Beyond network connectivity: A classification approach to brain age prediction with resting-state fMRI.

The brain is a complex, dynamic organ that shows differences in the same subject at various periods. Understanding how brain activity changes across age as a function of the brain networks has been greatly abetted by fMRI. Canonical analysis consists of determining how alterations in connectivity patterns (CPs) of certain regions are affected. An alternative approach is taken here by not considering connectivity but rather features computed from recordings at the regions of interest (ROIs). Using machine learning (ML) we assess how neural signals are altered by and prospectively predictive of age and sex via a methodology that is novel in drawing upon pairwise classification across six decades of subjects' chronological ages. ML is used to answer the equally important questions of what properties of the computed features are most predictive as well as which brain networks are most affected by aging. It was found that there is decreased differentiation among the neural signals of older subjects that are separated in age by the same number of years as younger subjects. Furthermore, the burstiness of the signals change at different rates between males and females. The findings provide insight into brain aging via an ROI-based analysis, the consideration of several feature groups, and a novel classification-based ML pipeline. There is also a contribution to understanding the effects of data aggregated from different recording centers on the conclusions of fMRI studies.

Brain age monotonicity and functional connectivity differences of healthy subjects.

Alterations in the brain's connectivity or the interactions among brain regions have been studied with the aid of resting state (rs)fMRI data attained from large numbers of healthy subjects of various demographics. This has been instrumental in providing insight into how a phenotype as fundamental as age affects the brain. Although machine learning (ML) techniques have already been deployed in such studies, novel questions are investigated in this work. We study whether young brains develop properties that progressively resemble those of aged brains, and if the aging dynamics of older brains provide information about the aging trajectory in young subjects. The degree of a prospective monotonic relationship will be quantified, and hypotheses of brain aging trajectories will be tested via ML. Furthermore, the degree of functional connectivity across the age spectrum of three datasets will be compared at a population level and across sexes. The findings scrutinize similarities and differences among the male and female subjects at greater detail than previously performed.

Machine Learning Identifies a Rat Model of Parkinson's Disease via Sleep-Wake Electroencephalogram.

Alpha-synuclein induced degeneration of the midbrain substantia nigra pars compact (SNc) dopaminergic neurons causes Parkinson's disease (PD). Rodent studies demonstrate that nigrostriatal dopamine stimulates pallidal neurons which, via the topographical pallidocortical pathway, regulate cortical activity and functions. We hypothesize that nigrostriatal dopamine acting at the basal ganglia regulates cortical activity in sleep and wake state, and its depletion systemically alters electroencephalogram (EEG) across frequencies during sleep-wake state. Compared to control rats, 6-hydroxydopamine induced selective SNc lesions increased overall EEG power (positive synchronization) across 0.5-60 Hz during wake, NREM (non-rapid eye movement) sleep, and REM sleep. Application of machine learning (ML) to seven EEG features computed at a single or combined spectral bands during sleep-wake differentiated SNc lesions from controls at high accuracy. ML algorithms construct a model based on empirical data to make predictions on subsequent data. The accuracy of the predictive results indicate that nigrostriatal dopamine depletion increases global EEG spectral synchronization in wake, NREM sleep, and REM sleep. The EEG changes can be exploited by ML to identify SNc lesions at a high accuracy.

Identifying Neural Signatures of Dopamine Signaling with Machine Learning.

The emergence of new tools to image neurotransmitters, neuromodulators, and neuropeptides has transformed our understanding of the role of neurochemistry in brain development and cognition, yet analysis of this new dimension of neurobiological information remains challenging. Here, we image dopamine modulation in striatal brain tissue slices with near infrared catecholamine nanosensors (nIRCat) and implement machine learning to determine which features of dopamine modulation are unique to changes in stimulation strength, and to different neuroanatomical regions. We trained a support vector machine and a random forest classifier to determine whether recordings were made from the dorsolateral striatum (DLS) versus the dorsomedial striatum (DMS) and find that machine learning is able to accurately distinguish dopamine release that occurs in DLS from that occurring in DMS in a manner unachievable with canonical statistical analysis. Furthermore, our analysis determines that dopamine modulatory signals including the number of unique dopamine release sites and peak dopamine released per stimulation event are most predictive of neuroanatomy yet note that integrated neuromodulator amount is the conventional metric currently used to monitor neuromodulation in animal studies. Lastly, our study finds that machine learning discrimination of different stimulation strengths or neuroanatomical regions is only possible in adult animals, suggesting a high degree of variability in dopamine modulatory kinetics during animal development. Our study highlights that machine learning could become a broadly-utilized tool to differentiate between neuroanatomical regions, or between neurotypical and disease states, with features not detectable by conventional statistical analysis.

Identifying Neural Signatures of Dopamine Signaling with Machine Learning.

The emergence of new tools to image neurotransmitters, neuromodulators, and neuropeptides has transformed our understanding of the role of neurochemistry in brain development and cognition, yet analysis of this new dimension of neurobiological information remains challenging. Here, we image dopamine modulation in striatal brain tissue slices with near-infrared catecholamine nanosensors (nIRCat) and implement machine learning to determine which features of dopamine modulation are unique to changes in stimulation strength, and to different neuroanatomical regions. We trained a support vector machine and a random forest classifier to decide whether the recordings were made from the dorsolateral striatum (DLS) versus the dorsomedial striatum (DMS) and find that machine learning is able to accurately distinguish dopamine release that occurs in DLS from that occurring in DMS in a manner unachievable with canonical statistical analysis. Furthermore, our analysis determines that dopamine modulatory signals including the number of unique dopamine release sites and peak dopamine released per stimulation event are most predictive of neuroanatomy. This is in light of integrated neuromodulator amount being the conventional metric used to monitor neuromodulation in animal studies. Lastly, our study finds that machine learning discrimination of different stimulation strengths or neuroanatomical regions is only possible in adult animals, suggesting a high degree of variability in dopamine modulatory kinetics during animal development. Our study highlights that machine learning could become a broadly utilized tool to differentiate between neuroanatomical regions or between neurotypical and disease states, with features not detectable by conventional statistical analysis.

Transcriptomic profiles of stress susceptibility and resilience in the amygdala and hippocampus.

A single, severe episode of stress can bring about myriad responses amongst individuals, ranging from cognitive enhancement to debilitating and persistent anxiety; however, the biological mechanisms that contribute to resilience versus susceptibility to stress are poorly understood. The dentate gyrus (DG) of the hippocampus and the basolateral nucleus of the amygdala (BLA) are key limbic regions that are susceptible to the neural and hormonal effects of stress. Previous work has also shown that these regions contribute to individual variability in stress responses; however, the molecular mechanisms underlying the role of these regions in susceptibility and resilience are unknown. In this study, we profiled the transcriptomic signatures of the DG and BLA of rats with divergent behavioral outcomes after a single, severe stressor. We subjected rats to three hours of immobilization with exposure to fox urine and conducted a behavioral battery one week after stress to identify animals that showed persistent, high anxiety-like behavior. We then conducted bulk RNA sequencing of the DG and BLA from susceptible, resilient, and unexposed control rats. Differential gene expression analyses revealed that the molecular signatures separating each of the three groups were distinct and non-overlapping between the DG and BLA. In the amygdala, key genes associated with insulin and hormonal signaling corresponded with vulnerability. Specifically, Inhbb, Rab31 , and Ncoa3 were upregulated in the amygdala of stress-susceptible animals compared to resilient animals. In the hippocampus, increased expression of Cartpt - which encodes a key neuropeptide involved in reward, reinforcement, and stress responses - was strongly correlated with vulnerability to anxiety-like behavior. However, few other genes distinguished stress-susceptible animals from control animals, while a larger number of genes separated stress-resilient animals from control and stress-susceptible animals. Of these, Rnf112, Tbx19 , and UBALD1 distinguished resilient animals from both control and susceptible animals and were downregulated in resilience, suggesting that an active molecular response in the hippocampus facilitates protection from the long-term consequences of severe stress. These results provide novel insight into the mechanisms that bring about individual variability in the behavioral responses to stress and provide new targets for the advancement of therapies for stress-induced neuropsychiatric disorders.

Quantitative and Correlational Analysis of Brain and Spleen Immune Cellular Responses Following Cerebral Ischemia.

Stroke is a multiphasic process, and the initial ischemic phase of neuronal damage is followed by secondary innate and adaptive responses that unfold over days after stroke, offer a longer time frame of intervention, and represent a novel therapeutic target. Therefore, revealing the distinct functions of immune cells in both brain and periphery is important for identification of immunotherapeutic targets for stroke to extend the treatment time window. In this paper an examination of the cellular dynamics of the immune response in the central nervous system (CNS) and periphery provoked by cerebral ischemia is provided. New data is presented for the number of immune cells in brain and spleen of mice during the 7 days following middle cerebral artery occlusion (MCAO). A novel analysis of the correlation among various cell types in the brain and spleen following stroke is presented. It is found that the infiltrated macrophages in the ischemic hemisphere positively correlate with neutrophils which implies their synergic effect in migrating into the brain after stroke onset. It is noted that during infiltration of adaptive immune cells, the number of neutrophils correlate positively with T cells, which suggests neutrophils contribute to T cell infiltration in the stroked brain. Furthermore, the correlation among neurological deficit and various immune cells suggests that microglia and splenic adaptive immune cells (T and B cells) are protective while infiltrating peripheral myeloid cells (macrophage and neutrophils) worsen stroke outcome. Comprehension of such immune responses post cerebral ischemia is crucial for differentiating the drivers of outcomes and also predicting the stroke outcome.

Object Recognition at Higher Regions of the Ventral Visual Stream via Dynamic Inference.

The ventral visual stream (VVS) is a fundamental pathway involved in visual object identification and recognition. In this work, we present a hypothesis of a sequence of computations performed by the VVS during object recognition. The operations performed by the inferior temporal (IT) cortex are represented as not being akin to a neural-network, but rather in-line with a dynamic inference instantiation of the untangling notion. The presentation draws upon a technique for dynamic maximum a posteriori probability (MAP) sequence estimation based on the Viterbi algorithm. Simulation results are presented to show that the decoding portion of the architecture that is associated with the IT can effectively untangle object identity when presented with synthetic data. More importantly, we take a step forward in visual neuroscience by presenting a framework for an inference-based approach that is biologically inspired via attributes implicated in primate object recognition. The analysis will provide insight in explaining the exceptional proficiency of the VVS.

A Layered Control Architecture of Sleep and Arousal.

Sleep and wakefulness are promoted not by a single neural pathway but via wake or sleep-promoting nodes distributed across layers of the brain. We equate each layer with a brain region in proposing a layered subsumption model for arousal based on a computational architecture. Beyond the brainstem the layers include the diencephalon (hypothalamus, thalamus), basal ganglia, and cortex. In light of existing empirical evidence, we propose that each layer have sleep and wake computations driven by similar high-level architecture and processing units. Specifically, an interconnected wake-promoting system is suggested as driving arousal in each brain layer with the processing converging to produce the features of wakefulness. In contrast, sleep-promoting GABAergic neurons largely project to and inhibit wake-promoting neurons. We propose a general pattern of caudal wake-promoting and sleep-promoting neurons having a strong effect on overall behavior. However, while rostral brain layers have less influence on sleep and wake, through descending projections, they can subsume the activity of caudal brain layers to promote arousal. The two models presented in this work will suggest computations for the layering and hierarchy. Through dynamic system theory several hypotheses are introduced for the interaction of controllers and systems that correspond to the different populations of neurons at each layer. The models will be drawn-upon to discuss future experiments to elucidate the structure of the hierarchy that exists among the sleep-arousal architecture.

Probabilistic Modeling of Pseudorabies Virus Infection in a Neural Circuit.

Viral transneuronal tracing methods effectively label synaptically connected neurons in a time-dependent manner. However, the modeling of viral vectors has been largely absent. An objective of this article is to motivate and initiate a basis for computational modeling of viral labeling and the questions that can be investigated through modeling of pseudorabies virus (PRV) virion progression in a neural circuit. In particular, a mathematical model is developed for quantitative analysis of PRV infection. Probability expressions are presented to evaluate the progression of viral labeling along the neural circuit. The analysis brings forth various parameters, the numerical values of which must be attained through future experiments. This is the first computational model for PRV viral labeling of a neural circuit.

Deconstruction of the beaten Path-Sidestep interaction network provides insights into neuromuscular system development.

An 'interactome' screen of all Drosophila cell-surface and secreted proteins containing immunoglobulin superfamily (IgSF) domains discovered a network formed by paralogs of Beaten Path (Beat) and Sidestep (Side), a ligand-receptor pair that is central to motor axon guidance. Here we describe a new method for interactome screening, the Bio-Plex Interactome Assay (BPIA), which allows identification of many interactions in a single sample. Using the BPIA, we 'deorphanized' four more members of the Beat-Side network. We confirmed interactions using surface plasmon resonance. The expression patterns of beat and side genes suggest that Beats are neuronal receptors for Sides expressed on peripheral tissues. side-VI is expressed in muscle fibers targeted by the ISNb nerve, as well as at growth cone choice points and synaptic targets for the ISN and TN nerves. beat-V genes, encoding Side-VI receptors, are expressed in ISNb and ISN motor neurons.