5 µg) or vehicle followed by FC and tests for conditioned freezing. One hour after FC, Arc protein expression was significantly elevated in the CeA and bed nucleus of the stria terminalis (BNST), interconnected structures that are key elements of the extended amygdala, in rats that received the combination of PACAP + FC. In contrast, Arc expression within the subdivisions of the hippocampus, or the LA/BLA, were unchanged. A subpopulation of Arc-positive cells in both the CeA and BNST also express PKCdelta, an intracellular marker that has been used to identify microcircuits that gate conditioned fear in the CeA. Consistent with our previous findings, on the following day conditioned freezing behavior was reduced in rats that had been given the combination of PACAP + FC-an amnestic-like effect-and Arc expression levels had returned to baseline. Given the established role of Arc in modifying synaptic plasticity and memory formation, our findings suggest that PACAP-induced overexpression of Arc following fear conditioning may disrupt neuroplastic changes within populations of CeA and BNST neurons normally responsible for encoding fear-related cues that, in this case, results in altered fear memory consolidation. Hence, PACAP systems may represent an axis on which stress and experience-driven neurotransmission converge to alter emotional memory, and mediate pathologies that are characteristic of psychiatric illnesses such as post-traumatic stress disorder.

Identifying disease foci from static and dynamic effective connectivity networks: Illustration in soldiers with trauma.

Brain connectivity studies report group differences in pairwise connection strengths. While informative, such results are difficult to interpret since our understanding of the brain relies on region-based properties, rather than on connection information. Given that large disruptions in the brain are often caused by a few pivotal sources, we propose a novel framework to identify the sources of functional disruption from effective connectivity networks. Our approach integrates static and time-varying effective connectivity modeling in a probabilistic framework, to identify aberrant foci and the corresponding aberrant connectomics network. Using resting-state fMRI, we illustrate the utility of this novel approach in U.S. Army soldiers (N = 87) with posttraumatic stress disorder (PTSD), mild traumatic brain injury (mTBI) and combat controls. Additionally, we employed machine-learning classification to identify those significant connectivity features that possessed high predictive ability. We identified three disrupted foci (middle frontal gyrus [MFG], insula, hippocampus), and an aberrant prefrontal-subcortical-parietal network of information flow. We found the MFG to be the pivotal focus of network disruption, with aberrant strength and temporal-variability of effective connectivity to the insula, amygdala and hippocampus. These connectivities also possessed high predictive ability (giving a classification accuracy of 81%); and they exhibited significant associations with symptom severity and neurocognitive functioning. In summary, dysregulation originating in the MFG caused elevated and temporally less-variable connectivity in subcortical regions, followed by a similar effect on parietal memory-related regions. This mechanism likely contributes to the reduced control over traumatic memories leading to re-experiencing, hyperarousal and flashbacks observed in soldiers with PTSD and mTBI. Hum Brain Mapp 39:264-287, 2018. © 2017 Wiley Periodicals, Inc.

Or47b-neurons promote male-mating success in Drosophila.

Drosophila performs elaborate well-defined rituals of courtship, which involve several types of sensory inputs. Here, we report that Or47b-neurons promote male-mating success. Males with Or47b-neurons silenced/ablated exhibit reduced copulation frequency and increased copulation latency. Copulation latency of Or47b-manipulated flies increased proportionately with size of the assay arena, whereas in controls it remained unchanged. While competing for mates, Or47b-ablated males are outperformed by intact controls. These results suggest the role of Or47b-neurons in promoting male-mating success.

Rheumatoid Arthritis and Periodontal Disease. An Update.

A review of the epidemiological, pathological and immunological relationships between two chronic inflammatory diseases: rheumatoid arthritis (RA) and periodontal disease (PD). RA is a chronic inflammatory disease of the joints, characterized by loss of connective tissue and mineralized structures, the so-called "synovial membrane." Periodontitis is the inflammatory destruction of the periodontal attachment and alveolar bone. While the etiology of these two diseases may differ, the underlying pathogenic mechanisms are similar. And it is possible that individuals manifesting both PD and RA may suffer from a unifying underlying systemic deregulation of the inflammatory response. There is an overproduction of a variety of cytokines and MMPs that appears to be common in both diseases. Oral health parameters should be more closely monitored in patients with RA, an autoimmune disease. Data suggest that periodontal therapies combined with routine RA treatments further improve RA status. Interventions to prevent, minimize or treat periodontitis in arthritis patients will definitely promise a better quality of life for these patients.

Expanding the canon: An inclusive neurobiology of thalamic and subthalamic fear circuits.

Appropriate expression of fear in the face of threats in the environment is essential for survival. The sustained expression of fear in the absence of threat signals is a central pathological feature of trauma- and anxiety-related disorders. Our understanding of the neural circuitry that controls fear inhibition coalesces around the amygdala, hippocampus, and prefrontal cortex. By discussing thalamic and sub-thalamic influences on fear-related learning and expression in this review, we suggest a more inclusive neurobiological framework that expands our canonical view of fear. First, we visit how fear-related learning and expression is influenced by the aforementioned canonical brain regions. Next, we review emerging data that shed light on new roles for thalamic and subthalamic nuclei in fear-related learning and expression. Then, we highlight how these neuroanatomical hubs can modulate fear via integration of sensory and salient stimuli, gating information flow and calibrating behavioral responses, as well as maintaining and updating memory representations. Finally, we propose that the presence of this thalamic and sub-thalamic neuroanatomy in parallel with the tripartite prefrontal cortex-amygdala-hippocampus circuit allows for dynamic modulation of information based on interoceptive and exteroceptive signals. This article is part of the Special Issue on "Fear, Anxiety and PTSD".

mSPD-NN: A Geometrically Aware Neural Framework for Biomarker Discovery from Functional Connectomics Manifolds.

Connectomics has emerged as a powerful tool in neuroimaging and has spurred recent advancements in statistical and machine learning methods for connectivity data. Despite connectomes inhabiting a matrix manifold, most analytical frameworks ignore the underlying data geometry. This is largely because simple operations, such as mean estimation, do not have easily computable closed-form solutions. We propose a geometrically aware neural framework for connectomes, i.e., the mSPD-NN, designed to estimate the geodesic mean of a collections of symmetric positive definite (SPD) matrices. The mSPD-NN is comprised of bilinear fully connected layers with tied weights and utilizes a novel loss function to optimize the matrix-normal equation arising from Fréchet mean estimation. Via experiments on synthetic data, we demonstrate the efficacy of our mSPD-NN against common alternatives for SPD mean estimation, providing competitive performance in terms of scalability and robustness to noise. We illustrate the real-world flexibility of the mSPD-NN in multiple experiments on rs-fMRI data and demonstrate that it uncovers stable biomarkers associated with subtle network differences among patients with ADHD-ASD comorbidities and healthy controls.

GIRUS-net: A Multimodal Deep Learning Model Identifying Imaging and Genetic Biomarkers Linked to Alzheimer's Disease Severity.

We introduce an explainable deep neural architecture that combines brain structure with genetic influence to improve disease severity prediction in Alzheimer's disease. Our framework consists of an encoder, a decoder, and a rank-consistent ordinal regression module. The encoder projects neural imaging and genetics data into a low-dimensional latent space regularized by the decoder. The ordinal regression module guides the feature embedding process to find discriminative patterns representative of disease severity. We also add a learnable dropout layer that learns feature importance and extracts explainable biomarkers from the data. We evaluate our model using structural MRI (sMRI) and Single Nucleotide Polymorphism (SNP) data provided by the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. In 2-class severity classification comparison, our model has a median F-score of 0.86 (baseline median F-score range: 0.57-0.81). In 3-class classification comparison, our model's median F-score is 0.50 (baseline range: 0.17 - 0.41). In 4-class classification comparison, our model's median F-score is 0.40 (baseline range: 0.14 - 0.39). We demonstrate that our model provides improved disease diagnosis alongside sparse and clinically relevant biomarkers.Clinical relevance-This study provides a deep-learning model that can predict Alzheimer's disease severity levels while identifying consistent and clinically relevant biomarkers.

DeepEZ: A Graph Convolutional Network for Automated Epileptogenic Zone Localization From Resting-State fMRI Connectivity.

OBJECTIVE: Epileptogenic zone (EZ) localization is a crucial step during diagnostic work up and therapeutic planning in medication refractory epilepsy. In this paper, we present the first deep learning approach to localize the EZ based on resting-state fMRI (rs-fMRI) data. METHODS: Our network, called DeepEZ, uses a cascade of graph convolutions that emphasize signal propagation along expected anatomical pathways. We also integrate domain-specific information, such as an asymmetry term on the predicted EZ and a learned subject-specific bias to mitigate environmental confounds. RESULTS: We validate DeepEZ on rs-fMRI collected from 14 patients with focal epilepsy at the University of Wisconsin Madison. Using cross validation, we demonstrate that DeepEZ achieves consistently high EZ localization performance (Accuracy: 0.88 ± 0.03; AUC: 0.73 ± 0.03) that far outstripped any of the baseline methods. This performance is notable given the variability in EZ locations and scanner type across the cohort. CONCLUSION: Our results highlight the promise of using DeepEZ as an accurate and noninvasive therapeutic planning tool for medication refractory epilepsy. SIGNIFICANCE: While prior work in EZ localization focused on identifying localized aberrant signatures, there is growing evidence that epileptic seizures affect inter-regional connectivity in the brain. DeepEZ allows clinicians to harness this information from noninvasive imaging that can easily be integrated into the existing clinical workflow.

BEATRICE: Bayesian Fine-mapping from Summary Data using Deep Variational Inference.

We introduce a novel framework BEATRICE to identify putative causal variants from GWAS summary statistics (https://github.com/sayangsep/Beatrice-Finemapping). Identifying causal variants is challenging due to their sparsity and to highly correlated variants in the nearby regions. To account for these challenges, our approach relies on a hierarchical Bayesian model that imposes a binary concrete prior on the set of causal variants. We derive a variational algorithm for this fine-mapping problem by minimizing the KL divergence between an approximate density and the posterior probability distribution of the causal configurations. Correspondingly, we use a deep neural network as an inference machine to estimate the parameters of our proposal distribution. Our stochastic optimization procedure allows us to simultaneously sample from the space of causal configurations. We use these samples to compute the posterior inclusion probabilities and determine credible sets for each causal variant. We conduct a detailed simulation study to quantify the performance of our framework across different numbers of causal variants and different noise paradigms, as defined by the relative genetic contributions of causal and non-causal variants. Using this simulated data, we perform a comparative analysis against two state-of-the-art baseline methods for fine-mapping. We demonstrate that BEATRICE achieves uniformly better coverage with comparable power and set sizes, and that the performance gain increases with the number of causal variants. Thus, BEATRICE is a valuable tool to identify causal variants from eQTL and GWAS summary statistics across complex diseases and traits.

EPViz: A flexible and lightweight visualizer to facilitate predictive modeling for multi-channel EEG.

Scalp Electroencephalography (EEG) is one of the most popular noninvasive modalities for studying real-time neural phenomena. While traditional EEG studies have focused on identifying group-level statistical effects, the rise of machine learning has prompted a shift in computational neuroscience towards spatio-temporal predictive analyses. We introduce a novel open-source viewer, the EEG Prediction Visualizer (EPViz), to aid researchers in developing, validating, and reporting their predictive modeling outputs. EPViz is a lightweight and standalone software package developed in Python. Beyond viewing and manipulating the EEG data, EPViz allows researchers to load a PyTorch deep learning model, apply it to EEG features, and overlay the output channel-wise or subject-level temporal predictions on top of the original time series. These results can be saved as high-resolution images for use in manuscripts and presentations. EPViz also provides valuable tools for clinician-scientists, including spectrum visualization, computation of basic data statistics, and annotation editing. Finally, we have included a built-in EDF anonymization module to facilitate sharing of clinical data. Taken together, EPViz fills a much needed gap in EEG visualization. Our user-friendly interface and rich collection of features may also help to promote collaboration between engineers and clinicians.

Prediction of lactate concentrations after cardiac surgery using machine learning and deep learning approaches.

Background: Although conventional prediction models for surgical patients often ignore intraoperative time-series data, deep learning approaches are well-suited to incorporate time-varying and non-linear data with complex interactions. Blood lactate concentration is one important clinical marker that can reflect the adequacy of systemic perfusion during cardiac surgery. During cardiac surgery and cardiopulmonary bypass, minute-level data is available on key parameters that affect perfusion. The goal of this study was to use machine learning and deep learning approaches to predict maximum blood lactate concentrations after cardiac surgery. We hypothesized that models using minute-level intraoperative data as inputs would have the best predictive performance. Methods: Adults who underwent cardiac surgery with cardiopulmonary bypass were eligible. The primary outcome was maximum lactate concentration within 24 h postoperatively. We considered three classes of predictive models, using the performance metric of mean absolute error across testing folds: (1) static models using baseline preoperative variables, (2) augmentation of the static models with intraoperative statistics, and (3) a dynamic approach that integrates preoperative variables with intraoperative time series data. Results: 2,187 patients were included. For three models that only used baseline characteristics (linear regression, random forest, artificial neural network) to predict maximum postoperative lactate concentration, the prediction error ranged from a median of 2.52 mmol/L (IQR 2.46, 2.56) to 2.58 mmol/L (IQR 2.54, 2.60). The inclusion of intraoperative summary statistics (including intraoperative lactate concentration) improved model performance, with the prediction error ranging from a median of 2.09 mmol/L (IQR 2.04, 2.14) to 2.12 mmol/L (IQR 2.06, 2.16). For two modelling approaches (recurrent neural network, transformer) that can utilize intraoperative time-series data, the lowest prediction error was obtained with a range of median 1.96 mmol/L (IQR 1.87, 2.05) to 1.97 mmol/L (IQR 1.92, 2.05). Intraoperative lactate concentration was the most important predictive feature based on Shapley additive values. Anemia and weight were also important predictors, but there was heterogeneity in the importance of other features. Conclusion: Postoperative lactate concentrations can be predicted using baseline and intraoperative data with moderate accuracy. These results reflect the value of intraoperative data in the prediction of clinically relevant outcomes to guide perioperative management.

Automated seizure activity tracking and onset zone localization from scalp EEG using deep neural networks.

We propose a novel neural network architecture, SZTrack, to detect and track the spatio-temporal propagation of seizure activity in multichannel EEG. SZTrack combines a convolutional neural network encoder operating on individual EEG channels with recurrent neural networks to capture the evolution of seizure activity. Our unique training strategy aggregates individual electrode level predictions for patient-level seizure detection and localization. We evaluate SZTrack on a clinical EEG dataset of 201 seizure recordings from 34 epilepsy patients acquired at the Johns Hopkins Hospital. Our network achieves similar seizure detection performance to state-of-the-art methods and provides valuable localization information that has not previously been demonstrated in the literature. We also show the cross-site generalization capabilities of SZTrack on a dataset of 53 seizure recordings from 14 epilepsy patients acquired at the University of Wisconsin Madison. SZTrack is able to determine the lobe and hemisphere of origin in nearly all of these new patients without retraining the network. To our knowledge, SZTrack is the first end-to-end seizure tracking network using scalp EEG.

Incerto-thalamic modulation of fear via GABA and dopamine.

Fear generalization and deficits in extinction learning are debilitating dimensions of Post-Traumatic Stress Disorder (PTSD). Most understanding of the neurobiology underlying these dimensions comes from studies of cortical and limbic brain regions. While thalamic and subthalamic regions have been implicated in modulating fear, the potential for incerto-thalamic pathways to suppress fear generalization and rescue deficits in extinction recall remains unexplored. We first used patch-clamp electrophysiology to examine functional connections between the subthalamic zona incerta and thalamic reuniens (RE). Optogenetic stimulation of GABAergic ZI → RE cell terminals in vitro induced inhibitory post-synaptic currents (IPSCs) in the RE. We then combined high-intensity discriminative auditory fear conditioning with cell-type-specific and projection-specific optogenetics in mice to assess functional roles of GABAergic ZI → RE cell projections in modulating fear generalization and extinction recall. In addition, we used a similar approach to test the possibility of fear generalization and extinction recall being modulated by a smaller subset of GABAergic ZI → RE cells, the A13 dopaminergic cell population. Optogenetic stimulation of GABAergic ZI → RE cell terminals attenuated fear generalization and enhanced extinction recall. In contrast, optogenetic stimulation of dopaminergic ZI → RE cell terminals had no effect on fear generalization but enhanced extinction recall in a dopamine receptor D1-dependent manner. Our findings shed new light on the neuroanatomy and neurochemistry of ZI-located cells that contribute to adaptive fear by increasing the precision and extinction of learned associations. In so doing, these data reveal novel neuroanatomical substrates that could be therapeutically targeted for treatment of PTSD.

Automated eloquent cortex localization in brain tumor patients using multi-task graph neural networks.

Localizing the eloquent cortex is a crucial part of presurgical planning. While invasive mapping is the gold standard, there is increasing interest in using noninvasive fMRI to shorten and improve the process. However, many surgical patients cannot adequately perform task-based fMRI protocols. Resting-state fMRI has emerged as an alternative modality, but automated eloquent cortex localization remains an open challenge. In this paper, we develop a novel deep learning architecture to simultaneously identify language and primary motor cortex from rs-fMRI connectivity. Our approach uses the representational power of convolutional neural networks alongside the generalization power of multi-task learning to find a shared representation between the eloquent subnetworks. We validate our method on data from the publicly available Human Connectome Project and on a brain tumor dataset acquired at the Johns Hopkins Hospital. We compare our method against feature-based machine learning approaches and a fully-connected deep learning model that does not account for the shared network organization of the data. Our model achieves significantly better performance than competing baselines. We also assess the generalizability and robustness of our method. Our results clearly demonstrate the advantages of our graph convolution architecture combined with multi-task learning and highlight the promise of using rs-fMRI as a presurgical mapping tool.