Inverse Independent Component Analysis Research Articles

Hybrid time‐reassigned multisynchrosqueezing transform‐Picard‐based automated electroencephalography artifact correction methodology for brain–computer interface applications

AbstractThe electroencephalographic activity is often contaminated with non‐cerebral origins resulting in erroneous classification in brain–computer interface (BCI) applications. In the present work, a novel automated Electroencephalogram artifact correction methodology is proposed. Preconditional independent component analysis (ICA) is utilized for de‐mixing the electroencephalographic activity into original components. The sparse entropy‐based thresholding criterion is developed to automatically identify the artifactual components and the time‐reassigned multisynchrosqueezing transform coefficients are estimated. The transform coefficients are processed to retain leaked cerebral activity into artifactual components. Finally, inverse transformation and inverse ICA are performed to attain artifact‐free electroencephalographic activity. The efficiency of the proposed methodology is validated on simulated and real electroencephalographic activity. The results illustrate the efficacy of the proposed methodology on both simulated and real electroencephalography activities in retaining useful cerebral information and rejecting artifactual origins, with a potential to be used in various BCI applications.

Automatic ECG Artefact Removal from EEG Signals

Abstract Electroencephalography (EEG) signals are frequently contaminated by ocular, muscle, and cardiac artefacts whose removal normally requires manual inspection or the use of reference channels (EOG, EMG, ECG). We present a novel, fully automatic method for the detection and removal of ECG artefacts that works without a reference ECG channel. Independent Component Analysis (ICA) is applied to the measured data and the independent components are examined for the presence of QRS waveforms using an adaptive threshold-based QRS detection algorithm. Detected peaks are subsequently classified by a rule-based classifier as ECG or non-ECG components. Components manifesting ECG activity are marked for removal, and then the artefact-free signal is reconstructed by removing these components before performing the inverse ICA. The performance of the proposed method is evaluated on a number of EEG datasets and compared to results reported in the literature. The average sensitivity of our ECG artefact removal method is above 99 %, which is better than known literature results.

Automatic Removal of Ocular Artifact from EEG with DWT and ICA Method

Ocular artifact (OA) is one of the main interferences in electroencephalogram (EEG) recordings. It appears as a big pulse and has a strong impact to EEG signals. To overcome OA interference in EEG data, a novel automatic method of OA removal, denoted as DWICA, was proposed in this paper. In DWICA, the discrete wavelet transform (DWT) is applied to every recorded signal to obtain multiple scale coefficients. Then the independent component analysis (ICA) algorithm is used, and its input is the coefficients connected in series. Thus the independent components are acquired quickly in wavelet domain. The criterion of angle cosine is introduced to recognize ocular artifact, and the corresponding component is set to zero. Furthermore, the artifact free components are projected to original electrodes with inverse ICA algorithm. Finally, DWT is inversed to obtain the artifact free brain signals. Quantitative studies about suppression of OA distorting the underlying cerebral activity and accurate evaluation on denoising effect of DWICA are finished in this paper. Experiment results show that DWICA is preferable and effective in automatic OA correction. Meanwhile, DWICA is powerful in noise immunity and fast in convergence rate, and it provides a preferable method for EEG preprocessing on-line.

Automatic Removal of Artifacts from Attention Deficit Hyperactivity Disorder Electroencephalograms Based on Independent Component Analysis

In this paper, we present an automatic artifact removal method by combining independent component analysis (ICA) and wavelet denoising to remove artifacts from attention deficit hyperactivity disorder (ADHD) electroencephalograms (EEGs). First, the ICA algorithm is used to analyze EEG signals to get the independent components (ICs); then, the wavelet method is applied to the demixed ICs as an intermediate step; finally, the EEG data are reconstructed from the new ICs (that are obtained in the intermediate step) by using the inverse ICA. The experimental results show that the automatic artifact removal method can not only remove eye artifacts, muscle artifacts and some unknown physiological artifacts sources, but also efficiently keep the weak neural activities from strong background artifacts. The results can establish a basis for ADHD clinical diagnose study.

Unmixing Dynamic Fluorescence Diffuse Optical Tomography Images With Independent Component Analysis

Dynamic fluorescence diffuse optical tomography (D-FDOT) is important for drug delivery research. However, the low spatial resolution of FDOT and the complex kinetics of drug limit the ability of D-FDOT in resolving metabolic processes of drug throughout whole body of small animals. In this paper, we propose an independent component analysis (ICA)-based method to perform D-FDOT studies. When applied to D-FDOT images, ICA not only generates a set of independent components (ICs) which can illustrate functional structures with different kinetic behaviors, but also provides a set of associated time courses (TCs) which can represent normalized time courses of drug in corresponding functional structures. Further, the drug concentration in specific functional structure at different time points can be recovered by an inverse ICA transformation. To evaluate the performance of the proposed algorithm in the study of drug kinetics at whole-body level, simulation study and phantom experiment are both performed on a full-angle FDOT imaging system with line-shaped excitation pattern. In simulation study, the nanoparticle delivery of indocynaine green (ICG) throughout whole body of a digital mouse is simulated and imaged. In phantom experiment, four tubes containing different ICG concentrations are imaged and used to imitate the uptake and excretion of ICG in organs. The results suggest that we can not only illustrate ICG distributions in different functional structures, but also recover ICG concentrations in specific functional structure at different time points, when ICA is applied to D-FDOT images.

A novel remote sensing image fusion method based on independent component analysis

Many research papers have reported problems with existing fusion techniques. The most significant problem is that fusion images produce spectral distortion when they retain high spatial resolution. Independent component analysis (ICA) can eliminate low-order and also high-order redundancy for data. According to statistical theory, the most important information from data is always included in the statistical characteristic of high order. Therefore, independent components (ICs) resulting from the transform of the ICA provide information on the data that is otherwise hidden in the large data set, and describe the essential structure of the data. For a colour or false colour composite image, the ICs represent the main body, spectral and spatial detail information. In this study, a novel remote sensing image fusion method based on ICA is presented. By replacing the main body IC of a multispectral image with the vector of a panchromatic (PAN) image, the new ICs contain both spatial information of the PAN image and spectral characteristics of the multispectral image. Then an inverse ICA transformation (IIT) is performed to attain the fusion image. IKONOS and Enhanced Thematic Mapper Plus (ETM + ) images are used to evaluate the proposed method and others (hue–saturation–value (HSV), principal component analysis and wavelets). The fusion results are compared graphically, visually and statistically, and show that the proposed method can retain spatial and spectral information simultaneously, and has a better balance between them.

Detection of Extra Components of T Wave by Independent Component Analysis in Congenital Long-QT Syndrome

The main ECG criteria for the diagnosis of long-QT syndrome (LQTS) include abnormal T-wave morphology as well as prolonged QT interval. The T wave in LQTS probably includes additional components of the myocardial repolarization process, which are derived from aberrant ion currents. We investigated whether independent component analysis (ICA) can extract such abnormal repolarization components. Digital ECG data were obtained as a time series from 10 channels using 20 surface electrodes in 22 patients with genetically confirmed LQTS type 1 (LQT1) and 30 normal subjects. In each case, T-wave area was analyzed by radical ICA after noise reduction by the wavelet thresholding method. Furthermore, inverse ICA was applied to determine the origin of each independent component (IC). Radical ICA revealed that a T-wave consisted of 4 basic ICs in all control subjects, whereas ≥ 5 (mostly 6) ICs were identified in all 22 patients with LQT1. The extra ICs, which were not evident in normal subjects, were assumed to contribute to the formation of abnormal T-wave morphology. The extra ICs were identified even in patients with normal QTc values and in those taking β-blockers. Inverse ICA indicated that the additional ICs originate predominantly from the late phase of the T wave of the left ventricle. Extra ICs appear during repolarization in all patients with LQT1 but not in normal subjects. ICA is a potentially useful multivariate statistical method to differentiate patients with LQT1 from normal subjects.

Inverse Independent Component Analysis Facilitates Clarification of the Accessory Conductive Pathway of Wolf–Parkinson–White Syndrome Electrocardiogram

Our aim was to demonstrate a digital analyzing method that could extract the potential of early excitation derived from accessory conductive pathway (ACP) from fusion of the QRS complex wave of the electrocardiogram of Wolf-Parkinson-White (WPW) syndrome. A 13-year-old boy with WPW syndrome received successful catheter ablation therapy. ECG was recorded and analyzed using independent component analysis (ICA) and inverse independent component analysis (I-ICA), at pretherapy and posttherapy. We identified the ACP potential and the following potential spread to the ventricle. Results agreed with those of intracardiac mapping, locating the ACP in the left posterior side of the heart. ICA and I-ICA might be useful for noninvasive analysis of WPW syndrome ECG and other electrocardiac abnormalities.

Multivariate streamflow forecasting using independent component analysis

Seasonal forecasting of streamflow provides many benefits to society, by improving our ability to plan and adapt to changing water supplies. A common approach to developing these forecasts is to use statistical methods that link a set of predictors representing climate state as it relates to historical streamflow, and then using this model to project streamflow one or more seasons in advance based on current or a projected climate state. We present an approach for forecasting multivariate time series using independent component analysis (ICA) to transform the multivariate data to a set of univariate time series that are mutually independent, thereby allowing for the much broader class of univariate models to provide seasonal forecasts for each transformed series. Uncertainty is incorporated by bootstrapping the error component of each univariate model so that the probability distribution of the errors is maintained. Although all analyses are performed on univariate time series, the spatial dependence of the streamflow is captured by applying the inverse ICA transform to the predicted univariate series. We demonstrate the technique on a multivariate streamflow data set in Colombia, South America, by comparing the results to a range of other commonly used forecasting methods. The results show that the ICA‐based technique is significantly better at representing spatial dependence, while not resulting in any loss of ability in capturing temporal dependence. As such, the ICA‐based technique would be expected to yield considerable advantages when used in a probabilistic setting to manage large reservoir systems with multiple inflows or data collection points.