Showing papers on "Artifact (error) published in 2018"
TL;DR: The Harvard Automated Processing Pipeline for EEG (HAPPE) is proposed as a standardized, automated pipeline compatible with EEG recordings of variable lengths and artifact contamination levels, including high-artifact and short EEG recordings from young children or those with neurodevelopmental disorders.
Abstract: Electroenchephalography (EEG) recordings collected with developmental populations present particular challenges from a data processing perspective. These EEGs have a high degree of artifact contamination and often short recording lengths. As both sample sizes and EEG channel densities increase, traditional processing approaches like manual data rejection are becoming unsustainable. Moreover, such subjective approaches preclude standardized metrics of data quality, despite the heightened importance of such measures for EEGs with high rates of initial artifact contamination. There is presently a paucity of automated resources for processing these EEG data and no consistent reporting of data quality measures. To address these challenges, we propose the Harvard Automated Processing Pipeline for EEG (HAPPE) as a standardized, automated pipeline compatible with EEG recordings of variable lengths and artifact contamination levels, including high-artifact and short EEG recordings from young children or those with neurodevelopmental disorders. HAPPE processes event-related and resting-state EEG data from raw files through a series of filtering, artifact rejection, and re-referencing steps to processed EEG suitable for time-frequency-domain analyses. HAPPE also includes a post-processing report of data quality metrics to facilitate the evaluation and reporting of data quality in a standardized manner. Here, we describe each processing step in HAPPE, perform an example analysis with EEG files we have made freely available, and show that HAPPE outperforms seven alternative, widely-used processing approaches. HAPPE removes more artifact than all alternative approaches while simultaneously preserving greater or equivalent amounts of EEG signal in almost all instances. We also provide distributions of HAPPE's data quality metrics in an 867 file dataset as a reference distribution and in support of HAPPE's performance across EEG data with variable artifact contamination and recording lengths. HAPPE software is freely available under the terms of the GNU General Public License at https://github.com/lcnhappe/happe.
428 citations
TL;DR: This editorial provides some constructive guidance across different positioning statements with actionable recommendations for DSR authors and reviewers to serve as a foundational step towards clarifying misconceptions about DSR contributions.
Abstract: With the rising interest in Design Science Research (DSR), it is crucial to engage in the ongoing debate on what constitutes an acceptable contribution for publishing DSR the design artifact, the design theory, or both. In this editorial, we provide some constructive guidance across different positioning statements with actionable recommendations for DSR authors and reviewers. We expect this editorial to serve as a foundational step towards clarifying misconceptions about DSR contributions and to pave the way for the acceptance of more DSR papers to top IS journals.
397 citations
TL;DR: In this paper, a deep residual learning network is proposed to remove aliasing artifacts from artifact corrupted images, which can work as an iterative k-space interpolation algorithm using framelet representation.
Abstract: Objective: Accelerated magnetic resonance (MR) image acquisition with compressed sensing (CS) and parallel imaging is a powerful method to reduce MR imaging scan time. However, many reconstruction algorithms have high computational costs. To address this, we investigate deep residual learning networks to remove aliasing artifacts from artifact corrupted images. Methods: The deep residual learning networks are composed of magnitude and phase networks that are separately trained. If both phase and magnitude information are available, the proposed algorithm can work as an iterative k-space interpolation algorithm using framelet representation. When only magnitude data are available, the proposed approach works as an image domain postprocessing algorithm. Results: Even with strong coherent aliasing artifacts, the proposed network successfully learned and removed the aliasing artifacts, whereas current parallel and CS reconstruction methods were unable to remove these artifacts. Conclusion: Comparisons using single and multiple coil acquisition show that the proposed residual network provides good reconstruction results with orders of magnitude faster computational time than existing CS methods. Significance: The proposed deep learning framework may have a great potential for accelerated MR reconstruction by generating accurate results immediately.
375 citations
18 Jul 2018
TL;DR: This study systematically validates ASR on ten EEG recordings in a simulated driving experiment and shows that the optimal ASR parameter is between 10 and 100, which is small enough to remove activities from artifacts and eye- related components and large enough to retain signals from brain-related components.
Abstract: One of the greatest challenges that hinder the decoding and application of electroencephalography (EEG) is that EEG recordings almost always contain artifacts - non-brain signals. Among existing automatic artifact-removal methods, artifact subspace reconstruction (ASR) is an online and realtime capable, component-based method that can effectively remove transient or large-amplitude artifacts. However, the effectiveness of ASR and the optimal choice of its parameter have not been evaluated and reported, especially on real EEG data. This study systematically validates ASR on ten EEG recordings in a simulated driving experiment. Independent component analysis (ICA) is applied to separate artifacts from brain signals to allow a quantitative assessment of ASR's effectiveness in removing various types of artifacts and preserving brain activities. Empirical results show that the optimal ASR parameter is between 10 and 100, which is small enough to remove activities from artifacts and eye-related components and large enough to retain signals from brain-related components. With the appropriate choice of the parameter, ASR can be a powerful and automatic artifact removal approach for offline data analysis or online real-time EEG applications such as clinical monitoring and brain-computer interfaces.
300 citations
TL;DR: This protocol can be used to reduce motion-related variance to near zero in studies of functional connectivity, providing up to a 100-fold improvement over minimal-processing approaches in large datasets.
Abstract: Participant motion during functional magnetic resonance image (fMRI) acquisition produces spurious signal fluctuations that can confound measures of functional connectivity. Without mitigation, motion artifact can bias statistical inferences about relationships between connectivity and individual differences. To counteract motion artifact, this protocol describes the implementation of a validated, high-performance denoising strategy that combines a set of model features, including physiological signals, motion estimates, and mathematical expansions, to target both widespread and focal effects of subject movement. This protocol can be used to reduce motion-related variance to near zero in studies of functional connectivity, providing up to a 100-fold improvement over minimal-processing approaches in large datasets. Image denoising requires 40 min to 4 h of computing per image, depending on model specifications and data dimensionality. The protocol additionally includes instructions for assessing the performance of a denoising strategy. Associated software implements all denoising and diagnostic procedures, using a combination of established image-processing libraries and the eXtensible Connectivity Pipeline (XCP) software. Ciric et al. describe a protocol for the removal of motion artifacts from functional MRI data. They introduce a software package that implements common denoising protocols and provides tools for assessing the efficacy of denoising.
298 citations
TL;DR: This paper demonstrates a fast, robust and generic algorithm for removal of EEG artifacts of various types, i.e. those that were annotated as unwanted by the user, with better performance than current state-of-the-art methods.
Abstract: Objective The electroencephalogram (EEG) is an essential neuro-monitoring tool for both clinical and research purposes, but is susceptible to a wide variety of undesired artifacts. Removal of these artifacts is often done using blind source separation techniques, relying on a purely data-driven transformation, which may sometimes fail to sufficiently isolate artifacts in only one or a few components. Furthermore, some algorithms perform well for specific artifacts, but not for others. In this paper, we aim to develop a generic EEG artifact removal algorithm, which allows the user to annotate a few artifact segments in the EEG recordings to inform the algorithm. Approach We propose an algorithm based on the multi-channel Wiener filter (MWF), in which the artifact covariance matrix is replaced by a low-rank approximation based on the generalized eigenvalue decomposition. The algorithm is validated using both hybrid and real EEG data, and is compared to other algorithms frequently used for artifact removal. Main results The MWF-based algorithm successfully removes a wide variety of artifacts with better performance than current state-of-the-art methods. Significance Current EEG artifact removal techniques often have limited applicability due to their specificity to one kind of artifact, their complexity, or simply because they are too 'blind'. This paper demonstrates a fast, robust and generic algorithm for removal of EEG artifacts of various types, i.e. those that were annotated as unwanted by the user.
264 citations
TL;DR: It is shown that the CNN-based information can be displayed in a novel artifact-free image format, enabling us to effectively remove reflection artifacts from photoacoustic images, which is not possible with traditional geometry-based beamforming.
Abstract: Interventional applications of photoacoustic imaging typically require visualization of point-like targets, such as the small, circular, cross-sectional tips of needles, catheters, or brachytherapy seeds. When these point-like targets are imaged in the presence of highly echogenic structures, the resulting photoacoustic wave creates a reflection artifact that may appear as a true signal. We propose to use deep learning techniques to identify these types of noise artifacts for removal in experimental photoacoustic data. To achieve this goal, a convolutional neural network (CNN) was first trained to locate and classify sources and artifacts in pre-beamformed data simulated with $k$ -Wave. Simulations initially contained one source and one artifact with various medium sound speeds and 2-D target locations. Based on 3,468 test images, we achieved a 100% success rate in classifying both sources and artifacts. After adding noise to assess potential performance in more realistic imaging environments, we achieved at least 98% success rates for channel signal-to-noise ratios (SNRs) of −9dB or greater, with a severe decrease in performance below −21dB channel SNR. We then explored training with multiple sources and two types of acoustic receivers and achieved similar success with detecting point sources. Networks trained with simulated data were then transferred to experimental waterbath and phantom data with 100% and 96.67% source classification accuracy, respectively (particularly when networks were tested at depths that were included during training). The corresponding mean ± one standard deviation of the point source location error was 0.40 ± 0.22 mm and 0.38 ± 0.25 mm for waterbath and phantom experimental data, respectively, which provides some indication of the resolution limits of our new CNN-based imaging system. We finally show that the CNN-based information can be displayed in a novel artifact-free image format, enabling us to effectively remove reflection artifacts from photoacoustic images, which is not possible with traditional geometry-based beamforming.
238 citations
TL;DR: This paper presents an extensive overview of the existing methods for ocular, muscle, and cardiac artifact identification and removal with their comparative advantages and limitations and reviewed the schemes developed for validating the performances of algorithms with simulated and real EEG data.
Abstract: Electroencephalogram (EEG), boasting the advantages of portability, low cost, and high-temporal resolution, is a non-invasive brain-imaging modality that can be used to measure different brain states. However, EEG recordings are always contaminated with artifacts from different sources other than neurons, which renders EEG data analysis more difficult, and which potentially results in misleading findings. Therefore, it is essential for many medical and practical applications to remove these artifacts in the preprocessing stage before analyzing EEG data. In the last thirty years, various methods have been developed to remove different types of artifacts from contaminated EEG data; still though, there is no standard method that can be used optimally, and therefore, the research remains attractive as well as challenging. This paper presents an extensive overview of the existing methods for ocular, muscle, and cardiac artifact identification and removal with their comparative advantages and limitations. We also reviewed the schemes developed for validating the performances of algorithms with simulated and real EEG data. In future studies, researchers should focus not only on the combining of different methods with multiple processing stages for efficient removal of artifactual interferences but also on the development of standard criteria for validation of recorded EEG signals.
191 citations
TL;DR: A hybrid method that takes advantage of different correction algorithms for hemodynamic response function estimation using the following metrics: mean-squared error, peak-to-peak error, Pearson’s correlation, and the area under the receiver operator characteristic curve is found.
Abstract: Motion artifact contamination in near-infrared spectroscopy (NIRS) data has become an important challenge in realizing the full potential of NIRS for real-life applications. Various motion correction algorithms have been used to alleviate the effect of motion artifacts on the estimation of the hemodynamic response function. While smoothing methods, such as wavelet filtering, are excellent in removing motion-induced sharp spikes, the baseline shifts in the signal remain after this type of filtering. Methods, such as spline interpolation, on the other hand, can properly correct baseline shifts; however, they leave residual high-frequency spikes. We propose a hybrid method that takes advantage of different correction algorithms. This method first identifies the baseline shifts and corrects them using a spline interpolation method or targeted principal component analysis. The remaining spikes, on the other hand, are corrected by smoothing methods: Savitzky-Golay (SG) filtering or robust locally weighted regression and smoothing. We have compared our new approach with the existing correction algorithms in terms of hemodynamic response function estimation using the following metrics: mean-squared error, peak-to-peak error ([Formula: see text]), Pearson's correlation ([Formula: see text]), and the area under the receiver operator characteristic curve. We found that spline-SG hybrid method provides reasonable improvements in all these metrics with a relatively short computational time. The dataset and the code used in this study are made available online for the use of all interested researchers.
182 citations
TL;DR: The proposed method, called MEMD-CCA, first utilizes MEMD and CCA to jointly decompose the few-channel EEG recordings into multivariate intrinsic mode functions (IMFs), and is applied to further decomposes the reorganized multivariate IMFs into the underlying sources.
Abstract: Electroencephalography (EEG) recordings are often contaminated by muscle artifacts. In the literature, a number of methods have been proposed to deal with this problem. Yet most denoising muscle artifact methods are designed for either single-channel EEG or hospital-based, high-density multichannel recordings, not the few-channel scenario seen in most ambulatory EEG instruments. In this paper, we propose utilizing interchannel dependence information seen in the few-channel situation by combining multivariate empirical mode decomposition and canonical correlation analysis (MEMD-CCA). The proposed method, called MEMD-CCA, first utilizes MEMD to jointly decompose the few-channel EEG recordings into multivariate intrinsic mode functions (IMFs). Then, CCA is applied to further decompose the reorganized multivariate IMFs into the underlying sources. Reconstructing the data using only artifact-free sources leads to artifact-attenuated EEG. We evaluated the performance of the proposed method through simulated, semisimulated, and real data. The results demonstrate that the proposed method is a promising tool for muscle artifact removal in the few-channel setting.
162 citations
TL;DR: The RL-ARCNN indicates that residual learning of CNN remarkably reduces metal artifacts and improves critical structure visualization and confidence of radiation oncologists in target delineation.
Abstract: Cervical cancer is the fifth most common cancer among women, which is the third leading cause of cancer death in women worldwide. Brachytherapy is the most effective treatment for cervical cancer. For brachytherapy, computed tomography (CT) imaging is necessary since it conveys tissue density information which can be used for dose planning. However, the metal artifacts caused by brachytherapy applicators remain a challenge for the automatic processing of image data for image-guided procedures or accurate dose calculations. Therefore, developing an effective metal artifact reduction (MAR) algorithm in cervical CT images is of high demand. A novel residual learning method based on convolutional neural network (RL-ARCNN) is proposed to reduce metal artifacts in cervical CT images. For MAR, a dataset is generated by simulating various metal artifacts in the first step, which will be applied to train the CNN. This dataset includes artifact-insert, artifact-free, and artifact-residual images. Numerous image patches are extracted from the dataset for training on deep residual learning artifact reduction based on CNN (RL-ARCNN). Afterwards, the trained model can be used for MAR on cervical CT images. The proposed method provides a good MAR result with a PSNR of 38.09 on the test set of simulated artifact images. The PSNR of residual learning (38.09) is higher than that of ordinary learning (37.79) which shows that CNN-based residual images achieve favorable artifact reduction. Moreover, for a 512 × 512 image, the average removal artifact time is less than 1 s. The RL-ARCNN indicates that residual learning of CNN remarkably reduces metal artifacts and improves critical structure visualization and confidence of radiation oncologists in target delineation. Metal artifacts are eliminated efficiently free of sinogram data and complicated post-processing procedure.
1 Mar 2018
TL;DR: A new data-driven algorithm to effectively remove ocular and muscular artifacts from single-channel EEG: the surrogate-based artifact removal (SuBAR), which provides a relative error 4 to 5 times lower than traditional techniques.
Abstract: Objective: the recent emergence and success of electroencephalography (EEG) in low-cost portable devices, has opened the door to a new generation of applications processing a small number of EEG channels for health monitoring and brain-computer interfacing. These recordings are, however, contaminated by many sources of noise degrading the signals of interest, thus compromising the interpretation of the underlying brain state. In this paper, we propose a new data-driven algorithm to effectively remove ocular and muscular artifacts from single-channel EEG: the surrogate-based artifact removal (SuBAR). Methods: by means of the time-frequency analysis of surrogate data, our approach is able to identify and filter automatically ocular and muscular artifacts embedded in single-channel EEG. Results: in a comparative study using artificially contaminated EEG signals, the efficacy of the algorithm in terms of noise removal and signal distortion was superior to other traditionally-employed single-channel EEG denoizing techniques: wavelet thresholding and the canonical correlation analysis combined with an advanced version of the empirical mode decomposition. Even in the presence of mild and severe artifacts, our artifact removal method provides a relative error 4 to 5 times lower than traditional techniques. Significance: in view of these results, the SuBAR method is a promising solution for mobile environments, such as ambulatory healthcare systems, sleep stage scoring, or anesthesia monitoring, where very few EEG channels or even a single channel is available.
16 Sep 2018
TL;DR: Results of a comparative study of the artifact subspace re-construction (ASR) method and two other popular methods dedicated to correct EEG artifacts show a significantly better level of artifact correction for the ASR method.
Abstract: The paper presents the results of a comparative study of the artifact subspace re-construction (ASR) method and two other popular methods dedicated to correct EEG artifacts: independent component analysis (ICA) and principal component analysis (PCA). The comparison is based on automatic rejection of EEG signal epochs performed on a dataset of motor imagery data. ANOVA results show a significantly better level of artifact correction for the ASR method. What is more, the ASR method does not cause serious signal loss compared to other methods.
TL;DR: This paper describes a novel methodology leveraging particle filters for the application of robust heart rate monitoring in the presence of motion artifacts, and formulate the heart rate itself as the only state to be estimated, and do not rely on multiple specific signal features.
Abstract: This paper describes a novel methodology leveraging particle filters for the application of robust heart rate monitoring in the presence of motion artifacts. Motion is a key source of noise that confounds traditional heart rate estimation algorithms for wearable sensors due to the introduction of spurious artifacts in the signals. In contrast to previous particle filtering approaches, we formulate the heart rate itself as the only state to be estimated, and do not rely on multiple specific signal features. Instead, we design observation mechanisms to leverage the known steady, consistent nature of heart rate variations to meet the objective of continuous monitoring of heart rate using wearable sensors. Furthermore, this independence from specific signal features also allows us to fuse information from multiple sensors and signal modalities to further improve estimation accuracy. The signal processing methods described in this work were tested on real motion artifact affected electrocardiogram and photoplethysmogram data with concurrent accelerometer readings. Results show promising average error rates less than 2 beats/min for data collected during intense running activities. Furthermore, a comparison with contemporary signal processing techniques for the same objective shows how the proposed implementation is also computationally more efficient for comparable performance.
28 Apr 2018
TL;DR: From the study, it is shown that combination of conventional with other methods is popularly used, as it is able to improve the removal of artifacts, and makes use of machine learning to provide an automated solution with higher efficiency.
Abstract: Electroencephalogram (EEG) is a signal collected from the human brain to study and analyze the brain activities. However, raw EEG may be contaminated with unwanted components such as noises and artifacts caused by power source, environment, eye blinks, heart rate and muscle movements, which are unavoidable. These unwanted components will effect the analysis of EEG and provide inaccurate information. Therefore, researchers have proposed all kind of approaches to eliminate unwanted noises and artifacts from EEG. In this paper, a literature review is carried out to study the works that have been done for noise and artifact removal from year 2010 up to the present. It is found that conventional approaches include ICA, wavelet based analysis, statistical analysis and others. However, the existing ways of artifacts removal cannot eliminate certain noise and will cause information lost by directly discard the contaminated components. From the study, it is shown that combination of conventional with other methods is popularly used, as it is able to improve the removal of artifacts. The current trend of artifacts removal makes use of machine learning to provide an automated solution with higher efficiency.
TL;DR: Two sparsity-based techniques namely morphological component analysis (MCA) and K-SVD-based artifact removal method have been evaluated and it is shown that without using any computationally expensive algorithms, only with the use of over-complete dictionaries the proposed sparsity to eliminate EB artifacts accurately from the EEG signals.
Abstract: Neural activities recorded using electroencephalography (EEG) are mostly contaminated with eye blink (EB) artifact. This results in undesired activation of brain–computer interface (BCI) systems. Hence, removal of EB artifact is an important issue in EEG signal analysis. Of late, several artifact removal methods have been reported in the literature and they are based on independent component analysis (ICA), thresholding, wavelet transformation, etc. These methods are computationally expensive and result in information loss which makes them unsuitable for online BCI system development. To address the above problems, we have investigated sparsity-based EB artifact removal methods. Two sparsity-based techniques namely morphological component analysis (MCA) and K-SVD-based artifact removal method have been evaluated in our work. MCA-based algorithm exploits the morphological characteristics of EEG and EB using predefined Dirac and discrete cosine transform (DCT) dictionaries. Next, in K-SVD-based algorithm an overcomplete dictionary is learned from the EEG data itself and is designed to model EB characteristics. To substantiate the efficacy of the two algorithms, we have carried out our experiments with both synthetic and real EEG data. We observe that the K-SVD algorithm, which uses a learned dictionary, delivers superior performance for suppressing EB artifacts when compared to MCA technique. Finally, the results of both the techniques are compared with the recent state-of-the-art FORCe method. We demonstrate that the proposed sparsity-based algorithms perform equal to the state-of-the-art technique. It is shown that without using any computationally expensive algorithms, only with the use of over-complete dictionaries the proposed sparsity-based algorithms eliminate EB artifacts accurately from the EEG signals.
Uppsala University1, Royal Institute of Technology2, University of Oxford3, Academy of Sciences of the Czech Republic4, Chalmers University of Technology5, European XFEL6, Technical University of Berlin7, Argonne National Laboratory8, SLAC National Accelerator Laboratory9, Science for Life Laboratory10, Brookhaven National Laboratory11, Lawrence Berkeley National Laboratory12
TL;DR: A three-dimensional reconstruction of the Melbournevirus affected by a strong artifact in the center of the particle was found to be probably caused by background scattering, while particle size and pulse-energy variation did not affect the quality of the reconstruction.
11 Apr 2018
TL;DR: The proposed convolutional neural networks framework enables the network to generalize well to both unseen simulated motion artifacts as well as real motion artifact-affected data and could easily be adapted to estimate a motion severity score, which could be used as a score of quality control or as a nuisance covariate in subsequent statistical analyses.
Abstract: Head motion during MRI acquisition presents significant problems for subsequent neuroimaging analyses. In this work, we propose to use convolutional neural networks (CNNs) to correct motion-corrupted images as well as investigate a possible improvement by augmenting L1 loss with adversarial loss. For training, in order to gain access to a ground-truth, we first selected a large number of motion-free images from the ABIDE dataset. We then added simulated motion artifacts on these images to produce motion corrupted data and a 3D regression CNN was trained to predict the motion-free volume as the output. We tested the CNN on unseen simulated data as well as real motion affected data. Quantitative evaluation was carried out using metrics such as Structural Similarity (SSIM) index, Correlation Coefficient (CC), and Tissue Contrast T-score (TCT). It was found that Gaussian smoothing as a conventional method did not significantly differ in SSIM, CC and RMSE from the uncorrected data. On the other hand, the two CNN models successfully removed the motion-related artifact as their SSIM and CC significantly increased after their correction and the error was reduced. The CNN displayed significantly larger TCT compared to the uncorrected images whereas the adversarial network, while improved did not show a significantly increased TCT, which may be explained also by its over-enhancement of edges. Our results suggest that the proposed CNN framework enables the network to generalize well to both unseen simulated motion artifacts as well as real motion artifact-affected data. The proposed method could easily be adapted to estimate a motion severity score, which could be used as a score of quality control or as a nuisance covariate in subsequent statistical analyses. ∗USC Stevens Neuroimaging and Informatics Institute, University of Southern California, CA 90033 †Department of Computer Engineering and Computer Science, California State University Long Beach, Long Beach, CA 90840 1st Conference on Medical Imaging with Deep Learning (MIDL 2018), Amsterdam, The Netherlands.
TL;DR: A novel software pipeline of real-time image processing suited for closed-loop experiments and a novel method to estimate baseline calcium signal using kernel density estimate, which reduces the number of parameters to be tuned.
Abstract: Two-photon calcium imaging has been extensively used to record neural activity in the brain. It has been long used solely with post-hoc analysis, but the recent efforts began to include closed-loop experiments. Closed-loop experiments pose new challenges because they require fast, real-time image processing without iterative parameter tuning. When imaging awake animals, one of the crucial steps of post hoc image analysis is correction of lateral motion artifacts. In most of the closed-loop experiments, this step has not been implemented and ignored due to technical difficulties. We recently reported the first experiments with real-time processing of calcium imaging that included lateral motion correction. Here, we report the details of the implementation of fast motion correction and present performance analysis across several algorithms with different parameters. Additionally, we introduce a novel method to estimate baseline calcium signal using kernel density estimate, which reduces the number of parameters to be tuned. Combined, we propose a novel software pipeline of real-time image processing suited for closed-loop experiments. The pipeline is also useful for rapid post hoc image processing.
TL;DR: A systematic, integrated methodology for removing MA from EEG signals recorded during treadmill (TM) and over-ground (OG) walking is introduced, and the approach is offered as a unified methodology for MA removal from EEG collected during gait trials.
Abstract: The appreciation for the need to record electroencephalographic (EEG) signals from humans while walking has been steadily growing in recent years, particularly in relation to understanding gait disturbances. Movement artefacts (MA) in EEG signals originate from mechanical forces applied to the scalp electrodes, inducing small electrode movements relative to the scalp which, in turn, cause the recorded voltage to change irrespectively of cortical activity. These mechanical forces, and thus MA, may have various sources (e.g., ground reaction forces, head movements, etc.) that are inherent to daily activities, notably walking. In this paper we introduce a systematic, integrated methodology for removing MA from EEG signals recorded during treadmill (TM) and over-ground (OG) walking, as well as quantify the prevalence of MA in different locomotion settings. In our experiments, participants performed walking trials at various speeds both OG and on a TM while wearing a 32-channel EEG cap and a 3-axis accelerometer, placed on the forehead. Data preprocessing included separating the EEG signals into statistically independent additive components using independent component analysis (ICA). We observed an increase in electro-physiological signals (e.g., neck EMG activations for stabilizing the head during heel-strikes) as the walking speed increased. These artefact independent-components (ICs), while not originating from electrode movement, still exhibit a similar spectral pattern to the MA ICs–a peak at the stepping frequency. MA was identified and quantified in each component using a novel method that utilizes the participant’s stepping frequency, derived from a forehead-mounted accelerometer. We then benchmarked the EEG data by applying newly established metrics to quantify the success of our method in cleaning the data. The results indicate that our approach can be successfully applied to EEG data recorded during TM and OG walking, and is offered as a unified methodology for MA removal from EEG collected during gait trials.
TL;DR: This work systematically review and discusses the methods available for removing DBS artifacts, and highlights an open-source toolbox incorporating most artifact removal methods, allowing users to combine different strategies.
TL;DR: A real-time artifact removal algorithm that is based on canonical correlation analysis (CCA), feature extraction, and the Gaussian mixture model (GMM) to improve the quality of EEG signals is proposed.
Abstract: Electroencephalogram (EEG) signals are usually contaminated with various artifacts, such as signal associated with muscle activity, eye movement, and body motion, which have a noncerebral origin. The amplitude of such artifacts is larger than that of the electrical activity of the brain, so they mask the cortical signals of interest, resulting in biased analysis and interpretation. Several blind source separation methods have been developed to remove artifacts from the EEG recordings. However, the iterative process for measuring separation within multichannel recordings is computationally intractable. Moreover, manually excluding the artifact components requires a time-consuming offline process. This work proposes a real-time artifact removal algorithm that is based on canonical correlation analysis (CCA), feature extraction, and the Gaussian mixture model (GMM) to improve the quality of EEG signals. The CCA was used to decompose EEG signals into components followed by feature extraction to extract representative features and GMM to cluster these features into groups to recognize and remove artifacts. The feasibility of the proposed algorithm was demonstrated by effectively removing artifacts caused by blinks, head/body movement, and chewing from EEG recordings while preserving the temporal and spectral characteristics of the signals that are important to cognitive research.
TL;DR: An efficient wavelet-based method for artifacts attenuation while minimizing distortions, using a stationary wavelet transform (SWT) modeling the wavelet coefficients as a Laplace distribution is proposed, and it outperforms existing approaches and it has a lower computational cost.
TL;DR: It is argued that BSS-REG may enable the development of novel BCI applications requiring high-density recordings, such as source-based neurofeedback and closed-loop neuromodulation, because of its low computational requirements.
Abstract: Objective The performance of brain-computer interfaces (BCIs) based on electroencephalography (EEG) data strongly depends on the effective attenuation of artifacts that are mixed in the recordings. To address this problem, we have developed a novel online EEG artifact removal method for BCI applications, which combines blind source separation (BSS) and regression (REG) analysis. Approach The BSS-REG method relies on the availability of a calibration dataset of limited duration for the initialization of a spatial filter using BSS. Online artifact removal is implemented by dynamically adjusting the spatial filter in the actual experiment, based on a linear regression technique. Main results Our results showed that the BSS-REG method is capable of attenuating different kinds of artifacts, including ocular and muscular, while preserving true neural activity. Thanks to its low computational requirements, BSS-REG can be applied to low-density as well as high-density EEG data. Significance We argue that BSS-REG may enable the development of novel BCI applications requiring high-density recordings, such as source-based neurofeedback and closed-loop neuromodulation.
TL;DR: An overview of the most common MRI artifacts and methods to fix or rectify them is presented and the original artifacts images and statistics from the Lithuanian University of Health Sciences Kaunas Clinical Hospital, Dept of Radiology are provided.
Abstract: Different kinds of artifacts can occur during a magnetic resonance imaging (MRI) scans due to hardware or software related problems, human physiologic phenomenon or physical restrictions. Some of them can seriously affecting diagnostic image quality, while others may simulate or be confused with different pathology. On another word artifact as an artificial feature appearing in an image that is not present in the original investigative object. It is important to recognize these artifacts according to a basic understanding of their origin, especially those mimicking pathology, as they can lead to incorrect diagnosis and cause serious after-effects on patient’s health and outcomes. We presented an overview of the most common MRI artifacts and methods to fix or rectify them. We also provide the original artifacts images and statistics from the Lithuanian University of Health Sciences Kaunas Clinical Hospital, Dept of Radiology, mainly obtained from image databases and some images from data base of other Lithuanian hospitals.
TL;DR: The relative sensitivity of 38 commonly used HRV measures to artifact was compared to determine which measures show the most change with increasing increments of artifact, and whether short‐term and long‐term HRV Measures share similarities in their sensitivity to artifact.
Abstract: Background
Artifact is common in cardiac RR interval data derived from 24-hr recordings and has a significant impact on heart rate variability (HRV) measures. However, the relative impact of progressively added artifact on a large group of commonly used HRV measures has not been assessed. This study compared the relative sensitivity of 38 commonly used HRV measures to artifact to determine which measures show the most change with increasing increments of artifact. A secondary aim was to ascertain whether short-term and long-term HRV measures, as groups, share similarities in their sensitivity to artifact.
Methods
Up to 10% of artifact was added to 20 artificial RR (ARR) files and 20 human cardiac recordings, which had been assessed for artifact by a cardiac technician. The added artifact simulated deletion of RR intervals and insertion of individual short RR intervals. Thirty-eight HRV measures were calculated for each file. Regression analysis was used to rank the HRV measures according to their sensitivity to artifact as determined by the magnitude of slope.
Results
RMSSD, SDANN, SDNN, RR triangular index and TINN, normalized power and relative power linear measures, and most nonlinear methods examined are most robust to artifact.
Conclusion
Short-term time domain HRV measures are more sensitive to added artifact than long-term measures. Absolute power frequency domain measures across all frequency bands are more sensitive than normalized and relative frequency domain measures. Most nonlinear HRV measures assessed were relatively robust to added artifact, with Poincare plot SD1 being most sensitive.
TL;DR: An EEG artifact detection model, the Fingerprint Method, is developed which identifies different spatial, temporal, spectral, and statistical features indicative of physiological artifacts and uses these features to automatically classify artifactual independent components in EEG based on a machine leaning approach.
Abstract: Data contamination due to physiological artifacts such as those generated by eyeblinks, eye movements, and muscle activity continues to be a central concern in the acquisition and analysis of electroencephalographic (EEG) data. This issue is further compounded in EEG sports science applications where the presence of artifacts is notoriously difficult to control because behaviors that generate these interferences are often the behaviors under investigation. Therefore, there is a need to develop effective and efficient methods to identify physiological artifacts in EEG recordings during sports applications so that they can be isolated from cerebral activity related to the activities of interest. We have developed an EEG artifact detection model, the Fingerprint Method, which identifies different spatial, temporal, spectral, and statistical features indicative of physiological artifacts and uses these features to automatically classify artifactual independent components in EEG based on a machine leaning approach. Here, we optimized our method using artifact-rich training data and a procedure to determine which features were best suited to identify eyeblinks, eye movements, and muscle artifacts. We then applied our model to an experimental dataset collected during endurance cycling. Results reveal that unique sets of features are suitable for the detection of distinct types of artifacts and that the Optimized Fingerprint Method was able to correctly identify over 90% of the artifactual components with physiological origin present in the experimental data. These results represent a significant advancement in the search for effective means to address artifact contamination in EEG sports science applications.
TL;DR: A quantitative and automated support vector machine (SVM)-based algorithm to accurately classify artifactual EEG epochs in awake rodent, canine and humans subjects and enables the determination of 'eyes open/closed' states in human subjects is developed.
[...]
TL;DR: This chapter presents an overview of the methods available for each process and discusses practical considerations for applying these methods to the EEG signals, with considerable attention paid to the state-of-the-art artifact removal methods.
Abstract: Preprocessing of the EEG signal, which is virtually a set of signal processing steps preceding main EEG data analyses, is essential to obtain only brain activity from the noisy EEG recordings. It has been shown that the design of preprocessing procedures can affect subsequent EEG data analysis outcomes. Preprocessing of EEG largely includes a number of processes, such as line noise removal, adjustment of referencing, elimination of bad EEG channels, and artifact removal. This chapter presents an overview of the methods available for each process and discusses practical considerations for applying these methods to the EEG signals. In particular, considerable attention is paid to the state-of-the-art artifact removal methods since there are still plenty of opportunities to enhance the artifact removal techniques for EEG, in the perspectives of both signal processing and neuroscience. It is desirable that this chapter provides the readers an overall view of EEG preprocessing pipelines and serves as a handbook guide for the practice of EEG preprocessing.
Proceedings Article•
1 Jan 2018TL;DR: An unifying meta-algorithm is developed for learning beam search policies using imitation learning that captures existing learning algorithms and suggests new ones and lets us show novel no-regret guarantees for learning beams search policies.
Abstract: Beam search is widely used for approximate decoding in structured prediction problems. Models often use a beam at test time but ignore its existence at train time, and therefore do not explicitly learn how to use the beam. We develop an unifying meta-algorithm for learning beam search policies using imitation learning. In our setting, the beam is part of the model and not just an artifact of approximate decoding. Our meta-algorithm captures existing learning algorithms and suggests new ones. It also lets us show novel no-regret guarantees for learning beam search policies.