scholarly journals Neural Circuits for Dynamics-Based Segmentation of Time Series

2022 ◽  
pp. 1-49
Author(s):  
Tiberiu Teşileanu ◽  
Siavash Golkar ◽  
Samaneh Nasiri ◽  
Anirvan M. Sengupta ◽  
Dmitri B. Chklovskii
Keyword(s):  
Time Series ◽  
Latent Variables ◽  
Prediction Error ◽  
Ground Truth ◽  
Brain Regions ◽  
Autoregressive Models ◽  
Data Sets ◽  
Model Free ◽  
Segmentation Task ◽  
The Brain

Abstract The brain must extract behaviorally relevant latent variables from the signals streamed by the sensory organs. Such latent variables are often encoded in the dynamics that generated the signal rather than in the specific realization of the waveform. Therefore, one problem faced by the brain is to segment time series based on underlying dynamics. We present two algorithms for performing this segmentation task that are biologically plausible, which we define as acting in a streaming setting and all learning rules being local. One algorithm is model based and can be derived from an optimization problem involving a mixture of autoregressive processes. This algorithm relies on feedback in the form of a prediction error and can also be used for forecasting future samples. In some brain regions, such as the retina, the feedback connections necessary to use the prediction error for learning are absent. For this case, we propose a second, model-free algorithm that uses a running estimate of the autocorrelation structure of the signal to perform the segmentation. We show that both algorithms do well when tasked with segmenting signals drawn from autoregressive models with piecewise-constant parameters. In particular, the segmentation accuracy is similar to that obtained from oracle-like methods in which the ground-truth parameters of the autoregressive models are known. We also test our methods on data sets generated by alternating snippets of voice recordings. We provide implementations of our algorithms at https://github.com/ttesileanu/bio-time-series.

scholarly journals Neural circuits for dynamics-based segmentation of time series

2021 ◽  
Author(s):  
Tiberiu Teşileanu ◽  
Siavash Golkar ◽  
Samaneh Nasiri ◽  
Anirvan M. Sengupta ◽  
Dmitri B. Chklovskii
Keyword(s):  
Time Series ◽  
Latent Variables ◽  
Prediction Error ◽  
Ground Truth ◽  
Brain Regions ◽  
Autoregressive Models ◽  
Model Free ◽  
Segmentation Task ◽  
Feedback Connections ◽  
The Brain

AbstractThe brain must extract behaviorally relevant latent variables from the signals streamed by the sensory organs. Such latent variables are often encoded in the dynamics that generated the signal rather than in the specific realization of the waveform. Therefore, one problem faced by the brain is to segment time series based on underlying dynamics. We present two algorithms for performing this segmentation task that are biologically plausible, which we define as acting in a streaming setting and all learning rules being local. One algorithm is model-based and can be derived from an optimization problem involving a mixture of autoregressive processes. This algorithm relies on feedback in the form of a prediction error, and can also be used for forecasting future samples. In some brain regions, such as the retina, the feedback connections necessary to use the prediction error for learning are absent. For this case, we propose a second, model-free algorithm that uses a running estimate of the autocorrelation structure of the signal to perform the segmentation. We show that both algorithms do well when tasked with segmenting signals drawn from autoregressive models with piecewise-constant parameters. In particular, the segmentation accuracy is similar to that obtained from oracle-like methods in which the ground-truth parameters of the autoregressive models are known. We provide implementations of our algorithms at https://github.com/ttesileanu/bio-time-series.

scholarly journals Prefrontal solution to the bias-variance tradeoff during reinforcement learning

2020 ◽  
Author(s):  
Dongjae Kim ◽  
Jaeseung Jeong ◽  
Sang Wan Lee
Keyword(s):  
Adaptive Control ◽  
Reinforcement Learning ◽  
Prediction Error ◽  
Brain Regions ◽  
Decision Task ◽  
Prediction Errors ◽  
Model Based ◽  
Model Free ◽  
Bias Variance ◽  
The Brain

AbstractThe goal of learning is to maximize future rewards by minimizing prediction errors. Evidence have shown that the brain achieves this by combining model-based and model-free learning. However, the prediction error minimization is challenged by a bias-variance tradeoff, which imposes constraints on each strategy’s performance. We provide new theoretical insight into how this tradeoff can be resolved through the adaptive control of model-based and model-free learning. The theory predicts the baseline correction for prediction error reduces the lower bound of the bias–variance error by factoring out irreducible noise. Using a Markov decision task with context changes, we showed behavioral evidence of adaptive control. Model-based behavioral analyses show that the prediction error baseline signals context changes to improve adaptability. Critically, the neural results support this view, demonstrating multiplexed representations of prediction error baseline within the ventrolateral and ventromedial prefrontal cortex, key brain regions known to guide model-based and model-free learning.One sentence summaryA theoretical, behavioral, computational, and neural account of how the brain resolves the bias-variance tradeoff during reinforcement learning is described.

scholarly journals Slow Cortical Waves via Cyclicity

2021 ◽  
Author(s):  
Ivan Abraham ◽  
Bahar Shahsavarani ◽  
Ben Zimmerman ◽  
Fatima Husain ◽  
yuliy baryshnikov
Keyword(s):  
Time Series ◽  
Dynamic Structure ◽  
Resting State Fmri ◽  
Brain Regions ◽  
Fine Grained ◽  
Human Connectome Project ◽  
Fmri Time Series ◽  
The Brain ◽  
Cortical Waves ◽  
Temporal Ordering

Fine-grained information about dynamic structure of cortical networks is crucial in unpacking brain function. Here,we introduced a novel analytical method to characterize the dynamic interaction between distant brain regions,based on cyclicity analysis, and applied it to data from the Human Connectome Project. Resting-state fMRI time series are aperiodic and, hence, lack a base frequency. Cyclicity analysis, which is time-reparametrization invariant, is effective in recovering dynamic temporal ordering of such time series along a circular trajectory without assuming any time scale. Our analysis detected the propagation of slow cortical waves across thebrain with consistent shifts in lead-lag relationships between specific brain regions. We also observed short bursts of strong temporal ordering that dominated overall lead-lag relationships between pairs of regions in the brain, which were modulated by tasks. Our results suggest the possible role played by slow waves of ordered information between brain regions that underlie emergent cognitive function.

scholarly journals 7,8-dihydroxyflavone Improves Neuropathological Changes in the Brain of Tg26 Mice, A Model for HIV-associated Neurocognitive Disorder

2020 ◽  
Author(s):  
Joseph Bryant ◽  
Sanketh Andhavarapu ◽  
Christopher Bever ◽  
Poornachander Guda ◽  
Akhil Katuri ◽  
...  
Keyword(s):  
Therapeutic Potential ◽  
Brain Regions ◽  
Mitochondrial Damage ◽  
Data Sets ◽  
Neurocognitive Disorder ◽  
Hiv Patients ◽  
Neuroprotective Effects ◽  
Downstream Signaling ◽  
Hiv Associated Neurocognitive Disorder ◽  
The Brain

Abstract Background: The combined antiretroviral therapy (cART) era has significantly increased the lifespan of HIV patients, turning a fatal disease to a chronic one. However, this lower but persistent level of HIV infection increases the susceptibility of HIV-associated neurocognitive disorder (HAND). Therefore, research is currently seeking improved treatment for this complication of HIV. In HIV+ patients, low levels of brain derived neurotrophic factor (BDNF) has been associated with worse neurocognitive impairment. Hence, BDNF administration has been gaining relevance as a possible adjunct therapy for HAND. However, systemic administration of BDNF is impractical because of poor pharmacological profile.Methods: We investigated the neuroprotective effects of BDNF-mimicking 7,8 dihydroxyflavone (DHF), a bioactive high-affinity TrkB agonist, in the memory-involved hippocampus and brain cortex of Tg26 mice, a murine model for HAND. We immunohistochemically stained brain tissue sections from vehicle-treated wild type (WT), vehicle-treated Tg26, and DHF (5 mg/kg, i.p)-treated Tg26 mice to examine activation of TrkB and downstream signaling, expression of HIV-1 chemokine co-receptors CXCR4 and CCR5, neuroinflammation, and mitochondrial damage. A one-way ANOVA with a Bonferroni Comparison post-hoc test was performed to analyze the data sets. Results: In the brain regions of Tg26 mice, we observed astrogliosis, increased CXCR4 and CCR5 expression, neuroinflammation, and mitochondrial damage. Hippocampi and cortices of DHF treated mice exhibited a reversal of these pathological changes, suggesting the therapeutic potential of DHF in HAND. Our data indicates that DHF increases the phosphorylation of TrkB, providing new insights about the role of the TrkB-Akt-NFkB signaling pathway in mediating these pathological hallmarks.Conclusions: Our study provides an overview of how targeting BDNF-TrkB signaling in the pathophysiology of HAND may be relevant for future therapies, and sheds light on 7,8 Dihydroxyflavone as a potential adjunct therapeutic agent to current antiviral therapy.

scholarly journals An Adversarial and Densely Dilated Network for Connectomes Segmentation

Symmetry ◽  
2018 ◽  
Vol 10 (10) ◽  
pp. 467 ◽  
Author(s):  
Ke Chen ◽  
Dandan Zhu ◽  
Jianwei Lu ◽  
Ye Luo
Keyword(s):  
State Of The Art ◽  
Piriform Cortex ◽  
Receptive Fields ◽  
Ground Truth ◽  
Training Data ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Boundary Map ◽  
Segmentation Task ◽  
The Brain

Automatic reconstructing of neural circuits in the brain is one of the most crucial studies in neuroscience. Connectomes segmentation plays an important role in reconstruction from electron microscopy (EM) images; however, it is rather challenging due to highly anisotropic shapes with inferior quality and various thickness. In our paper, we propose a novel connectomes segmentation framework called adversarial and densely dilated network (ADDN) to address these issues. ADDN is based on the conditional Generative Adversarial Network (cGAN) structure which is the latest advance in machine learning with power to generate images similar to the ground truth especially when the training data is limited. Specifically, we design densely dilated network (DDN) as the segmentor to allow a deeper architecture and larger receptive fields for more accurate segmentation. Discriminator is trained to distinguish generated segmentation from manual segmentation. During training, such adversarial loss function is optimized together with dice loss. Extensive experimental results demonstrate that our ADDN is effective for such connectomes segmentation task, helping to retrieve more accurate segmentation and attenuate the blurry effects of generated boundary map. Our method obtains state-of-the-art performance while requiring less computation on ISBI 2012 EM dataset and mouse piriform cortex dataset.

scholarly journals L-DOPA Reduces Model-Free Control of Behavior by Attenuating the Transfer of Value to Action

2016 ◽  
Author(s):  
Nils B. Kroemer ◽  
Ying Lee ◽  
Shakoor Pooseh ◽  
Ben Eppinger ◽  
Thomas Goschke ◽  
...  
Keyword(s):  
Reinforcement Learning ◽  
Prediction Error ◽  
Behavioral Control ◽  
Decision Task ◽  
Learning Models ◽  
Model Free ◽  
Markov Decision ◽  
Model Free Control ◽  
Reinforcement Learning Models ◽  
The Brain

AbstractDopamine is a key neurotransmitter in reinforcement learning and action control. Recent findings suggest that these components are inherently entangled. Here, we tested if increases in dopamine tone by administration of L-DOPA upregulate deliberative “model-based” control of behavior or reflexive “model-free” control as predicted by dual-control reinforcement-learning models. Alternatively, L-DOPA may impair learning as suggested by “value” or “thrift” theories of dopamine. To this end, we employed a two-stage Markov decision-task to investigate the effect of L-DOPA (randomized cross-over) on behavioral control while brain activation was measured using fMRI. L-DOPA led to attenuated model-free control of behavior as indicated by the reduced impact of reward on choice and increased stochasticity of model-free choices. Correspondingly, in the brain, L-DOPA decreased the effect of reward while prediction-error signals were unaffected. Taken together, our results suggest that L-DOPA reduces model-free control of behavior by attenuating the transfer of value to action.

scholarly journals Mean deviation based identification of activated voxels from time-series fMRI data of schizophrenia patients

F1000Research ◽  
2018 ◽  
Vol 7 ◽  
pp. 1615 ◽  
Author(s):  
Indranath Chatterjee
Keyword(s):  
Time Series ◽  
Superior Temporal Gyrus ◽  
Brain Regions ◽  
Medical Intervention ◽  
Fmri Data ◽  
Mean Deviation ◽  
Functional Changes ◽  
Statistical Measures ◽  
The Brain ◽  
Over Time

Background: Schizophrenia is a serious mental illness affecting different regions of the brain, which causes symptoms such as hallucinations and delusions. Functional magnetic resonance imaging (fMRI) is the most popular technique to study the functional activation patterns of the brain. The fMRI data is four-dimensional, composed of 3D brain images over time. Each voxel of the 3D brain volume is associated with a time series of signal intensity values. This study aimed to identify the distinct voxels from time-series fMRI data that show high functional activation during a task. Methods: In this study, a novel mean-deviation based approach was applied to time-series fMRI data of 34 schizophrenia patients and 34 healthy subjects. The statistical measures such as mean and median were used to find the functional changes in each voxel over time. The voxels that show significant changes for each subject were selected and thus used as the feature set during the classification of schizophrenia patients and healthy controls. Results: The proposed approach identifies a set of relevant voxels that are used to distinguish between healthy and schizophrenia subjects with high classification accuracy. The study shows functional changes in brain regions such as superior frontal gyrus, cuneus, medial frontal gyrus, middle occipital gyrus, and superior temporal gyrus. Conclusions: This work describes a simple yet novel feature selection algorithm for time-series fMRI data to identify the activated brain voxels that are generally affected in schizophrenia. The brain regions identified in this study may further help clinicians to understand the illness for better medical intervention. It may be possible to explore the approach to fMRI data of other psychological disorders.

scholarly journals Detecting brain connectivity changes during seizures using Granger causality measures on scalp EEG

2015 ◽  
Author(s):  
Christos Koutlis
Keyword(s):  
Time Series ◽  
Granger Causality ◽  
Time Window ◽  
Brain Connectivity ◽  
Characteristic Curve ◽  
Multivariate Time Series ◽  
Brain Regions ◽  
Brain Network ◽  
Network Feature ◽  
The Brain

In this work the objective is to detect brain connectivity changes during epileptic seizures using methods of multivariate time series analysis on scalp multi-channel EEG. Different brain regions represented by the electrode positions interact in terms of Granger causality and these directed connections formulate the brain network at a certain time window. The numerous proposed network features are believed to capture the information of many network characteristics. The ability of a single network feature of the brain network to detect the transition of brain activity from preictal to ictal is examined. The connectivity of the brain is estimated by 13 Granger causality indices on 7 epochs from multivariate time series (19 channels per epoch) at 15 time windows of 20 seconds (5 min in total) before seizure and during the seizure. The characteristics of the networks are estimated by 379 network features. Finally, the discrimination task (preictal vs. ictal) for each network feature is evaluated by the area under receiver operating characteristic curve (AUROC).

scholarly journals Generative adversarial network based on chaotic time series

2019 ◽  
Vol 9 (1) ◽  
Author(s):  
Makoto Naruse ◽  
Takashi Matsubara ◽  
Nicolas Chauvet ◽  
Kazutaka Kanno ◽  
Tianyu Yang ◽  
...  
Keyword(s):  
Time Series ◽  
Semiconductor Lasers ◽  
Latent Variables ◽  
Chaotic Time Series ◽  
Training Data ◽  
Generative Adversarial Networks ◽  
Data Sets ◽  
Deep Convolutional Neural Networks ◽  
Generative Adversarial Network ◽  
Adversarial Network

Abstract Generative adversarial networks (GANs) are becoming increasingly important in the artificial construction of natural images and related functionalities, wherein two types of networks called generators and discriminators evolve through adversarial mechanisms. Using deep convolutional neural networks and related techniques, high-resolution and highly realistic scenes, human faces, etc. have been generated. GANs generally require large amounts of genuine training data sets, as well as vast amounts of pseudorandom numbers. In this study, we utilized chaotic time series generated experimentally by semiconductor lasers for the latent variables of a GAN, whereby the inherent nature of chaos could be reflected or transformed into the generated output data. We show that the similarity in proximity, which describes the robustness of the generated images with respect to minute changes in the input latent variables, is enhanced, while the versatility overall is not severely degraded. Furthermore, we demonstrate that the surrogate chaos time series eliminates the signature of the generated images that is originally observed corresponding to the negative autocorrelation inherent in the chaos sequence. We also address the effects of utilizing chaotic time series to retrieve images from the trained generator.

scholarly journals Retinal Drug Delivery: Rethinking Outcomes for the Efficient Replication of Retinal Behavior

2020 ◽  
Vol 10 (12) ◽  
pp. 4258 ◽  
Author(s):  
Eliana B. Souto ◽  
Elena Sanchez-Lopez ◽  
Joana R. Campos ◽  
Raquel da Ana ◽  
Marta Espina ◽  
...  
Keyword(s):  
Computational Models ◽  
Vascular System ◽  
Data Sets ◽  
Neuronal System ◽  
Significant Development ◽  
Model Free ◽  
Disease Mechanisms ◽  
Free Data ◽  
Efficient Replication ◽  
The Brain

The retina is a highly organized structure that is considered to be "an approachable part of the brain." It is attracting the interest of development scientists, as it provides a model neurovascular system. Over the last few years, we have been witnessing significant development in the knowledge of the mechanisms that induce the shape of the retinal vascular system, as well as knowledge of disease processes that lead to retina degeneration. Knowledge and understanding of how our vision works are crucial to creating a hardware-adaptive computational model that can replicate retinal behavior. The neuronal system is nonlinear and very intricate. It is thus instrumental to have a clear view of the neurophysiological and neuroanatomic processes and to take into account the underlying principles that govern the process of hardware transformation to produce an appropriate model that can be mapped to a physical device. The mechanistic and integrated computational models have enormous potential toward helping to understand disease mechanisms and to explain the associations identified in large model-free data sets. The approach used is modulated and based on different models of drug administration, including the geometry of the eye. This work aimed to review the recently used mathematical models to map a directed retinal network.

Sign in / Sign up

Export Citation Format

Share Document