Conventional Independent Component Analysis Research Articles

High-Throughput Independent Component Analysis Processor for Full Duplex Systems

This paper presents the algorithm and very-large-scale integration (VLSI) architecture of a high-throughput and highly efficient independent component analysis (ICA) processor for self-interference cancellation (SIC) in in-band full-duplex (IBFD) systems. This is the first VLSI architecture reported in the literature based on the state-of-the-art entropy bound minimization (EBM) approach. A novel ICA algorithm is presented in this paper with momentum gradient descent optimization. Simulation results show that the number of iterations for the proposed algorithm is significantly reduced compared to the conventional ICA algorithms. Furthermore, a novel early-distribution estimation scheme is proposed in the designed ICA processor to compute multiple distribution functions with low latency and low complexity. The processing flow and the efficiency for the hardware utilization are specifically designed so that the processing speed is maximized with minimum employment of hardware components. The proposed ICA processor is designed and implemented based on the application-specific-integrated circuit (ASIC) flow. The post-layout estimations show that compared with the conventional EBM-based scheme, the proposed design improves the throughput and efficiency by 30x. In addition, compared to prior designs shown in the literature, the proposed ICA processor also demonstrates a significant enhancement in terms of throughput and efficiency.

Removal of site effects and enhancement of signal using dual projection independent component analysis for pooling multi-site MRI data.

Combining magnetic resonance imaging (MRI) data from multi-site studies is a popular approach for constructing larger datasets to greatly enhance the reliability and reproducibility of neuroscience research. However, the scanner/site variability is a significant confound that complicates the interpretation of the results, so effective and complete removal of the scanner/site variability is necessary to realise the full advantages of pooling multi-site datasets. Independent component analysis (ICA) and general linear model (GLM) based harmonisation methods are the two primary methods used to eliminate scanner/site effects. Unfortunately, there are challenges with both ICA-based and GLM-based harmonisation methods to remove site effects completely when the signals of interest and scanner/site effects-related variables are correlated, which may occur in neuroscience studies. In this study, we propose an effective and powerful harmonisation strategy that implements dual projection (DP) theory based on ICA to remove the scanner/site effects more completely. This method can separate the signal effects correlated with site variables from the identified site effects for removal without losing signals of interest. Both simulations and vivo structural MRI datasets, including a dataset from Autism Brain Imaging Data Exchange II and a travelling subject dataset from the Strategic Research Program for Brain Sciences, were used to test the performance of a DP-based ICA harmonisation method. Results show that DP-based ICA harmonisation has superior performance for removing site effects and enhancing the sensitivity to detect signals of interest as compared with GLM-based and conventional ICA harmonisation methods.

Independent vector analysis: Model, applications, challenges

This paper overviews an appealing unsupervised learning method named independent vector analysis (IVA) for its promising applications, such as in audio/speech signal separation, medical signal processing, remote sensing, video/image processing, wireless communication processing, and so on. As a useful data-driven technique in blind source separation (BSS) field, IVA has played an increasingly vital role in dealing with the problems of convolutive mixture separation, multivariate latent variable analysis and multivariate data fusion. IVA extends the conventional independent component analysis (ICA) to multidimensional components, which can result in more available information utilization. Compared with ICA mechanism, IVA is not only to utilize the statistical independence of multivariate signals but also the statistical inner dependency of each multivariate signal. With this generalization, IVA can manipulate some prominent ill-pose issues faced in sensor receiving models and has the advantage to overcome the inherent random permutation ambiguity problem in joint BSS. Motivated by the flexible and versatile technology superiorities of IVA, this paper concentrates on reviewing the IVA model in details, associated methods briefly, and its potential applications as well as prospects. Moreover, some significant open problems about IVA challenges are also discussed in this paper.

Machine learning may predict individual hand motor activation from resting-state fMRI in patients with brain tumors in perirolandic cortex

The study aimed to evaluate the predictive validity of the neural network (NN) method for presurgical mapping of motor areas using resting-state functional MRI (rs-fMRI) data of patients with brain tumor located in the perirolandic cortex (PRC). A total of 109 patients with brain tumors occupying PRC underwent rs-fMRI and hand movement task-based fMRI (tb-fMRI) scans. Using a NN model trained on fMRI data of 47 healthy controls, individual task activation maps were predicted from their rs-fMRI data. NN-predicted maps were compared with task activation and independent component analysis (ICA)-derived maps. Spatial Pearson's correlation coefficients (CC) matrices and Dice coefficients (DC) between task activation and predicted activation using NN (DCNN_Act) and ICA (DCICA_Act) were calculated and compared using non-parametric tests. The effects of tumor types and head motion on predicted maps were demonstrated. The CC matrix of NN-predicted maps showed higher diagonal values compared with ICA-derived maps (p < 0.001). DCNN_Act were higher than DCICA_Act (p < 0.001) for patients with or without motor deficits. Lower DCs were found in subjects with head motion greater than one voxel. DCs were higher on the nontumor side than on the tumor side (p < 0.001), especially in the glioma group compared with meningioma and metastatic groups. This study indicated that the NN approach could predict individual motor activation using rs-fMRI data and could have promising clinical applications in brain tumor patients with anatomical and functional reorganizations. • The neural network machine learning approach successfully predicted hand motor activation in patients with a tumor in the perirolandic cortex, despite space-occupying effects and possible functional reorganization. • Compared to the conventional independent component analysis, the neural network approach utilizing resting-state fMRI data yielded a higher correlation to the active task hand activation data. • The Dice coefficient of machine learning-predicted activation vs. task fMRI activation was different between tumor and nontumor side, also between tumor types, which might indicate different effects of possible neurovascular uncoupling on resting-state and task fMRI.

Refinement of High-Gamma EEG Features From TBI Patients With Hemicraniectomy Using an ICA Informed by Simulated Myoelectric Artifacts.

Recent studies have shown the ability to record high-γ signals (80–160 Hz) in electroencephalogram (EEG) from traumatic brain injury (TBI) patients who have had hemicraniectomies. However, extraction of the movement-related high-γ remains challenging due to a confounding bandwidth overlap with surface electromyogram (EMG) artifacts related to facial and head movements. In our previous work, we described an augmented independent component analysis (ICA) approach for removal of EMG artifacts from EEG, and referred to as EMG Reduction by Adding Sources of EMG (ERASE). Here, we tested this algorithm on EEG recorded from six TBI patients with hemicraniectomies while they performed a thumb flexion task. ERASE removed a mean of 52 ± 12% (mean ± S.E.M) (maximum 73%) of EMG artifacts. In contrast, conventional ICA removed a mean of 27 ± 19% (mean ± S.E.M) of EMG artifacts from EEG. In particular, high-γ synchronization was significantly improved in the contralateral hand motor cortex area within the hemicraniectomy site after ERASE was applied. A more sophisticated measure of high-γ complexity is the fractal dimension (FD). Here, we computed the FD of EEG high-γ on each channel. Relative FD of high-γ was defined as that the FD in move state was subtracted by FD in idle state. We found relative FD of high-γ over hemicraniectomy after applying ERASE were strongly correlated to the amplitude of finger flexion force. Results showed that significant correlation coefficients across the electrodes related to thumb flexion averaged ~0.76, while the coefficients across the homologous electrodes in non-hemicraniectomy areas were nearly 0. After conventional ICA, a correlation between relative FD of high-γ and force remained high in both hemicraniectomy areas (up to 0.86) and non-hemicraniectomy areas (up to 0.81). Across all subjects, an average of 83% of electrodes significantly correlated with force was located in the hemicraniectomy areas after applying ERASE. After conventional ICA, only 19% of electrodes with significant correlations were located in the hemicraniectomy. These results indicated that the new approach isolated electrophysiological features during finger motor activation while selectively removing confounding EMG artifacts. This approach removed EMG artifacts that can contaminate high-gamma activity recorded over the hemicraniectomy.

Dictionary Learning-Based fMRI Data Analysis for Capturing Common and Individual Neural Activation Maps

In this paper, a novel dictionary learning (DL) method is proposed to estimate sparse neural activations from multi-subject fMRI data sets. By exploiting the label information such as the patient and the normal healthy groups, the activation maps that are commonly shared across the groups as well as those that can explain the group differences are both captured. The proposed method was tested using real fMRI data sets consisting of schizophrenic subjects and healthy controls. The DL approach not only reproduced most of the maps obtained from the conventional independent component analysis (ICA), but also identified more maps that are significantly group-different, including a number of novel ones that were not revealed by ICA. The stability analysis of the DL method and the correlation analysis with separate neuropsychological test scores further strengthen the validity of our analysis.

Snowball ICA: A Model Order Free Independent Component Analysis Strategy for Functional Magnetic Resonance Imaging Data.

In independent component analysis (ICA), the selection of model order (i.e., number of components to be extracted) has crucial effects on functional magnetic resonance imaging (fMRI) brain network analysis. Model order selection (MOS) algorithms have been used to determine the number of estimated components. However, simulations show that even when the model order equals the number of simulated signal sources, traditional ICA algorithms may misestimate the spatial maps of the signal sources. In principle, increasing model order will consider more potential information in the estimation, and should therefore produce more accurate results. However, this strategy may not work for fMRI because large-scale networks are widely spatially distributed and thus have increased mutual information with noise. As such, conventional ICA algorithms with high model orders may not extract these components at all. This conflict makes the selection of model order a problem. We present a new strategy for model order free ICA, called Snowball ICA, that obviates these issues. The algorithm collects all information for each network from fMRI data without the limitations of network scale. Using simulations and in vivo resting-state fMRI data, our results show that component estimation using Snowball ICA is more accurate than traditional ICA. The Snowball ICA software is available at https://github.com/GHu-DUT/Snowball-ICA.

Mainlobe jamming suppression with polarimetric multi-channel radar via independent component analysis

Mainlobe jamming will reduce the capabilities of modern radar significantly. Conventional independent component analysis (ICA)-based methods will suffer from the ineffectiveness when the angle of the target is same as that of the jammer. In this paper, exploring polarization characteristics, we propose an approach based on ICA for polarimetric multi-channel radar to suppress mainlobe interference. Specifically, the signal model of polarimetric multi-channel radar accounting for the target and jamming signals is derived. Then, the approach based on ICA is utilized to separate the target component and jamming component while achieving the mainlobe jamming suppression. Finally, the effectiveness and capacities of proposed method are demonstrated by simulations, and the results indicate that different kinds of mainlobe jammings can be effectively suppressed and the proposed method outperforms conventional jamming cancellation methods.

Randomized independent component analysis and linear discriminant analysis dimensionality reduction methods for hyperspectral image classification

Dimensionality reduction is an important step in processing the hyperspectral images (HSI) to overcome the curse of dimensionality problem. Linear dimensionality reduction methods such as Independent component analysis (ICA) and Linear discriminant analysis (LDA) are commonly employed to reduce the dimensionality of HSI. These methods fail to capture non-linear dependency in the HSI data, as data lies in the nonlinear manifold. To handle this, nonlinear transformation techniques based on kernel methods were introduced for dimensionality reduction of HSI. However, the kernel methods involve cubic computational complexity while computing the kernel matrix, and thus its potential cannot be explored when the number of pixels (samples) are large. In literature a fewer number of pixels are randomly selected to partial to overcome this issue, however this sub-optimal strategy might neglect important information in the HSI. In this paper, we propose randomized solutions to the ICA and LDA dimensionality reduction methods using Random Fourier features, and we label them as RFFICA and RFFLDA. Our proposed method overcomes the scalability issue and to handle the non-linearities present in the data more efficiently. Experiments conducted with two real-world hyperspectral datasets demonstrates that our proposed randomized methods outperform the conventional kernel ICA and kernel LDA in terms overall, per-class accuracies and computational time.

Introducing a Deflationary Approach to Space-Time ICA that uses temporal methods in Brain Signals Processing.

For the extraction of underlying sources of brain activity, time structure-based techniques for applying Independent Component Analysis (ICA) have been demonstrably more robust than state-of-the-art statistical-based methods, such as FastICA. Since the early application of conventional ICA on electroencephalogram (EEG) recordings, Space-Time ICA (ST-ICA) has emerged as more capable approach for extracting complex underlying activity, but not without the 'curse of dimensionality'. The challenges in the future development of ST-ICA will require a focus on the optimisation of the mixing matrix, and on component clustering techniques. This paper proposes a new optimisation approach for the mixing matrix, which makes ST-ICA more tractable, when using a time structure-based ICA technique, LSDIAG. Such techniques rely on constructing a multi-layer covariance matrix, Cxk of the original dataset to generate the inverse of the mixing matrix; Csk = WCxkWT. This means a simple truncation of the mixing matrix is not appropriate. To overcome this, we propose a deflationary approach to optimise a much smaller mixing matrix - based on the absolute values of the diagonals of the co-variance matrix, Csk, to represent the underlying sources. The preliminary results of the new technique applied to different channels of EEG recorded using the standard 10-20 system - including the full selection of all channels - are very promising.Clinical Relevance-The potential of this deflationary approach for Space-Time ICA, seeks to allow clinicians to identify underlying sources in the brain - that both spatially and spectrally overlap - to be identified, whilst making the 'dimensionality' challenges more tractable. In the long run, applications of this technique could enhance certain brain-computer interface paradigms.

Photonic independent component analysis using an on-chip microring weight bank.

Independent component analysis (ICA) is a general-purpose technique for analyzing multi-dimensional data to reveal the underlying hidden factors that are maximally independent from each other. We report the first photonic ICA on mixtures of unknown signals by employing an on-chip microring (MRR) weight bank. The MRR weight bank performs so-called weighted addition (i.e., multiply-accumulate) operations on the received mixtures, and outputs a single reduced-dimensional representation of the signal of interest. We propose a novel ICA algorithm to recover independent components solely based on the statistical information of the weighted addition output, while remaining blind to not only the original sources but also the waveform information of the mixtures. We investigate both channel separability and near-far problems, and our two-channel photonic ICA experiment demonstrates our scheme holds comparable performance with the conventional software-based ICA method. Our numerical simulation validates the fidelity of the proposed approach, and studies noise effects to identify the operating regime of our method. The proposed technique could open new domains for future research in blind source separation, microwave photonics, and on-chip information processing.

Modeling motor task activation from resting-state fMRI using machine learning in individual subjects.

Resting-state functional MRI (rs-fMRI) has provided important insights into brain physiology. It has become an increasingly popular method for presurgical mapping, as an alternative to task-based functional MRI wherein the subject performs a task while being scanned. However, there is no commonly acknowledged gold standard approach for detecting eloquent brain areas using rs-fMRI data in clinical settings. In this study, a general linear model-based machine learning (GLM-ML) approach was tested to predict individual motor task activation based on rs-fMRI data. Its accuracy was then compared to a conventional independent component analysis (ICA) approach. 47 healthy subjects were scanned using resting state, active and passive motor task fMRI experiments using a clinically applicable low-resolution fMRI protocol. The model was trained to associate rs-fMRI network maps with that of hand movement task fMRI, then used to predict task activation maps for unseen subjects solely based on their rs-fMRI data. Our results showed that the GLM-ML approach can accurately predict individual differences in task activation using rs-fMRI data and outperform conventional ICA to detect task activation in the primary sensorimotor region. Furthermore, the predicted activation maps using the GLM -ML model matched well with the activation of passive hand movement fMRI on an individual basis. These results suggest that GLM-ML approach can robustly predict individual differences of task activation based on conventional low-resolution rs-fMRI data and has important implications for future clinical applications.

Illumination normalization using independent component analysis and filtering

ABSTRACTIn this work, we separate the illumination and reflectance components of a single input image which is non-uniformly illuminated. Considering the input image and its blurred version as two different combinations of illumination and reflectance components, we use the conventional independent component analysis (ICA) to separate these two components. The separated reflectance component, which is an illumination normalized version of the input image, can then be used as an effective pre-processed (illumination normalized) image for different computer vision tasks e.g. face recognition. To this end, we present simulation results to show that our proposed pre-processing method called illumination normalization using ICA increases the accuracy rate of several baseline face recognition systems (FRSs). The proposed method showed improved performance of baseline FRSs when using the Extended Yale-B databases.

Blind channel estimation for massive MIMO

In order to scale with the demand of higher data rates and improved spectral efficiency in next generation wireless communication systems, a large-scale multiple-input and multiple-output (MIMO) technology called massive MIMO has been proposed. In massive MIMO, appropriate signal-to-noise ratio (SNR) values can be achieved by the addition of base station (BS) antennas in place of increasing transmit power. Pilot-based channel estimation is widely used in conventional MIMO systems, where pilot signal sequences are sent from the user terminals (UTs) to the BS to estimate the channel. In massive MIMO-based cellular networks, channel estimation in a given cell will be impaired by the pilot signal sequences transmitted by users in other cells--rendering the addition of antennas or transmit power ineffective. This effect is called pilot contamination. Therefore, pilot-based channel estimation limits the performance of massive MIMO. Semi-blind and blind methods are alternatives to pilot-based channel estimation that perform channel estimation with short pilot signal sequences and without pilot signal sequences, respectively. Blind channel estimation is one of the promising solutions to the pilot contamination problem in massive MIMO. This paper compares, using MATLAB simulations of a cluster-based COST 2100 channel model, the performance of pilot-based, semi-blind, blind, and adaptive-blind channel estimation methods. The pilot contamination effect on different channel estimation methods and how channel estimation methods can be used to overcome pilot contamination are shown. Finally, an adaptive independent component analysis (ICA)-based channel estimation method, which outperforms conventional ICA in terms of computational complexity, is proposed.

Kernel-Regularized ICA for Computing Functional Topography from Resting-state fMRI.

Topographic regularity is a fundamental property in brain connectivity. In this work, we present a novel method for studying topographic regularity of functional connectivity based on resting-state fMRI (rfMRI), which is widely available and easy to acquire in large-scale studies. The main idea in our method is the incorporation of topographically regular structural connectivity for independent component analysis (ICA). This is enabled by the recent development of novel tractography and tract filtering algorithms that can generate highly organized fiber bundles connecting different brain regions. By leveraging these cutting-edge tractography algorithms, here we develop a kernel-regularized ICA method for the extraction of functional topography with rfMRI signals. In our experiments, we use rfMRI scans of 35 unrelated, right-handed subjects from the Human Connectome Project (HCP) to study the functional topography of the motor cortex. We first demonstrate that our method can generate functional connectivity maps with more regular topography than conventional group ICA. We also show that the components extracted by our algorithm are able to capture co-activation patterns that respect the organized topography of the motor cortex across the hemisphere. Finally, we show that our method achieves improved reproducibility as compared to conventional group ICA.

An improved artifacts removal method for high dimensional EEG

BackgroundMultiple noncephalic electrical sources superpose with brain signals in the recorded EEG. Blind source separation (BSS) methods such as independent component analysis (ICA) have been shown to separate noncephalic artifacts as unique components. However, robust and objective identification of artifact components remains a challenge in practice. In addition, with high dimensional data, ICA requires a large number of observations for stable solutions. Moreover, using signals from long recordings to provide the large observation set might violate the stationarity assumption of ICA due to signal changes over time. New methodInstead of decomposing all channels simultaneously, subsets of channels are randomly selected and decomposed with ICA. With reduced dimensionality of the subsets, much less amount of data is required to derive stable components. To characterize each independent component, an artifact relevance index (ARI) is calculated by template matching each component with a model of the artifact. Automatic artifact identification is then implemented based on the statistical distribution of ARI of the numerous components generated. ResultsThe proposed permutation resampling for identification matching (PRIM) method effectively removed eye blink artifacts from both simulated and real EEG. Comparison with existing methodThe average topomap correlation coefficient between the cleaned EEG and the ground truth is 0.89±0.01 for PRIM, compared with 0.64±0.05 for conventional ICA based method. The average relative root-mean-square error is 0.40±0.01 for PRIM, compared with 0.66±0.10 for conventional method. ConclusionsThe proposed method overcame limitations of conventional ICA based method and succeeded in removing eye blink artifacts automatically.

Blind Channel Estimation Using Wavelet Denoising of Independent Component Analysis for LTE

&lt;p&gt;A new proposal of blind channel estimation method for long term evoluation (LTE) based on combining advantages of denoising property of wavelet transform (WT) with blind estimation capability of independent component analysis (ICA) called wavelet denoising of ICA (WD-ICA) was presented. This new method increased the spectral efficiency compared to training based methods, and provided considerable performance enhancement over conventional ICA methods. The conventional blind channel estimation methods based on ICA were performed individually for each orthogonal frequency division multiplexing (OFDM) subcarrier. To reduce complexity of implementation of WD-ICA method, channel interpolation was used. This method was presented for multiple-input-multiple-output (MIMO) downlink LTE system. WD-ICA method was compared to conventional ICA methods and the Performance was evaluated by calculating normalized mean square error (NMSE) and bit error rate (BER). WD-ICA method as compared to the other known ICA channel estimation methods has lower complexity, lower value of NMSE, and lower value of BER, which indicates the superiority of the proposed method.&lt;/p&gt;

Single-channel blind signal separation and blind estimation of symbol sequences of communication signals

A new method for the blind separation of multiple communication signals received by a single sensor is presented. In this paper based on the different rate and waveform of communication signals, we employs oversampling and high order statistics to turn a single de-mixing problem into the estimation problem of the waveform and symbol that can be solved by means of a conventional ICA (Independent component analysis) algorithm. Simulation results demonstrate the effectiveness and superiority of the proposed algorithm.

A novel method for device-related electroencephalography artifact suppression to explore cochlear implant-related cortical changes in single-sided deafness

BackgroundQuantitative electroencephalography (qEEG) is effective when used to analyze ongoing cortical oscillations in cochlear implant (CI) users. However, localization of cortical activity in such users via qEEG is confounded by the presence of artifacts produced by the device itself. Typically, independent component analysis (ICA) is used to remove CI artifacts in auditory evoked EEG signals collected upon brief stimulation and it is effective for auditory evoked potentials (AEPs). However, AEPs do not reflect the daily environments of patients, and thus, continuous EEG data that are closer to such environments are desirable. In this case, device-related artifacts in EEG data are difficult to remove selectively via ICA due to over-completion of EEG data removal in the absence of preprocessing. New methodsEEGs were recorded for a long time under conditions of continuous auditory stimulation. To obviate the over-completion problem, we limited the frequency of CI artifacts to a significant characteristic peak and apply ICA artifact removal. ResultsTopographic brain mapping results analyzed via band-limited (BL)-ICA exhibited a better energy distribution, matched to the CI location, than data obtained using conventional ICA. Also, source localization data verified that BL-ICA effectively removed CI artifacts. Comparison with existing methodThe proposed method selectively removes CI artifacts from continuous EEG recordings, while ICA removal method shows residual peak and removes important brain activity signals. ConclusionCI artifacts in EEG data obtained during continuous passive listening can be effectively removed with the aid of BL-ICA, opening up new EEG research possibilities in subjects with CIs.

Efficient online target speech extraction using DOA-constrained independent component analysis of stereo data for robust speech recognition

This paper describes an efficient online target-speech-extraction method used as a preprocessing step for robust automatic speech recognition (ASR). Because a target speaker is located relatively close to microphones in many ASR applications, acoustic paths to microphones are moderately reverberant, and the target speaker direction can easily be estimated. In this situation, noise estimation is effectively performed by forming a directional null to the target speaker. Required weights for extracting target speech, independent of the estimated noise, are then determined using an adaptation rule derived from a modified version of the cost function for independent component analysis (ICA), while retaining the minimal distortion principle. In particular, an online natural-gradient learning rule with a nonholonomic constraint and normalization by a smoothed power estimate of the input signal is derived for stable convergence, even for dynamically changing speech levels, with much less computational complexity than conventional ICA. Furthermore, stereo mixtures are considered as input data for further reduction of computational loads and fast convergence. Although the method may suffer from the underdetermined problem, the weights are adapted to obtain signal-to-noise-ratio-maximization beamformers for successful target speech estimation. The experimental results obtained for various conditions demonstrate the effectiveness of the proposed method.