Adaptive neural network classifier for decoding MEG signals.
Ontology highlight
ABSTRACT:
SUBMITTER: Zubarev I
PROVIDER: S-EPMC6609925 | biostudies-literature | 2019 Aug
REPOSITORIES: biostudies-literature
Publications
Adaptive neural network classifier for decoding MEG signals.
Zubarev Ivan I Zetter Rasmus R Halme Hanna-Leena HL Parkkonen Lauri L
NeuroImage 20190504
We introduce two Convolutional Neural Network (CNN) classifiers optimized for inferring brain states from magnetoencephalographic (MEG) measurements. Network design follows a generative model of the electromagnetic (EEG and MEG) brain signals allowing explorative analysis of neural sources informing classification. The proposed networks outperform traditional classifiers as well as more complex neural networks when decoding evoked and induced responses to different stimuli across subjects. Impor ...[more]
Similar Datasets
Project description:The application of deep learning to neuroimaging big data will help develop computer-aided diagnosis of neurological diseases. Pattern recognition using deep learning can extract features of neuroimaging signals unique to various neurological diseases, leading to better diagnoses. In this study, we developed MNet, a novel deep neural network to classify multiple neurological diseases using resting-state magnetoencephalography (MEG) signals. We used the MEG signals of 67 healthy subjects, 26 patients with spinal cord injury, and 140 patients with epilepsy to train and test the network using 10-fold cross-validation. The trained MNet succeeded in classifying the healthy subjects and those with the two neurological diseases with an accuracy of 70.7 ± 10.6%, which significantly exceeded the accuracy of 63.4 ± 12.7% calculated from relative powers of six frequency bands (δ: 1-4 Hz; θ: 4-8 Hz; low-α: 8-10 Hz; high-α: 10-13 Hz; β: 13-30 Hz; low-γ: 30-50 Hz) for each channel using a support vector machine as a classifier (p = 4.2 × 10-2). The specificity of classification for each disease ranged from 86-94%. Our results suggest that this technique would be useful for developing a classifier that will improve neurological diagnoses and allow high specificity in identifying diseases.
| S-EPMC6433906 | biostudies-literature
Project description:Recordings of extracellular electrical, and later also magnetic, brain signals have been the dominant technique for measuring brain activity for decades. The interpretation of such signals is however nontrivial, as the measured signals result from both local and distant neuronal activity. In volume-conductor theory the extracellular potentials can be calculated from a distance-weighted sum of contributions from transmembrane currents of neurons. Given the same transmembrane currents, the contributions to the magnetic field recorded both inside and outside the brain can also be computed. This allows for the development of computational tools implementing forward models grounded in the biophysics underlying electrical and magnetic measurement modalities. LFPy (LFPy.readthedocs.io) incorporated a well-established scheme for predicting extracellular potentials of individual neurons with arbitrary levels of biological detail. It relies on NEURON (neuron.yale.edu) to compute transmembrane currents of multicompartment neurons which is then used in combination with an electrostatic forward model. Its functionality is now extended to allow for modeling of networks of multicompartment neurons with concurrent calculations of extracellular potentials and current dipole moments. The current dipole moments are then, in combination with suitable volume-conductor head models, used to compute non-invasive measures of neuronal activity, like scalp potentials (electroencephalographic recordings; EEG) and magnetic fields outside the head (magnetoencephalographic recordings; MEG). One such built-in head model is the four-sphere head model incorporating the different electric conductivities of brain, cerebrospinal fluid, skull and scalp. We demonstrate the new functionality of the software by constructing a network of biophysically detailed multicompartment neuron models from the Neocortical Microcircuit Collaboration (NMC) Portal (bbp.epfl.ch/nmc-portal) with corresponding statistics of connections and synapses, and compute in vivo-like extracellular potentials (local field potentials, LFP; electrocorticographical signals, ECoG) and corresponding current dipole moments. From the current dipole moments we estimate corresponding EEG and MEG signals using the four-sphere head model. We also show strong scaling performance of LFPy with different numbers of message-passing interface (MPI) processes, and for different network sizes with different density of connections. The open-source software LFPy is equally suitable for execution on laptops and in parallel on high-performance computing (HPC) facilities and is publicly available on GitHub.com.
| S-EPMC6305460 | biostudies-literature
Project description:Background:Glucose is a crucial energy source. In humans, it is the primary sugar for high energy demanding cells in brain, muscle and peripheral neurons. Deviations of blood glucose levels from normal levels for an extended period of time is dangerous or even fatal, so regulation of blood glucose levels is a biological imperative. The vagus nerve, comprised of sensory and motor fibres, provides a major anatomical substrate for regulating metabolism. While prior studies have implicated the vagus nerve in the neurometabolic interface, its specific role in either the afferent or efferent arc of this reflex remains elusive. Methods:Here we use recently developed methods to isolate and decode specific neural signals acquired from the surface of the vagus nerve in BALB/c wild type mice to identify those that respond robustly to hypoglycemia. We also attempted to decode neural signals related to hyperglycemia. In addition to wild type mice, we analyzed the responses to acute hypo- and hyperglycemia in transient receptor potential cation channel subfamily V member 1 (TRPV1) cell depleted mice. The decoding algorithm uses neural signals as input and reconstructs blood glucose levels. Results:Our algorithm was able to reconstruct the blood glucose levels with high accuracy (median error 18.6?mg/dl). Hyperglycemia did not induce robust vagus nerve responses, and deletion of TRPV1 nociceptors attenuated the hypoglycemia-dependent vagus nerve signals. Conclusion:These results provide insight to the sensory vagal signaling that encodes hypoglycemic states and suggest a method to measure blood glucose levels by decoding nerve signals. Trial registration:Not applicable.
| S-EPMC7098244 | biostudies-literature
Project description:Insects and vertebrates harbor specific neurons that encode the animal's head direction (HD) and provide an internal compass for spatial navigation. Each HD cell fires most strongly in one preferred direction. As the animal turns its head, however, HD cells in rat anterodorsal thalamic nucleus (ADN) and other brain areas fire already before their preferred direction is reached, as if the neurons anticipated the future HD. This phenomenon has been explained at a mechanistic level, but a functional interpretation is still missing. To close this gap, we use a computational approach based on the movement statistics of male rats and a simple model for the neural responses within the ADN HD network. Network activity is read out using population vectors in a biologically plausible manner, so that only past spikes are taken into account. We find that anticipatory firing improves the representation of the present HD by reducing the motion-induced temporal bias inherent in causal decoding. The amount of anticipation observed in ADN enhances the precision of the HD compass read-out by up to 40%. More generally, our theoretical framework predicts that neural integration times not only reflect biophysical constraints, but also the statistics of behaviorally relevant stimuli; in particular, anticipatory tuning should be found wherever neurons encode sensory signals that change gradually in time.SIGNIFICANCE STATEMENT Across different brain regions, populations of noisy neurons encode dynamically changing stimuli. Decoding a time-varying stimulus from the population response involves a trade-off: For short read-out times, stimulus estimates are unreliable as the number of stochastic spikes is small; for long read-outs, estimates are biased because they lag behind the true stimulus. We show that optimal decoding of temporally correlated stimuli not only relies on finding the right read-out time window but requires neurons to anticipate future stimulus values. We apply this general framework to the rodent head-direction system and show that the experimentally observed anticipation of future head directions can be explained at a quantitative level from the neuronal tuning properties, network size, and the animal's head-movement statistics.
| S-EPMC6462449 | biostudies-literature
Project description:We present an adaptive optical neural network based on a large-scale event-driven architecture. In addition to changing the synaptic weights (synaptic plasticity), the optical neural network's structure can also be reconfigured enabling various functionalities (structural plasticity). Key building blocks are wavelength-addressable artificial neurons with embedded phase-change materials that implement nonlinear activation functions and nonvolatile memory. Using multimode focusing, the activation function features both excitatory and inhibitory responses and shows a reversible switching contrast of 3.2 decibels. We train the neural network to distinguish between English and German text samples via an evolutionary algorithm. We investigate both the synaptic and structural plasticity during the training process. On the basis of this concept, we realize a large-scale network consisting of 736 subnetworks with 16 phase-change material neurons each. Overall, 8398 neurons are functional, highlighting the scalability of the photonic architecture.
| S-EPMC10588940 | biostudies-literature
Project description:The nervous system maintains physiological homeostasis through reflex pathways that modulate organ function. This process begins when changes in the internal milieu (e.g., blood pressure, temperature, or pH) activate visceral sensory neurons that transmit action potentials along the vagus nerve to the brainstem. IL-1β and TNF, inflammatory cytokines produced by immune cells during infection and injury, and other inflammatory mediators have been implicated in activating sensory action potentials in the vagus nerve. However, it remains unclear whether neural responses encode cytokine-specific information. Here we develop methods to isolate and decode specific neural signals to discriminate between two different cytokines. Nerve impulses recorded from the vagus nerve of mice exposed to IL-1β and TNF were sorted into groups based on their shape and amplitude, and their respective firing rates were computed. This revealed sensory neural groups responding specifically to TNF and IL-1β in a dose-dependent manner. These cytokine-mediated responses were subsequently decoded using a Naive Bayes algorithm that discriminated between no exposure and exposures to IL-1β and TNF (mean successful identification rate 82.9 ± 17.8%, chance level 33%). Recordings obtained in IL-1 receptor-KO mice were devoid of IL-1β-related signals but retained their responses to TNF. Genetic ablation of TRPV1 neurons attenuated the vagus neural signals mediated by IL-1β, and distal lidocaine nerve block attenuated all vagus neural signals recorded. The results obtained in this study using the methodological framework suggest that cytokine-specific information is present in sensory neural signals within the vagus nerve.
| S-EPMC6003492 | biostudies-literature
Project description:G-protein-coupled receptor-regulated cAMP production from endosomes can specify signaling to the nucleus by moving the source of cAMP without changing its overall amount. How this is possible remains unknown because cAMP gradients dissipate over the nanoscale, whereas endosomes typically localize micrometers from the nucleus. We show that the key location-dependent step for endosome-encoded transcriptional control is nuclear entry of cAMP-dependent protein kinase (PKA) catalytic subunits. These are sourced from punctate accumulations of PKA holoenzyme that are densely distributed in the cytoplasm and titrated by global cAMP into a discrete metastable state, in which catalytic subunits are bound but dynamically exchange. Mobile endosomes containing activated receptors collide with the metastable PKA puncta and pause in close contact. We propose that these properties enable cytoplasmic PKA to act collectively like a semiconductor, converting nanoscale cAMP gradients generated from endosomes into microscale elevations of free catalytic subunits to direct downstream signaling.
| S-EPMC8084946 | biostudies-literature
Project description:BackgroundDealing with the high dimension of both neuroimaging data and genetic data is a difficult problem in the association of genetic data to neuroimaging. In this article, we tackle the latter problem with an eye toward developing solutions that are relevant for disease prediction. Supported by a vast literature on the predictive power of neural networks, our proposed solution uses neural networks to extract from neuroimaging data features that are relevant for predicting Alzheimer's Disease (AD) for subsequent relation to genetics. The neuroimaging-genetic pipeline we propose is comprised of image processing, neuroimaging feature extraction and genetic association steps. We present a neural network classifier for extracting neuroimaging features that are related with the disease. The proposed method is data-driven and requires no expert advice or a priori selection of regions of interest. We further propose a multivariate regression with priors specified in the Bayesian framework that allows for group sparsity at multiple levels including SNPs and genes.ResultsWe find the features extracted with our proposed method are better predictors of AD than features used previously in the literature suggesting that single nucleotide polymorphisms (SNPs) related to the features extracted by our proposed method are also more relevant for AD. Our neuroimaging-genetic pipeline lead to the identification of some overlapping and more importantly some different SNPs when compared to those identified with previously used features.ConclusionsThe pipeline we propose combines machine learning and statistical methods to benefit from the strong predictive performance of blackbox models to extract relevant features while preserving the interpretation provided by Bayesian models for genetic association. Finally, we argue in favour of using automatic feature extraction, such as the method we propose, in addition to ROI or voxelwise analysis to find potentially novel disease-relevant SNPs that may not be detected when using ROIs or voxels alone.
| S-EPMC10311793 | biostudies-literature
Project description:Decoding speech envelopes from electroencephalogram (EEG) signals holds potential as a research tool for objectively assessing auditory processing, which could contribute to future developments in hearing loss diagnosis. However, current methods struggle to meet both high accuracy and interpretability. We propose a deep learning model called the auditory decoding transformer (ADT) network for speech envelope reconstruction from EEG signals to address these issues. The ADT network uses spatio-temporal convolution for feature extraction, followed by a transformer decoder to decode the speech envelopes. Through anticausal masking, the ADT considers only the current and future EEG features to match the natural relationship of speech and EEG. Performance evaluation shows that the ADT network achieves average reconstruction scores of 0.168 and 0.167 on the SparrKULee and DTU datasets, respectively, rivaling those of other nonlinear models. Furthermore, by visualizing the weights of the spatio-temporal convolution layer as time-domain filters and brain topographies, combined with an ablation study of the temporal convolution kernels, we analyze the behavioral patterns of the ADT network in decoding speech envelopes. The results indicate that low- (0.5-8 Hz) and high-frequency (14-32 Hz) EEG signals are more critical for envelope reconstruction and that the active brain regions are primarily distributed bilaterally in the auditory cortex, consistent with previous research. Visualization of attention scores further validated previous research. In summary, the ADT network balances high performance and interpretability, making it a promising tool for studying neural speech envelope tracking.
| S-EPMC11489951 | biostudies-literature
Project description:BackgroundIn silico promoter prediction represents an important challenge in bioinformatics as it provides a first-line approach to identifying regulatory elements to support wet-lab experiments. Historically, available promoter prediction software have focused on sigma factor-associated promoters in the model organism E. coli. As a consequence, traditional promoter predictors yield suboptimal predictions when applied to other prokaryotic genera, such as Pseudomonas, a Gram-negative bacterium of crucial medical and biotechnological importance.ResultsWe developed SAPPHIRE, a promoter predictor for σ70 promoters in Pseudomonas. This promoter prediction relies on an artificial neural network that evaluates sequences on their similarity to the - 35 and - 10 boxes of σ70 promoters found experimentally in P. aeruginosa and P. putida. SAPPHIRE currently outperforms established predictive software when classifying Pseudomonas σ70 promoters and was built to allow further expansion in the future.ConclusionsSAPPHIRE is the first predictive tool for bacterial σ70 promoters in Pseudomonas. SAPPHIRE is free, publicly available and can be accessed online at www.biosapphire.com . Alternatively, users can download the tool as a Python 3 script for local application from this site.
| S-EPMC7510298 | biostudies-literature