Classification Of Independent Components Research Articles

One out of ten independent components shows flipped polarity with poorer data quality: EEG database study.

Independent component analysis (ICA) is widely used today for scalp-recorded EEG analysis. One of the limitations of ICA-based analysis is polarity indeterminacy. It is not easy to find detailed documentations that explains engineering solutions of how the polarity indeterminacy is addressed in a given implementation. We investigated how it is implemented in the case of EEGLAB and also the relation between the outcome of the polarity determination and classification of independent components (ICs) in terms of the estimated nature of the sources (brain, muscle, eye, etc.) using an open database of n = 212 EEG dataset of resting state recordings. We found that (1) about 91% of ICs showed positive-dominant IC scalp topographies; (2) positive-dominant ICs were more associated with brain-originated signals; (3) positive-dominant ICs showed more radial (peaked at 10-30 degrees deviations from the radial axis) dipolar projection pattern with less residual variance from fitting the equivalent current dipole. In conclusion, using the EEGLAB's default ICA algorithm, one out of 10 ICs results in flipping its polarity to negative, which is associated with non-radial dipole orientation with higher residual variance. Thus, we determined EEGLAB biases toward positive polarity in decomposing high-quality brain ICs.

Is Brain Dynamics Preserved in the EEG After Automated Artifact Removal? A Validation of the Fingerprint Method and the Automatic Removal of Cardiac Interference Approach Based on Microstate Analysis.

The assessment of a method for removing artifacts from electroencephalography (EEG) datasets often disregard verifying that global brain dynamics is preserved. In this study, we verified that the recently introduced optimized fingerprint method and the automatic removal of cardiac interference (ARCI) approach not only remove physiological artifacts from EEG recordings but also preserve global brain dynamics, as assessed with a new approach based on microstate analysis. We recorded EEG activity with a high-resolution EEG system during two resting-state conditions (eyes open, 25 volunteers, and eyes closed, 26 volunteers) known to exhibit different brain dynamics. After signal decomposition by independent component analysis (ICA), the independent components (ICs) related to eyeblinks, eye movements, myogenic interference, and cardiac electromechanical activity were identified with the optimized fingerprint method and ARCI approach and statistically compared with the outcome of the expert classification of the ICs by visual inspection. Brain dynamics in two different groups of denoised EEG signals, reconstructed after having removed the artifactual ICs identified by either visual inspection or the automated methods, was assessed by calculating microstate topographies, microstate metrics (duration, occurrence, and coverage), and directional predominance (based on transition probabilities). No statistically significant differences between the expert and the automated classification of the artifactual ICs were found (p > 0.05). Cronbach’s α values assessed the high test–retest reliability of microstate parameters for EEG datasets denoised by the automated procedure. The total EEG signal variance explained by the sets of global microstate templates was about 80% for all denoised EEG datasets, with no significant differences between groups. For the differently denoised EEG datasets in the two recording conditions, we found that the global microstate templates and the sequences of global microstates were very similar (p < 0.01). Descriptive statistics and Cronbach’s α of microstate metrics highlighted no significant differences and excellent consistency between groups (p > 0.5). These results confirm the ability of the optimized fingerprint method and the ARCI approach to effectively remove physiological artifacts from EEG recordings while preserving global brain dynamics. They also suggest that microstate analysis could represent a novel approach for assessing the ability of an EEG denoising method to remove artifacts without altering brain dynamics.

Automatic Identification of Brain Independent Components in Electroencephalography Data Collected while Standing in a Virtually Immersive Environment - A Deep Learning-Based Approach.

Electroencephalography (EEG) is a commonly used method for monitoring brain activity. Automating an EEG signal processing pipeline is imperative to the exploration of real-time brain computer interface (BCI) applications. EEG analysis demands substantial training and time for removal of distinct unwanted independent components (ICs), generated via independent component analysis, corresponding to artifacts. The considerable subject-wise variations across these components motivates defining a procedural way to identify and eliminate these artifacts. We propose DeepIC-virtual, a convolutional neural network (CNN) deep learning classifier to automatically identify brain components in the ICs extracted from the subject's EEG data gathered while they are being immersed in a virtual reality (VR) environment. This work examined the feasibility of DL techniques to provide automated ICs classification on noisy and visually engaging upright stance EEG data. We collected the EEG data for six subjects while they were standing upright in a VR testing setup simulating pseudo-randomized variations in height and depth conditions and induced perturbations. An extensive 1432 IC representation images data set was generated and manually labelled via an expert as brain components or one of the six distinct removable artifacts. The supervised CNN architecture was utilized to categorize good brain ICs and bad artifactual ICs via generated images of topographical maps. Our model categorizing good versus bad IC topographical maps resulted in a binary classification accuracy and area under curve of 89.20% and 0.93 respectively. Despite significant imbalance, only 1 out of the 57 present brain ICs in the withheld testing set was miss-classified as an artifact. These results will hopefully encourage clinicians to integrate BCI methods and neurofeedback to control anxiety and provide a treatment of acrophobia, given the viability of automatic classification of artifactual ICs.

ICA-based denoising for ASL perfusion imaging

Arterial Spin Labelling (ASL) imaging derives a perfusion image by tracing the accumulation of magnetically labeled blood water in the brain. As the image generated has an intrinsically low signal to noise ratio (SNR), multiple measurements are routinely acquired and averaged, at a penalty of increased scan duration and opportunity for motion artefact. However, this strategy alone might be ineffective in clinical settings where the time available for acquisition is limited and patient motion are increased. This study investigates the use of an Independent Component Analysis (ICA) approach for denoising ASL data, and its potential for automation.72 ASL datasets (pseudo-continuous ASL; 5 different post-labeling delays: 400, 800, 1200, 1600, 2000 m s; total volumes = 60) were collected from thirty consecutive acute stroke patients. The effects of ICA-based denoising (manual and automated) where compared to two different denoising approaches, aCompCor, a Principal Component-based method, and Enhancement of Automated Blood Flow Estimates (ENABLE), an algorithm based on the removal of corrupted volumes. Multiple metrics were used to assess the changes in the quality of the data following denoising, including changes in cerebral blood flow (CBF) and arterial transit time (ATT), SNR, and repeatability. Additionally, the relationship between SNR and number of repetitions acquired was estimated before and after denoising the data.The use of an ICA-based denoising approach resulted in significantly higher mean CBF and ATT values (p < 0.001), lower CBF and ATT variance (p < 0.001), increased SNR (p < 0.001), and improved repeatability (p < 0.05) when compared to the raw data. The performance of manual and automated ICA-based denoising was comparable. These results went beyond the effects of aCompCor or ENABLE. Following ICA-based denoising, the SNR was higher using only 50% of the ASL-dataset collected than when using the whole raw data.The results show that ICA can be used to separate signal from noise in ASL data, improving the quality of the data collected. In fact, this study suggests that the acquisition time could be reduced by 50% without penalty to data quality, something that merits further study. Independent component classification and regression can be carried out either manually, following simple criteria, or automatically.

Deep Convolutional Neural Networks for Feature-Less Automatic Classification of Independent Components in Multi-Channel Electrophysiological Brain Recordings.

Interpretation of the electroencephalographic (EEG) and magnetoencephalographic (MEG) signals requires off-line artifacts removal. Since artifacts share frequencies with brain activity, filtering is insufficient. Blind source separation, mainly through independent component analysis (ICA), is the gold-standard procedure for the identification of artifacts in multi-dimensional recordings. However, a classification of brain and artifactual independent components (ICs) is still required. Since ICs exhibit recognizable patterns, classification is usually performed by experts' visual inspection. This procedure is time consuming and prone to errors. Automatic ICs classification has been explored, often through complex ICs features extraction prior to classification. Relying on deep-learning ability of self-extracting the features of interest, we investigated the capabilities of convolutional neural networks (CNNs) for off-line, automatic artifact identification through ICs without feature selection. A CNN was applied to spectrum and topography of a large dataset of few thousand samples of ICs obtained from multi-channel EEG and MEG recordings acquired during heterogeneous experimental settings and on different subjects. CNN performances, when applied to EEG, MEG, and combined EEG and MEG ICs, were explored and compared with state-of-the-art feature-based automatic classification. Beyond state-of-the-art automatic classification accuracies were demonstrated through cross validation (92.4% EEG, 95.4% MEG, 95.6% EEG+MEG). High CNN classification performances were achieved through heuristical selection of machinery hyperparameters and through the CNN self-selection of the features of interest. Considering the large data availability of multi-channel EEG and MEG recordings, CNNs may be suited for classification of ICs of multi-channel brain electrophysiological recordings.

Objective selection of epilepsy-related independent components from EEG data

BackgroundIndependent Component Analysis (ICA) is commonly used for the identification of sources of interest in electroencephalographic (EEG) data, but the selection of the relevant components remains an open issue depending on the specific application. New MethodWe propose a novel approach for the objective selection of epilepsy-related independent components (ICs) from EEG data collected during functional Magnetic Resonance Imaging (fMRI) acquisitions, called PROJection onto Independent Components (PROJIC). Inter-ictal epileptiform discharges (IEDs) are identified on a reference EEG dataset collected outside the MRI scanner by an expert neurophysiologist, and the resulting average IED is projected onto the IC space of the EEG data collected simultaneously with fMRI. The power of the IED projection is then used to inform a k-means clustering algorithm of the ICs, allowing for the classification of epilepsy-related ICs. Comparison with existing methodsThe performance of PROJIC was compared with two methods previously proposed for the objective selection of EEG ICs of interest, which are based on the explicit similarity of the ICs with spatio-temporal templates of the events of interest, instead of the projection power. ResultsThe proposed PROJIC method outperformed the others for both artificial and real data (19 datasets collected from 6 patients with drug-refractory focal epilepsy), with an average accuracy of 98.6%. ConclusionsThe ability of our method to accurately and objectively select epilepsy-related ICs makes it an important contribution for simultaneous EEG–fMRI epilepsy studies, with potential applications in the analysis of event-related EEG activity more generally, and also in EEG artefact correction.

Dimensionality of ICA in resting-state fMRI investigated by feature optimized classification of independent components with SVM.

Different machine learning algorithms have recently been used for assisting automated classification of independent component analysis (ICA) results from resting-state fMRI data. The success of this approach relies on identification of artifact components and meaningful functional networks. A limiting factor of ICA is the uncertainty of the number of independent components (NIC). We aim to develop a framework based on support vector machines (SVM) and optimized feature-selection for automated classification of independent components (ICs) and use the framework to investigate the effects of input NIC on the ICA results. Seven different resting-state fMRI datasets were studied. 18 features were devised by mimicking the empirical criteria for manual evaluation. The five most significant (p < 0.01) features were identified by general linear modeling and used to generate a classification model for the framework. This feature-optimized classification of ICs with SVM (FOCIS) framework was used to classify both group and single subject ICA results. The classification results obtained using FOCIS and previously published FSL-FIX were compared against manually evaluated results. On average the false negative rate in identifying artifact contaminated ICs for FOCIS and FSL-FIX were 98.27 and 92.34%, respectively. The number of artifact and functional network components increased almost linearly with the input NIC. Through tracking, we demonstrate that incrementing NIC affects most ICs when NIC < 33, whereas only a few limited ICs are affected by direct splitting when NIC is incremented beyond NIC > 40. For a given IC, its changes with increasing NIC are individually specific irrespective whether the component is a potential resting-state functional network or an artifact component. Using FOCIS, we investigated experimentally the ICA dimensionality of resting-state fMRI datasets and found that the input NIC can critically affect the ICA results of resting-state fMRI data.

On the interpretation of the independent components underlying the abdominal phonogram: a study of their physiological relevance

Recorded by positioning a sensitive acoustic sensor over the maternal womb, the abdominal phonogram is a signal that contains valuable information for foetal surveillance (e.g. heart rate), which is hidden by maternal and environmental sources. To recover such information, previous work used single-channel independent component analysis (SCICA) to separate the abdominal phonogram into statistically independent components (ICs) that, once acquired, must be objectively associated with the real sources underlying the abdominal phonogram—either physiological or environmental. This is a typical challenge for blind source separation methodologies and requires further research on the signals of interest to find a suitable solution. Here, we have conducted a joint study on 75 sets of ICs by means of statistical, spectral, complexity and time-structure analysis methods. As a result, valuable and consistent characteristics of the components separated from the abdominal phonogram by SCICA have been revealed: (1) the ICs are spectrally disjoint and sorted according to their frequency content, (2) only the ICs with lower frequency content present strong regular patterns and (3) such regular patterns are driven by well-known physiological processes given by the maternal breathing rate, the maternal heart rate and the foetal heart rate. This information was so promising that it has been used in current work for automatic classification of ICs and recovery of the traces of the physiological sources underlying the abdominal phonogram. Future work will look for the extraction of information useful for surveillance (e.g. heart rate), not only about foetal well-being, but also about maternal condition.