Artifact Removal Method Research Articles

Correction of the physiological artefacts at pre-surgical clinical functional MR

Pre-surgical functional MRI (fMRI) is an important modality of examinations before brain surgery. There are several artefacts (e.g. motion, susceptibility) which may hinder the evaluation of fMRI data. Physiological artefacts (breathing, pulsation) also affect the sensitivity and specificity of anatomical localization. The aim of this study is to demonstrate the efficiency of physiological artefact identification and removal methods for presurgical evaluation. Siemens Magnetom Verio 3T MRI scanner was used to collect data. The physiological parameters (breathing, pulse) were recorded with the MRI system's built-in devices. Data from fourteen patients - with primary brain tumour - were evaluated with SPM12 utilizing the RETROICOR/RVHR tool to detect and decrease the effect of physiological artefacts. We compared the statistical maps obtained with and without the physiological correction using the Jaccard similarity coefficient, and ROI analyses. Significant differences were found in the mean ROI values (p<0.0016) and the extensions of eloquent activations (p<0.0013), when using the physiological correction (RETORICOR/RVHR) based on convolution method. On the other hand, no significant differences were found between the ROIs' standard deviations (F=0.28). The RETROICOR/ RVHR method helps to define the precise localisation of eloquent areas (p<0.009). The number of irrelevant (non-significant) voxels were increased (p<0.001). . Minimising of physiological artefacts in fMRI data calculations, the (RETROICOR/RVHR) method based on convolution has been successfully adapted. This algorithm could be helpful before neurosurgical intervention. The activity pattern became more reliable.

Learning a Single Model with a Wide Range of Quality Factors for JPEG Image Artifacts Removal.

Lossy compression brings artifacts into the compressed image and degrades the visual quality. In recent years, many compression artifacts removal methods based on convolutional neural network (CNN) have been developed with great success. However, these methods usually train a model based on one specific value or a small range of quality factors. Obviously, if the test images quality factor does not match to the assumed value range, then degraded performance will be resulted. With this motivation and further consideration of practical usage, a highly robust compression artifacts removal network is proposed in this paper. Our proposed network is a single model approach that can be trained for handling a wide range of quality factors while consistently delivering superior or comparable image artifacts removal performance. To demonstrate, we focus on the JPEG compression with quality factors, ranging from 1 to 60. Note that a turnkey success of our proposed network lies in the novel utilization of the quantization tables as part of the training data. Furthermore, it has two branches in parallel-i.e., the restoration branch and the global branch. The former effectively removes the local artifacts, such as ringing artifacts removal. On the other hand, the latter extracts the global features of the entire image that provides highly instrumental image quality improvement, especially effective on dealing with the global artifacts, such as blocking, color shifting. Extensive experimental results performed on color and grayscale images have clearly demonstrated the effectiveness and efficacy of our proposed single-model approach on the removal of compression artifacts from the decoded image.

Image Denoising and Ring Artifacts Removal for Spectral CT via Deep Neural Network

The spectral computed tomography (CT) based on photon counting detectors can collect the incident photons with different energy ranges. However, due to the low photon counts in narrow energy bin and the unhomogeneous response problem of detector cells, there are severe noise and ring artifacts in reconstructed spectral CT images. We proposed an image denoising and ring artifacts removal method via improved Fully Convolutional Pyramid Residual Network (FCPRN). In our study, we scanned a mouse specimen with spectral CT based on photon counting detector, and reconstructed mouse CT images as data set. Then we use the data set to train our network for image denoising and ring artifacts removal. Experimental results demonstrated that the proposed method could reduce noise and suppress ring artifacts of spectral CT images concurrently in different energy ranges. And the performance of the FCPRN is better than that of some networks for CT image denoising.

Removing Ring Artefacts for Photon-Counting Detectors Using Neural Networks in Different Domains

The development of energy-resolving photon-counting detectors provides a new approach for obtaining spectral information in computed tomography. However, the responses of different photon counting detector pixels can be inconsistent, which will always cause stripe artefacts in projection domain and concentric ring artefacts in image domain. Traditional ring artifacts processing methods are mostly based on averaging and filtering. In this paper, we propose to use deep learning methods for ring artifacts removal respectively in image domain, projection domain and the polar coordinate system. Besides, by incorporating reconstruction process into neural networks, we unite the information from image domain and projection domain for ring artifacts removal under the framework of deep learning for the first time. A traditional ring artifacts removal method, which is based on wavelet and Fourier transform, is implemented for comparison. Quantitative analysis is performed on simulation and experimental results and it shows that deep learning based methods are promising in solving the problem of non-uniformity correction for photon-counting detectors.

Motion Artifact Removal in Ambulatory ECG Signal for Heart Rate Variability Analysis

This work presents an efficient method for motion artifact removal from ambulatory electrocardiogram (ECG) signal for heart rate variability (HRV) in wearable/portable healthcare devices. HRV is the fluctuation in the time interval between the adjacent heartbeats. Motion artifacts affect HRV analysis by creating some outliers. A two-phase method using stationary wavelet transform with level thresholding (SWT-LT) is used to remove motion artifact from the ECG signal. Multi-channel system prototype is used for ambulatory ECG signal recording which is developed using commercial integrated circuit components. Motion artifact affected ECG signals are recorded by emulating daily activity movements. Recorded ECG database (60 signals) and Motion Artifact Contaminated ECG Database (27 signals) are used for validation of the proposed SWT-LT method. Implemented results show that the proposed SWT-LT method removes various in-band motion artifacts efficiently with an average correlation coefficient of 0.9337 and an average normalized mean square error of 0.012 which are better than the other reported methods. The proposed method has shown improvement in features of HRV analysis by removing outliers due to motion artifact from the ECG signal which is verified using MATLAB app HRVTool 1.03 developed by Marcus Vollmer.

Improved EOG Artifact Removal Using Wavelet Enhanced Independent Component Analysis.

Electroencephalography (EEG) signals are frequently contaminated with unwanted electrooculographic (EOG) artifacts. Blinks and eye movements generate large amplitude peaks that corrupt EEG measurements. Independent component analysis (ICA) has been used extensively in manual and automatic methods to remove artifacts. By decomposing the signals into neural and artifactual components and artifact components can be eliminated before signal reconstruction. Unfortunately, removing entire components may result in losing important neural information present in the component and eventually may distort the spectral characteristics of the reconstructed signals. An alternative approach is to correct artifacts within the independent components instead of rejecting the entire component, for which wavelet transform based decomposition methods have been used with good results. An improved, fully automatic wavelet-based component correction method is presented for EOG artifact removal that corrects EOG components selectively, i.e., within EOG activity regions only, leaving other parts of the component untouched. In addition, the method does not rely on reference EOG channels. The results show that the proposed method outperforms other component rejection and wavelet-based EOG removal methods in its accuracy both in the time and the spectral domain. The proposed new method represents an important step towards the development of accurate, reliable and automatic EOG artifact removal methods.

Removal of computed tomography ring artifacts via radial basis function artificial neural networks

Ring artifacts in computed tomography (CT) images are caused by the undesirable response of detector pixels, which leads to the degradation of CT images. Accordingly, it affects the image interpretation, post-processing, and quantitative analysis. In this study, a radial basis function neural network (RBFNN) was used to remove ring artifacts. The proposed method employs polar coordinate transformation. First, ring artifacts were transformed into linear artifacts by polar coordinate transformation. Then, smoothing operators were applied to locate these artifacts exactly. Subsequently, RBFNN was operated on each linear artifact. The neuron numbers of the input, hidden, and output layers of the neural network were 8, 40, and 1, respectively. Neurons in the input layer were selected according to the characteristics of the artifact itself and its relationship with the surrounding normal pixels. For the training of the neural network, a hybrid of adaptive gradient descent algorithm (AGDA) and gravitational search algorithm (GSA) was adopted. After the corrected image was obtained using the updated neural network, the inverse coordinate transformation was implemented. The experimental data were divided into simulated ring artifacts and real ring artifacts, which were based on brain and abdomen CT images. Compared with current artifact removal methods, the proposed method removed ring artifacts more effectively and retained the maximum detail of normal tissues. In addition, for index analysis, the performance of proposed method was superior to that of the other methods.

Comparison of two methods of removing EOG artifacts for use in a motor imagery-based brain computer interface

Blind Source Separation (BSS) methods are signal processing tools that are widely utilized for Electroencephalogram (EEG) data analysis. The BSS separates a set of recorded EEG signals into a set of components. Note that the EEG measurements are often contaminated with various types of artifacts such as eye movements and blinks (Electrooculogram, EOG) that make data analysis very difficult. Therefore, we present a comparative study between two methods of removing EOG artifacts to be use in a Motor Imagery-based Brain Computer Interface (MI-BCI). In literature, a method of artifact removal based on Independent Component Analysis (ICA) using the concepts of Renyi’s entropy and kurtosis has been presented that we call this method ICA (the first method). However, applying the ICA method to EEG signal directly can cause the significant distortion of the clean EEG signal because this method completely removes independent components capturing artifacts. In this study, we improve this method by applying Wavelet Denoising (WD) only to segments affected by artifacts from the detected artifactual components instead of the complete removal of these components. This method is called ICA–WD (the second method). The quality of removing the artifact is evaluated using the Relative Root-Mean-Squared Error and Average Correlation Coefficient criteria which are computed between the original and processed signals. The results show that the ICA–WD removes the EOG artifacts better than the ICA. Next, we apply the ICA–WD method to dataset 2a of BCI competition IV for the EOG artifact removal. Then, the cleaned EEG signals are employed for feature extraction and classification. The results indicate that the proposed method outperforms the three best competitors and another study performed on the BCI competition IV dataset 2a.

A Novel Motion Artifact Removal Method via Joint Basis Pursuit Linear Program to Accurately Monitor Heart Rate

Photoplethysmography (PPG)-based heart rate (HR) estimation during physical exercise is challenging as PPG signals are often contaminated by motion artifacts (MA). This study develops a novel HR estimation method to effectively attenuate the impact of MA on PPG signals and accurately identify HR variations during physical exercise. First, a new signal reconstruction procedure is implemented based on a joint basis pursuit linear program (BPLP) to decompose PPG signal into different time series. Furthermore, an adaptive MA removal technique is developed, where the correlation between the acceleration signals and PPG time series are calculated and used as a reference to eliminate MA. Then, a new sparse spectra reconstruction method is designed to rebuild the spectrum of the current window based on the previous time frame. Furthermore, a simple HR estimation method with only one tuning parameter is designed to select the HR associated peak from the reconstructed spectra. Finally, a postprocessing technique is applied to further boost the accuracy of detection. The performance of the proposed algorithm is compared with three popular methods in recent studies using both training and testing sets from 2015 IEEE Signal Processing Cup. The proposed method provides the average absolute error of 1.79 beats per minutes (BPM) on all 22 recordings. With respect to testing datasets with stronger MA, the average absolute error is computed as 2.61BPM. The proposed HR tracking algorithm shows good robustness as it only involves a small set of parameters and can provide accurate estimations when PPG signals are contaminated by strong MA.

Automated artifact rejection algorithms harm P3 Speller brain-computer interface performance

ABSTRACTBrain-Computer Interfaces (BCIs) have been used to restore communication and control to people with severe paralysis. However, noninvasive BCIs based on electroencephalogram (EEG) are particularly vulnerable to noise artifacts. These artifacts, including electro-oculogram (EOG), can be orders of magnitude larger than the signal to be detected. Many automated methods have been proposed to remove EOG and other artifacts from EEG recordings, most based on blind source separation. This work presents a performance comparison of ten different automated artifact removal methods. Unfortunately, all tested methods substantially and significantly reduced P3 Speller BCI performance, and all methods were more likely to reduce performance than increase it. The least harmful methods were titled SOBI, JADER, and EFICA, but even these methods caused an average of approximately ten percentage points drop in BCI accuracy. Possible mechanistic causes for this empirical performance reduction are proposed.

GEVD Based on Multichannel Wiener Filter for Removal of EEG Artifacts

The electroencephalography (EEG) signals are contaminated by ocular artifacts usually called as ElectroOculoGraphy(EOG) artifacts. This occurs due to an eye movement and repeatedly blinking eyes, it is a major barrier to overcome when analyzing ElectroEncephaloGram (EEG) data. In this paper, Generalized Eigen Value Decomposition (GEVD) algorithm based on Multichannel Wiener filter (MWF) was proposed. In the GEVD algorithm, the covariance matrix of the artifact is identified and substituted by low rank approximation. For both real and hybrid EEG data is demonstrated using this algorithm and also compared with other existing methods for removal of artifacts. This paper determines generic, robust and fast algorithm for artifact removal of various types of EEG signals. Signal to Error Ratio (SER) and Artifact to Residue Ratio (ARR) both are expressed in dBs. The better performance of artifact removal is expressed with high SER which measures clean EEG distortion and ARR measures the artifact estimation.

A 9.2-g Fully-Flexible Wireless Ambulatory EEG Monitoring and Diagnostics Headband With Analog Motion Artifact Detection and Compensation.

An 8-channel wearable wireless device for ambulatory surface EEG monitoring and analysis is presented. The entire multi-channel recording, quantization, and motion artifact removal circuitries are implemented on a 4-layer polyimide flexible substrate. The recording electrodes and active shielding are also integrated on the same substrate, yielding the smallest form factor compared to the state of the art. Thanks to the dry non-contact electrodes, the system is quickly mountable with minimal assistance required, making it an ideal ambulatory front- and temporal-lobe EEG monitoring device. The flexible main board is connected to a rechargeable battery on one end and to a 13×17mm 2 rigid board on the other end. The mini rigid board hosts a low-power programmable FPGA and a BLE 5.0 transceiver, which add diagnostic capability and wireless connectivity features to the device, respectively. Design considerations for a wearable EEG monitoring and diagnostic device are discussed in details. The theory of the novel fully-analog method for motion artifact detection and removal is described and the detailed circuit implementation is presented. The device performance in terms of voltage gain (260 V/V), bandwidth (DC-300 Hz), motion artifact removal, and wireless communication throughput (up to 1Mbps) is experimentally validated. The entire wearable solution with the battery weighs 9.2 grams.

Removal of Muscle Artifacts From the EEG: A Review and Recommendations

Electroencephalography (EEG) has been widely used for studying brain function. As cortical signals recorded by the EEG are very weak, they are often obscured by motion artifacts and various non-brain physiological activities, such as eye movement, heart rhythms, and muscle activity, adversely affecting subsequent analysis and interpretation. Over the past decades, a number of techniques have been developed for preprocessing EEG recordings to improve the signal-to-noise ratio. However, based on our extensive literature survey, despite an increasing trend recently, only 14.72% of published studies on EEG preprocessing were involved with the removal of electromygraphic (EMG) artifacts. Given that ambulatory healthcare systems are continuously emerging, artifacts induced by muscle contraction become unavoidable, whereas in the past data tended to be collected in well-controlled clinical/laboratory settings. Motivated by the fact that EMG artifact removal is becoming an important issue to be addressed, we investigated the state-of-the-art muscle artifact removal methods systematically and comparatively, from the perspective of signal processing. In this review, we first present the signal characteristics from brain and muscle activity, and highlight the importance of this issue for subsequent artifact removal. We then provide an overview and taxonomy of representative methods. Based on the results in reported studies, we describe the pros and cons of different methods and give suggestions on selecting a suitable technique in different scenarios (e.g., single-channel, few-channel, and multichannel). Finally, we discuss remaining challenges and provide feasible recommendations for further exploration in this field.

ReMAE: User-Friendly Toolbox for Removing Muscle Artifacts From EEG

This paper describes a user-friendly toolbox, ReMAE, for removing muscle artifacts from electroencephalogram (EEG), running under the MATLAB environment. It implements a series of state-of-the-art methods for muscle artifact removal from EEG in the literature, and provides a graphical user interface (GUI). According to the taxonomy of the existing studies, this toolbox contains three denoising modes based on the number of input EEG channels, i.e., multi-channel, single-channel, and few-channel. Furthermore, this toolbox modularizes the denoising methods and visualizes each module. This means that users can readily observe the detailed denoising performance in each step, and even design a customized combined method in terms of their own understanding. In the current literature, there exists no method applicable for all situations due to the complexity of muscle artifacts. The main motivation of this work is to connect neuroscientists, psychologists, and clinicians with both the well-established and cutting-edge methods through a simple and intuitive GUI, and encourage them to extensively investigate different methods in a variety of real scenarios.

Faster Gait Speeds Reduce Alpha and Beta EEG Spectral Power From Human Sensorimotor Cortex.

ObjectiveOur aim was to determine if walking speed affected human sensorimotor electrocortical dynamics using mobile high-density electroencephalography (EEG).MethodsTo overcome limitations associated with motion and muscle artifact contamination in EEG recordings, we compared solutions for artifact removal using novel dual layer EEG electrodes and alternative signal processing methods. Dual layer EEG simultaneously recorded human electrocortical signals and isolated motion artifacts using pairs of mechanically coupled and electrically independent electrodes. For electrical muscle activity removal, we incorporated electromyographic (EMG) recordings from the neck into our mobile EEG data processing pipeline. We compared artifact removal methods during treadmill walking at four speeds (0.5, 1.0, 1.5, and 2.0 m/s).ResultsLeft and right sensorimotor alpha and beta spectral power increased in contralateral limb single support and push off, and decreased during contralateral limb swing at each speed. At faster walking speeds, sensorimotor spectral power fluctuations were less pronounced across the gait cycle with reduced alpha and beta power (p<0.05) compared to slower speeds. Isolated noise recordings and neck EMG spectral power fluctuations matched gait events and showed broadband spectral power increases at faster speeds.Conclusion and significanceDual layer EEG enabled us to isolate changes in human sensorimotor electrocortical dynamics across walking speeds. A comparison of signal processing approaches revealed similar intrastride cortical fluctuations when applying common (e.g. Artifact Subspace Reconstruction) and novel artifact rejection methods. Dual layer EEG, however, allowed us to document and rule out residual artifacts, which exposed sensorimotor spectral power changes across gait speeds.

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 Cardiac Interference (ARCI): A New Approach for EEG Data.

EEG recordings are generally affected by interference from physiological and non-physiological sources which may obscure underlying brain activity and hinder effective EEG analysis. In particular, cardiac interference can be caused by the electrical activity of the heart and/or cardiovascular activity related to blood flow. Successful EEG application in sports science settings requires a method for artifact removal that is automatic and flexible enough to be applied in a variety of acquisition conditions without requiring simultaneous ECG recordings that could restrict movement. We developed an automatic method for classifying and removing both electrical cardiac and cardiovascular artifacts (ARCI) that does not require additional ECG recording. Our method employs independent component analysis (ICA) to isolate data independent components (ICs) and identifies the artifactual ICs by evaluating specific IC features in the time and frequency domains. We applied ARCI to EEG datasets with cued artifacts and acquired during an eyes-closed condition. Data were recorded using a standard EEG wet cap with either 128 or 64 electrodes and using a novel dry electrode cap with either 97 or 64 dry electrodes. All data were decomposed into different numbers of components to evaluate the effect of ICA decomposition level on effective cardiac artifact detection. ARCI performance was evaluated by comparing automatic ICs classifications with classifications performed by experienced investigators. Automatic and investigator classifications were highly consistent resulting in an overall accuracy greater than 99% in all datasets and decomposition levels, and an average sensitivity greater than 90%. Best results were attained when data were decomposed into a fewer number of components where the method achieved perfect sensitivity (100%). Performance was also evaluated by comparing automatic component classification with externally recorded ECG. Results showed that ICs automatically classified as artifactual were significantly correlated with ECG activity whereas the other ICs were not. We also assessed that the interference affecting EEG signals was reduced by more than 82% after automatic artifact removal. Overall, ARCI represents a significant step in the detection and removal of cardiac-related EEG artifacts and can be applied in a variety of acquisition settings making it ideal for sports science applications.

U-net-based blocked artifacts removal method for dynamic computed tomography.

Airplane engines are vital aircraft components, so regular inspections of the engines are required to ensure their stable operation. A dynamic computed tomography (CT) system has been proposed by our group for in situ nondestructive testing of airplane engines, which takes advantage of the rotor's self-rotation. However, static parts of the engines cause blocked artifacts in the reconstructed image, leading to misinterpretations of the condition of engines. In this paper, in order to remove the artifacts produced by the projection of the static parts in CT reconstruction, two deep-learning-based methods are proposed, which use U-Net to perform correction in the projection domain. The projection of static parts can be estimated by a well-trained U-Net and subsequently can be subtracted from the projections of the engine. Finally, the rotor can be reconstructed from the corrected projections. The results shown in this paper indicate that the proposed methods are practical and effective for removing those blocked artifacts and recovering the details of rotating parts, which will, in turn, maximize the utilization of the dynamic CT system for in situ engine tests.

Continuous Stress Detection Using Wearable Sensors in Real Life: Algorithmic Programming Contest Case Study.

The negative effects of mental stress on human health has been known for decades. High-level stress must be detected at early stages to prevent these negative effects. After the emergence of wearable devices that could be part of our lives, researchers have started detecting extreme stress of individuals with them during daily routines. Initial experiments were performed in laboratory environments and recently a number of works took a step outside the laboratory environment to the real-life. We developed an automatic stress detection system using physiological signals obtained from unobtrusive smart wearable devices which can be carried during the daily life routines of individuals. This system has modality-specific artifact removal and feature extraction methods for real-life conditions. We further tested our system in a real-life setting with collected physiological data from 21 participants of an algorithmic programming contest for nine days. This event had lectures, contests as well as free time. By using heart activity, skin conductance and accelerometer signals, we successfully discriminated contest stress, relatively higher cognitive load (lecture) and relaxed time activities by using different machine learning methods.

Adaptive Artifact Removal From Intracortical Channels for Accurate Decoding of a Force Signal in Freely Moving Rats.

Intracortical data recorded with multi-electrode arrays provide rich information about kinematic and kinetic states of movement in the brain–machine interface (BMI) systems. Direct estimation of kinetic information such as the force from cortical data has the same importance as kinematic information to make a functional BMI system. Various types of the information including single unit activity (SUA), multiunit activity (MUA) and local field potential (LFP) can be used as an input information to extract motor commands for control of the external devices in BMI. Here we combine LFP and MUA information to improve decoding accuracy of the force signal from the multi-channel intracortical data of freely moving rats. We suggest a weighted common average referencing (CAR) algorithm in order to valid interpretation of the force decoding from different data types. The proposed spatial filter adaptively identifies contribution of the common noise on the channels employing Kalman filter method. We evaluated the efficacy of the proposed artifact algorithm on both simulation and real data. In the simulation study, the average R2 between the original and reconstructed signal of all channels after applying the proposed artifact removal method was computed for input SNRs in the range of −45 to 0 dB. Weighted CAR method can effectively reconstruct the original signal with average R2 higher than 0.5 for input SNRs higher than −s10 dB in case of adding simulated outlier and motion artifacts. We also show that the proposed artifact removal algorithm 33% improves the accuracy of force decoding in terms of R2 value compared to standard CAR filters.