Hybrid Independent Component Analysis Research Articles

Cognitive and psychiatric relevance of dynamic functional connectivity states in a large (N > 10,000) children population.

Children's brains dynamically adapt to the stimuli from the internal state and the external environment, allowing for changes in cognitive and mental behavior. In this work, we performed a large-scale analysis of dynamic functional connectivity (DFC) in children aged 9~11 years, investigating how brain dynamics relate to cognitive performance and mental health at an early age. A hybrid independent component analysis framework was applied to the Adolescent Brain Cognitive Development (ABCD) data containing 10,988 children. We combined a sliding-window approach with k-means clustering to identify five brain states with distinct DFC patterns. Interestingly, the occurrence of a strongly connected state with the most within-network synchrony and the anticorrelations between networks, especially between the sensory networks and between the cerebellum and other networks, was negatively correlated with cognitive performance and positively correlated with dimensional psychopathology in children. Meanwhile, opposite relationships were observed for a DFC state showing integration of sensory networks and antagonism between default-mode and sensorimotor networks but weak segregation of the cerebellum. The mediation analysis further showed that attention problems mediated the effect of DFC states on cognitive performance. This investigation unveils the neurological underpinnings of DFC states, which suggests that tracking the transient dynamic connectivity may help to characterize cognitive and mental problems in children and guide people to provide early intervention to buffer adverse influences.

Functional connectivity uniqueness and variability? Linkages with cognitive and psychiatric problems in children

Brain functional connectivity (FC) derived from functional magnetic resonance imaging has been serving as a potential ‘fingerprint’ for adults. However, cross-scan variation of FC can be substantial and carries biological information, especially during childhood. Here we performed a large-scale cross-sectional analysis on cross-scan FC stability and its associations with a diverse range of health measures in children. Functional network connectivity (FNC) was extracted via a hybrid independent component analysis framework on 9,071 participants and compared across four scans. We found that FNC can identify a given child from a large group with high accuracy (maximum >94%) and replicated the results across multiple scans. We then performed a linear mixed-effects model to investigate how cross-scan FNC stability was predictive of children’s behaviour. Although we could not find strong relationships between FNC stability and children’s behaviour, we observed significant but small associations between them (maximum r = 0.1070), with higher stability correlated with better cognitive performance, longer sleep duration and less psychotic expression. Via a multivariate analysis method, we captured larger effects between FNC stability and children’s cognitive performance (maximum r = 0.2932), which further proved the relevance of FNC stability to neurocognitive development. Overall, our findings show that a child’s connectivity profile is not only intrinsic but also exhibits reliable variability across scans, regardless of brain growth and development. Cross-scan connectivity stability may serve as a valuable neuroimaging feature to draw inferences on early cognitive and psychiatric behaviours in children.

Hybrid independent component analysis (H-ICA) with simultaneous analysis of high-order and second-order statistics for industrial process monitoring

Both the Gaussian and non-Gaussian information exist simultaneously in many real industrial processes. However, the standard independent component analysis (ICA) cannot effectively explore process information when there is more than one independent component (IC) following Gaussian distribution. Besides, process may present an obvious temporal structure due to wide applications of controllers, which has not been investigated by standard ICA. To solve the above problem, the paper proposed a hybrid ICA (H-ICA) algorithm for process monitoring by concurrent analysis of both high-order and second-order statistics. The time-delayed correlation matrices are first diagonalized for robust whitening to remove the effect of the additive white noise. Then, non-Gaussian information is explored by FastICA with the utilization of higher-order statistics, and Gaussian information is extracted by analyzing the time structure information contained in second-order nonzero-delayed covariance matrices. By the hybrid analysis of high-order and second-order statistics, both non-Gaussian-distributed and Gaussian-distributed ICs are extracted. In this way, more comprehensive process information is fully investigated and analyzed for the development of process monitoring strategy. The proposed H-ICA process monitoring algorithm is verified by both a numerical example and a real thermal power plant process which illustrates its feasibility and efficacy.

Novel Spectral Efficient Technique for MIMO-OFDM Channel Estimation with Reference to PAPR and BER Analysis

With emerge of increasing research in domain of future wireless communications, massive multiple input multiple output (MIMO) attracted most of researchers interests. Massive MIMO is nothing but high speed wireless communication standards. The performance of MIMO systems is based on techniques used for channel estimation. Efficient channel estimation leads to spectral efficient wireless communications. There are number of channel estimation techniques presented recently in literature with pros and cons. The recent method shows the spectral and bit error rate (BER) efficiency, however apart from this, there is need of improving the peak to average power ratio (PAPR). Recently we proposed, novel channel estimation method as the existing channel estimation techniques failed to effectively solve the inter symbol interference (ISI) problem. The presence of ISI in MIMO-OFDM may leads to worst performance. Our proposed blind channel estimation is combined with independent component analysis (ICA) hence this method is called as hybrid ICA (HICA) to minimize the ISI effect. The extensive simulation analysis of proposed HICA required to claiming the scalability as well as reliability. In this paper, proposed study on additional performance metrics such as PAPR and computational costs (energy) along with BER and spectral efficiency performances. The result claims that HICA is not improving the PAPR and energy performances significantly.

Spectral Efficient Blind Channel Estimation Technique for MIMO-OFDM Communications

<p><em>With emerge of increasing research in the domain of future wireless communications, massive MIMO (multiple inputs multiple outputs) attracted most of researchers interests. Massive MIMO is high-speed wireless communication standards. A channel estimation technology plays the essential role in the MIMO systems. Efficient channel estimation leads to spectral efficient wireless communications. The critics of Inter-Symbol Interference (ISI) are the challenging tasks while designing the channel estimation methods. To mitigate the challenges of ISI, we proposed the novel blind channel estimation method which based on Independent component analysis (ICA) in this paper. Proposed channel estimation it works for both blind interference cancellation and ISI cancellation. The proposed Hybrid ICA (HICA) method depends on pulse shape filtering and ambiguity removal to improve the spectral efficiency and reliability for MIMO communications. The Kurtosis operation is used to measure the complex data at first to estimate the common signals. Then we exploited the advantages of 3rd and 4th order Higher Order Statistics (HOS) to priorities the common signals during the channel estimation. In this paper, we present the detailed design and evaluation of HICA blind channel estimation method. We showed the simulation results of HICA against the state-of-art techniques for channel estimation using BER, MSE, and PAPR.</em><em></em></p>

Nicotinic modulation of salience network connectivity and centrality in schizophrenia

Although functional abnormalities of the salience network are associated with schizophrenia, the acute effects of nicotine on its function and network dynamics during the resting state in patients are poorly understood. In this study, the effects of a 7 mg nicotine patch (vs. placebo) on salience network connectivity were examined in 17 patients with schizophrenia and 19 healthy subjects. We hypothesized abnormal connectivity between the salience network and other major networks (e.g. executive network) in patients under placebo administration and amelioration of this difference after nicotine. We also examined effects of nicotine on betweenness centrality (a measure of the influence of a region on information transfer throughout the brain) and local efficiency (a measure of local information transfer) of the network. A hybrid independent component analysis (ICA)/seed-based connectivity approach was implemented in which the salience network was extracted by ICA and cortical network peaks (anterior cingulate cortex (ACC), left and right insula) were used as seeds for whole-brain seed-to-voxel connectivity analysis. Significant drug X diagnosis interactions were observed between the ACC seed and superior parietal lobule and ventrolateral prefrontal cortex. A significant interaction effect was also observed between the left insula seed and middle cingulate cortex. During placebo conditions, abnormal connectivity predicted negative symptom severity and lower global functioning in patients. A significant drug X diagnosis interaction was also observed for betweenness centrality of the ACC. These results suggest that nicotine may target abnormalities in functional connectivity between salience and executive network areas in schizophrenia as well as affect the ability of the salience network to act as an integrator of global signaling in the disorder.

Hybrid ICA-Regression: Automatic Identification and Removal of Ocular Artifacts from Electroencephalographic Signals.

Electroencephalography (EEG) is a portable brain-imaging technique with the advantage of high-temporal resolution that can be used to record electrical activity of the brain. However, it is difficult to analyze EEG signals due to the contamination of ocular artifacts, and which potentially results in misleading conclusions. Also, it is a proven fact that the contamination of ocular artifacts cause to reduce the classification accuracy of a brain-computer interface (BCI). It is therefore very important to remove/reduce these artifacts before the analysis of EEG signals for applications like BCI. In this paper, a hybrid framework that combines independent component analysis (ICA), regression and high-order statistics has been proposed to identify and eliminate artifactual activities from EEG data. We used simulated, experimental and standard EEG signals to evaluate and analyze the effectiveness of the proposed method. Results demonstrate that the proposed method can effectively remove ocular artifacts as well as it can preserve the neuronal signals present in EEG data. A comparison with four methods from literature namely ICA, regression analysis, wavelet-ICA (wICA), and regression-ICA (REGICA) confirms the significantly enhanced performance and effectiveness of the proposed method for removal of ocular activities from EEG, in terms of lower mean square error and mean absolute error values and higher mutual information between reconstructed and original EEG.

A Blind Carrier Frequency Offset Estimation Scheme for OFDM Systems via Hybrid-ICA Algorithm

This paper addresses a new iterative blind carrier frequency offset (CFO) estimator in orthogonal frequency-division multiplexing (OFDM) systems. In our scenario, relying on a hybrid independent component analysis (ICA) scheme, a blind frequency offset estimator is proposed for OFDM systems employing single antenna. The proposed estimator is a combination of the fast fixed-point ICA (FastICA) algorithm and the natural-gradient ICA algorithm, which inherits the fast convergence and high accuracy of theses algorithms, respectively. As a major advantage, the proposed estimator works in an efficient way in practical high-data-rate OFDM systems especially with large number of subcarriers. To demonstrate the efficiency of our proposed algorithm, numerical results are provided from computer simulations and comparisons are made with other existing CFO estimators in the literature. It is shown that by applying our proposed algorithm, the normalized mean-square error of the CFO estimator is decreased to a high extent. DOI: http://dx.doi.org/10.11591/ijece.v4i5.6533

A hybrid method of application of independent component analysis to in vivo1H MR spectra of childhood brain tumours

Independent component analysis (ICA) can automatically extract individual metabolite, macromolecular and lipid (MMLip) components from a series of in vivo MR spectra. The traditional feature extraction (FE)-based ICA approach is limited, in that a large sample size is required and a combination of metabolite and MMLip components can appear in the same independent component. The alternative ICA approach, based on blind source separation (BSS), is weak when dealing with overlapping peaks. Combining the advantages of both BSS and FE methods may lead to better results. Thus, we propose an ICA approach involving a hybrid of the BSS and FE techniques for the automated decomposition of a series of MR spectra. Experiments were performed on synthesised and patient in vivo childhood brain tumour MR spectra datasets. The hybrid ICA method showed an improvement in the decomposition ability compared with BSS-ICA or FE-ICA, with an increased correlation between the independent components and simulated metabolite and MMLip signals. Furthermore, we were able to automatically extract metabolites from the patient MR spectra dataset that were not in commonly used basis sets (e.g. guanidinoacetate).

Hybrid independent component analysis and twin support vector machine learning scheme for subtle gesture recognition

Myoelectric signal classification is one of the most difficult pattern recognition problems because large variations in surface electromyogram features usually exist. In the literature, attempts have been made to apply various pattern recognition methods to classify surface electromyography into components corresponding to the activities of different muscles, but this has not been very successful, as some muscles are bigger and more active than others. This results in dataset discrepancy during classification. Multicategory classification problems are usually solved by solving many, one-versus-rest binary classification tasks. These subtasks unsurprisingly involve unbalanced datasets. Consequently, we need a learning methodology that can take into account unbalanced datasets in addition to large variations in the distributions of patterns corresponding to different classes. Here, we attempt to address the above issues using hybrid features extracted from independent component analysis and twin support vector machine techniques.

Hybrid ICA–Bayesian network approach reveals distinct effective connectivity differences in schizophrenia

We utilized a discrete dynamic Bayesian network (dDBN) approach (Burge, J., Lane, T., Link, H., Qiu, S., Clark, V.P., 2007. Discrete dynamic Bayesian network analysis of fMRI data. Hum Brain Mapp.) to determine differences in brain regions between patients with schizophrenia and healthy controls on a measure of effective connectivity, termed the approximate conditional likelihood score (ACL) (Burge, J., Lane, T., 2005. Learning Class-Discriminative Dynamic Bayesian Networks. Proceedings of the International Conference on Machine Learning, Bonn, Germany, pp. 97-104.). The ACL score represents a class-discriminative measure of effective connectivity by measuring the relative likelihood of the correlation between brain regions in one group versus another. The algorithm is capable of finding non-linear relationships between brain regions because it uses discrete rather than continuous values and attempts to model temporal relationships with a first-order Markov and stationary assumption constraint (Papoulis, A., 1991. Probability, random variables, and stochastic processes. McGraw-Hill, New York.). Since Bayesian networks are overly sensitive to noisy data, we introduced an independent component analysis (ICA) filtering approach that attempted to reduce the noise found in fMRI data by unmixing the raw datasets into a set of independent spatial component maps. Components that represented noise were removed and the remaining components reconstructed into the dimensions of the original fMRI datasets. We applied the dDBN algorithm to a group of 35 patients with schizophrenia and 35 matched healthy controls using an ICA filtered and unfiltered approach. We determined that filtering the data significantly improved the magnitude of the ACL score. Patients showed the greatest ACL scores in several regions, most markedly the cerebellar vermis and hemispheres. Our findings suggest that schizophrenia patients exhibit weaker connectivity than healthy controls in multiple regions, including bilateral temporal, frontal, and cerebellar regions during an auditory paradigm.

고정점 알고리즘과 시간적 상관성의 적응조정 견실 알고리즘을 조합한 독립성분분석

본 논문에서는 고정점 알고리즘과 신호의 시간적 상관성을 적응 조정한 견실 알고리즘의 조합형 독립성분분석을 제안하였다. 여기서 고정점 알고리즘은 뉴우턴법의 경신규칙에 기초한 방법으로 빠른 분석속도와 우수한 분석성능을 얻기 위함이고, 견실 알고리즘은 시간적 상호 의존성이나 낮은 쿠토시스를 가지는 신호도 효과적으로 분석하기 위함이다. 특히 견실 알고리즘에서 경험적으로 설정되던 최대지연시간을 신호상호간의 자기상관함수를 이용하여 적응 조정되도록 함으로써 그 성능을 더욱 더 개선하였다. 제안된 독립성분분석을 500개 샘플을 가시는 4개의 신호와 <TEX>$512\times512$</TEX> 픽셀의 10개 영상으로부터 임의의 혼합행렬에 따라 발생되는 혼합신호와 혼합영상 각각의 분리에 적용한 결과, 고정점 알고리즘의 독립성분분석 및 고정점 알고리즘과 최대시간지연을 경험적으로 설정하는 기존의 견실 알고리즘을 단순히 조합한 독립성분분석에 비해 분리속도와 분리률에서 개선된 성능이 있음을 확인하였다. 특히 문제의 규모가 증가할수록 분석성능의 개선정도도 증가함을 확인하였다. This paper proposes a hybrid independent component analysis(ICA) of fixed-point(FP) algorithm and robust algorithm. The FP algorithm is applied for improving the analysis speed and performance, and the robust algorithm is applied for preventing performance degradations by means of very small kurtosis and temporal correlations between components. And the adaptive adaptation of temporal correlations has been proposed for solving limits of the conventional robust algorithm dependent on the maximum time delay. The proposed ICA has been applied to the problems for separating the 4-mixed signals of 500 samples and 10-mixed images of <TEX>$512\times512$</TEX>pixels, respectively. The experimental results show that the proposed ICA has a characteristics of adaptively adapting the maximum time delay, and has a superior separation performances(speed, rate) to conventional FP-ICA and hybrid ICA of heuristic correlation. Especially, the proposed ICA gives the larger degree of improvement as the problem size increases.

Hybrid independent component analysis by adaptive LUT activation function neurons

The aim of this paper is to present an efficient implementation of unsupervised adaptive-activation function neurons dedicated to one-dimensional probability density estimation, with application to independent component analysis. The proposed implementation is a computationally light improvement to adaptive pseudo-polynomial neurons, recently presented in Fiori, S. (2000a). Blind signal processing by the adaptive activation function neurons. Neural Networks, 13(6), 597–611, and is based upon the concept of ‘look-up table’ (LUT) neurons.

Detection of Consistently Task-Related Activations in fMRI Data with Hybrid Independent Component Analysis

fMRI data are commonly analyzed by testing the time course from each voxel against specific hypothesized waveforms, despite the fact that many components of fMRI signals are difficult to specify explicitly. In contrast, purely data-driven techniques, by focusing on the intrinsic structure of the data, lack a direct means to test hypotheses of interest to the examiner. Between these two extremes, there is a role for hybrid methods that use powerful data-driven techniques to fully characterize the data, but also use some a priori hypotheses to guide the analysis. Here we describe such a hybrid technique, HYBICA, which uses the initial characterization of the fMRI data from Independent Component Analysis and allows the experimenter to sequentially combine assumed task-related components so that one can gracefully navigate from a fully data-derived approach to a fully hypothesis-driven approach. We describe the results of testing the method with two artificial and two real data sets. A metric based on the diagnostic Predicted Sum of Squares statistic was used to select the best number of spatially independent components to combine and utilize in a standard regressional framework. The proposed metric provided an objective method to determine whether a more data-driven or a more hypothesis-driven approach was appropriate, depending on the degree of mismatch between the hypothesized reference function and the features in the data. HYBICA provides a robust way to combine the data-derived independent components into a data-derived activation waveform and suitable confounds so that standard statistical analysis can be performed.