Background Noise Model Research Articles

A Model of Functional Brain Connectivity and Background Noise as a Biomarker for Cognitive Phenotypes: Application to Autism

We present an efficient approach to discriminate between typical and atypical brains from macroscopic neural dynamics recorded as magnetoencephalograms (MEG). Our approach is based on the fact that spontaneous brain activity can be accurately described with stochastic dynamics, as a multivariate Ornstein-Uhlenbeck process (mOUP). By fitting the data to a mOUP we obtain: 1) the functional connectivity matrix, corresponding to the drift operator, and 2) the traces of background stochastic activity (noise) driving the brain. We applied this method to investigate functional connectivity and background noise in juvenile patients (n = 9) with Asperger’s syndrome, a form of autism spectrum disorder (ASD), and compared them to age-matched juvenile control subjects (n = 10). Our analysis reveals significant alterations in both functional brain connectivity and background noise in ASD patients. The dominant connectivity change in ASD relative to control shows enhanced functional excitation from occipital to frontal areas along a parasagittal axis. Background noise in ASD patients is spatially correlated over wide areas, as opposed to control, where areas driven by correlated noise form smaller patches. An analysis of the spatial complexity reveals that it is significantly lower in ASD subjects. Although the detailed physiological mechanisms underlying these alterations cannot be determined from macroscopic brain recordings, we speculate that enhanced occipital-frontal excitation may result from changes in white matter density in ASD, as suggested in previous studies. We also venture that long-range spatial correlations in the background noise may result from less specificity (or more promiscuity) of thalamo-cortical projections. All the calculations involved in our analysis are highly efficient and outperform other algorithms to discriminate typical and atypical brains with a comparable level of accuracy. Altogether our results demonstrate a promising potential of our approach as an efficient biomarker for altered brain dynamics associated with a cognitive phenotype.

A Time-Domain Model of Background Noise for In-Home MIMO PLC Networks

Multiple-input multiple-output techniques have recently become an important research field for enhancing the performance of in-home power-line communication (PLC) systems by exploiting the additional protective earth wire. The development of such systems requires an accurate description of the channel noise. In this paper, we have presented a model for PLC background noise based on an extensive set of measurements. We have adopted the framework of the multivariate time series to model the PLC background noise. This paper employs the vector autoregressive modeling technique to extract noise model parameters from the measured noise. We have verified the accuracy of the noise model by comparing time- and frequency-domain correlation of measured and modeled noise.

Mining the UKIDSS Galactic Plane Survey: star formation and embedded clusters

Data mining techniques must be developed and applied to analyse the large public data bases containing hundreds to thousands of millions entries. The aim of this study is to develop methods for locating previously unknown stellar clusters from the UKIDSS Galactic Plane Survey catalogue data. The cluster candidates are computationally searched from pre-filtered catalogue data using a method that fits a mixture model of Gaussian densities and background noise using the Expectation Maximization algorithm. The catalogue data contains a significant number of false sources clustered around bright stars. A large fraction of these artefacts were automatically filtered out before or during the cluster search. The UKIDSS data reduction pipeline tends to classify marginally resolved stellar pairs and objects seen against variable surface brightness as extended objects (or "galaxies" in the archive parlance). 10% or 66 x 10^6 of the sources in the UKIDSS GPS catalogue brighter than 17 magnitudes in the K band are classified as "galaxies". Young embedded clusters create variable NIR surface brightness because the gas/dust clouds in which they were formed scatters the light from the cluster members. Such clusters appear therefore as clusters of "galaxies" in the catalogue and can be found using only a subset of the catalogue data. The detected "galaxy clusters" were finally screened visually to eliminate the remaining false detections due to data artefacts. Besides the embedded clusters the search also located locations of non clustered embedded star formation. The search covered an area of 1302 square degrees and 137 previously unknown cluster candidates and 30 previously unknown sites of star formation were found.

SU‐E‐I‐151: Performance Evaluation of a New Commercially Available Portable Pixelated Gamma Camera for 99mTc‐Scintimammography

Purpose: To characterize the imaging performance of a pixelated gamma camera (Ergo, Digirad) for 99mTc‐scintimammography. Methods: The 31×40cm̂2 Ergo detector consists of 6mm thick 3.31×3.24mm̂2 CsI(Tl) crystals coupled to silicon photodiodes. The sensitivity and resolution was measured according to NEMA NU 1‐2007. Scintimammography performance was evaluated as hot tumor detection capabilities. We acquired images of different 99mTc‐spheres (inner diameters 3.95, 4.95, 6.23, 7.86, 9.89, 12.43mm) suspended in a 6.6cm thick 500mL 99mTc‐water bath. Images were acquired at tumor depths of 1.7 and 3.7mm with different sphere‐to‐background ratios (SBR: 5.0, 8.0, 10.1, 12.5, 18.8) for 10min acquisition using low‐energy collimation (LEHR). Acquisitions for 10, 5, 2.5, 1.25min were performed at SBR=5 where the 99mTc‐activity concentration (uCi/mL) in spheres/background was 34/6.8. The sphere‐CNR was estimated as: Sphere‐CNR = (SphereCounts_Max ‐ BackgroundCounts_Mean)/ BackgroundNoise. BackgroundNoise was the average standard deviation of 10 different 9‐pixel ROIs distributed across the image. BackgroundNoise was modeled as a power‐law function of BackgroundCounts_Mean and used to predict the sphere‐CNR at different activity levels and acquisition times. A preliminary visual assessment of sphere visibility was used to establish the threshold‐CNR for detection. Results: The sensitivity and resolution at 10cm using LEHR was measured to be 128cpm/uCi and 7.5mm. BackgroundNoise was found to vary as BackgroundCounts_Mean̂−0.67 suggesting a Poisson noise dominated imaging system. Threshold‐CNR of ∼8 corresponded to visible spheres. For SBR=5, depth=3.7cm, acquisition=10min, the measured sphere‐CNR for diameters 6.23, 7.86, 9.89, 12.43mm were 12.1, 15.9, 22.8, 33, respectively; that decreased to <3, 6.6, 10.1, 14.6 for acquisition=2.5min. The CNR predicted by the power‐law model for BackgroundNoise was in good agreement (differences of 9‐26%) with measured sphere‐CNR at lower acquisition times. Conclusions: The Ergo system is suitable for 99mTc‐ scintimammography with predictable performance at different imaging conditions. Small tumors (<1cm) could be visualized at a range of SBR under tested conditions.

Detecting Abnormal Network Traffic in the Secure Event Management Systems

State-of-the-art intrusion detection and monitoring systems produce hundreds or even thousands of events every day. Unfortunately, most of these events are false positives, or irrelevant and can be considered as background noise, which makes their correlation, analysis and investigation very complicated and resource consuming. This paper attempts to simulate the modeling of background noise using the non-stationary time series analysis with lag smoothing Kalman filter. Then introduce and compare a second technique applying a multi-layered perceptron neural network with back ropagation network; an approach that is used for the first time in modeling and correlating the background noise. DARPA Dataset is used to analyze and compare both techniques and finally a verification experiment is conducted using a gathered dataset from real network environment.

Cepstrum-Domain Model Combination Based on Decomposition of Speech and Noise Using MMSE-LSA for ASR in Noisy Environments

This paper presents an efficient method for combining models of speech and noise for robust speech recognition applications in noisy environments. This method decomposes the cepstrum domain representation of noise-corrupted speech into clean speech cepstrum and background noise cepstrum components using a minimum mean squared error-log spectral amplitude (MMSE-LSA) criterion. Speech recognition is then performed on noisy cepstrum domain observations using a model that is formed by parallel combination of cepstrum domain clean speech distributions and background noise distributions estimated using this MMSE-LSA based noise decomposition. This method is far more efficient than other parallel model combination (PMC) procedures because model combination is performed directly in the cepstrum domain rather than in the linear spectral domain. Whereas background noise model estimation is addressed as a separate issue in existing PMC procedures, this method explicitly incorporates a mechanism to continually update background noise models and signal-to-noise ratio (SNR) estimates over time. The performance of the proposed cepstrum-domain model combination method is compared with a well known implementation of PMC which uses a log-normal approximation when combining speech and background noise model means and variances on a connected digit string recognition task which is subjected to mismatched channel and environment conditions. As a result, it is shown that the proposed model combination technique gives a word error rate that is comparable to PMC when background noise information and SNR are known prior to estimation. The paper will also present the results of experiments where a combination of cepstrum-domain feature compensation and model combination are applied to this task.

PLC Channel: Impulsive Noise Modelling and Its Performance Evaluation Under Different Array Coding Schemes

Power-line communications (PLC) is a field that has raised a lot of research in the past years. In this paper, we introduce array codes into the PLC environment and we study the channel's performance. In particular, generalized array codes (GAC), as well as row and column array codes (RAC) are applied. We examine their performance by obtaining three different code rates from each category. Therefore, the (8, 4, 4), (12, 7, 4) and (16, 11, 4) GAC codes as well as the (9, 4, 4), (12, 6, 4), and (15, 8, 4) RAC codes are used, meaning that we obtain code rates of (1/2), (7/12), (11/16), and (4/9), (6/12), (8/15) respectively. In addition, for reasons of completeness, convolutional codes are also being applied under the same channel conditions. Moreover, we suggest a hybrid coding technique, which combines the (8, 4, 4) GAC and the (15, 8, 4) RAC code in order to meet the requirements of the PLC time-varying channel and improve its performance. Concerning the system's design, we take into consideration Zimmermann's model for the PLC channel. We apply Middleton's model for the channel's background and impulsive noise, while we also introduce a novel way of estimating the system's impulsive noise. Finally, the well-known transmission technique of orthogonal frequency-division multiplexing is used. The channel's performance is evaluated in terms of the bit-error rate for different E <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">b</sub> /N <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> values via computer simulations.

LTD windows of the STDP learning rule and synaptic connections having a large transmission delay enable robust sequence learning amid background noise

Spike-timing-dependent synaptic plasticity (STDP) is a simple and effective learning rule for sequence learning. However, synapses being subject to STDP rules are readily influenced in noisy circumstances because synaptic conductances are modified by pre- and postsynaptic spikes elicited within a few tens of milliseconds, regardless of whether those spikes convey information or not. Noisy firing existing everywhere in the brain may induce irrelevant enhancement of synaptic connections through STDP rules and would result in uncertain memory encoding and obscure memory patterns. We will here show that the LTD windows of the STDP rules enable robust sequence learning amid background noise in cooperation with a large signal transmission delay between neurons and a theta rhythm, using a network model of the entorhinal cortex layer II with entorhinal-hippocampal loop connections. The important element of the present model for robust sequence learning amid background noise is the symmetric STDP rule having LTD windows on both sides of the LTP window, in addition to the loop connections having a large signal transmission delay and the theta rhythm pacing activities of stellate cells. Above all, the LTD window in the range of positive spike-timing is important to prevent influences of noise with the progress of sequence learning.

Modeling and Analysis of Noise Effects on Broadband Power-Line Communications

Power line noise is known to affect the performance of broadband power-line communications significantly. This paper presents a frequency-domain approach to characterize and model the statistical variation of power-line noise. The model considers both the background noise and the impulsive noise. The background noise model is based on statistical analysis of the results from two long-term measurements of noise spectrum conducted at two separate sites of a laboratory and a residential apartment. On the other hand, the impulsive noise model is obtained by direct measurements from the noise sources (i.e., various electrical household appliances). The amount of impulse noise reaching a power-line communications (PLC) receiver can then be determined with consideration of the channel transfer characteristics between the noise sources and the PLC receiver. Using these noise models, the performance of two major classes of digital modulation schemes, namely single-carrier modulation and multicarrier modulation, are analyzed and compared. It is found that the multicarrier scheme performs better than the single-carrier scheme when subjected to the observed power-line noise with non-Gaussian statistics.

Significance Test of Clusters in Gene Expression Profile Data

Clustering is a popular technique used in gene expression analyses. However, conventionally well-used clustering methods often produce variety in results depending on, e.g., initial algorithmic condition. Due to this instability, analysis results may be questionable. For example [3], although B-cell lymphomas had been classified into two distinct classes using hierarchical clustering, another analysis by a different clustering method failed to find these classes [1, 2]. In spite of such instability, the preprocessing or visualization of data by a clustering technique is still important for gene expression analyses. For utilizing technique with certain reliability, it is required to develop not only a stable method but also a method to evaluate quantitatively the reliability of the clustering results. In this report, we propose a quantitative evaluation method of clusters, which determines whether a data group, seemingly a cluster, constructs a statistically significant cluster. In particular, we propose a statistical test examining the existence possibility of a cluster within a parametric model of background noise. A seeming cluster is expressed by a window function represented by characteristic parameters. This statistical test can be extended to examining of feature spaces; we also propose a significance test of feature subspace extracted from PCA analysis.

Background noise model of the piezoresistive microphone

A model has been developed to describe the sources of background noise in piezoresistive microphones. Three noise sources have been identified: mechanical Johnson noise due to the Brownian motion of air molecules impinging on the diaphragm, electrical Johnson noise in the bridge resistors, and generation–recombination noise in the bridge resistors. The model includes a set of lumped elements to represent the acoustic impedance of the diaphragm, another set of lumped elements to account for the stress across the bridge resistors, and a controlled source to yield an output voltage proportional to the stress due to the piezoresistive effect. Predictions based on the manufacturer’s data agree with experimental measurements. Representing the piezoresistive microphone as a nonreciprocal device, the model can be used to determine the microphone sensitivity and frequency response, as well as the background noise.