Baseline Wander Removal Research Articles

Prediction of QT interval in ECGs through baseline wander removal and deep learning technique

Abstract Background In electrocardiograms (ECGs), the QT interval, calculated from the onset of the Q-wave to the end of the T-wave, reflects the repolarization status of the heart. Currently, the QT interval automatically measured in 12-lead electrocardiograms is inaccurately derived as the median value, rather than utilizing the more precise mean value of QT intervals from all leads. This approach fails to provide information on heterogeneous QT prolongation crucial for predicting arrhythmias, as it does not accurately represent the values of QT intervals in each lead. Purpose Traditional analysis of ECGs based on QT intervals necessitates direct interpretation by cardiologists, leading to potential errors in diagnosis. In response to this challenge, we present a model that integrates low-pass filtering and continuous wavelet transformation to eliminate baseline wander noise occurring during ECG measurements. Moreover, to account for temporal features, we incorporate Convolutional Neural Networks (CNN) and Bidirectional Long Short-Term Memory (Bi-LSTM) to estimate QT intervals. Methods The dataset used in this study was the QT Database, comprising ECG data for 105 patients recorded at a sampling rate of 250Hz over a duration of 15 minutes. Each ECG was associated with annotations for two leads and information regarding the P, QRS, and T waves. We structured the dataset into a ratio of approximately 8:1:1, resulting in 51 patients for training (15,490 samples), 6 patients for validation (1,564 samples), and 7 patients for testing (1,682 samples). Results For the evaluation of deep learning, we selected Accuracy and F1-Score as metrics. We experimented with two configurations: the proposed model with two Bi-LSTM layers and an alternative with a single Bi-LSTM layer of 100 units. We also tested a similar recurrent neural network, GRU, with the same unit configuration. The results indicated that the method with two Bi-LSTM layers outperformed others in both Accuracy and F1-Score. Conclusion By incorporating CNN and Bi-LSTM, we conducted QT interval estimation, leveraging temporal features. This methodology facilitated the elimination of diverse noise types during ECG measurements, resulting in enhanced accuracy for QT interval estimation through the consideration of temporal ECG characteristics. In our future research endeavors, we plan to deploy this model for predicting arrhythmias, specifically targeting conditions like long QT syndrome and drug-induced QT prolongation.Methods and result

Comparative Evaluation of Neural Network Models for Optimizing ECG Signal in Non-Uniform Sampling Domain

Electrocardiographic signals (ECG) are ubiquitous, which justifies the research of their optimal storage and transmission. However, proposals for non-uniform signal sampling must take into account the priority of diagnostic data accuracy and record integrity, as well as robustness to noise and interference. In this study, two novel methods are introduced, each utilizing a distinct neural network architecture for optimizing non-uniform sampling of ECG signal. A transformer model refines each time point selection through an iterative process using gradient descent optimization, with the goal of minimizing the mean squared error between the original and resampled signals. It adaptively modifies time points, which improves the alignment between both signals. In contrast, the Temporal Convolutional Network model trains on the original signal, and gradient descent optimization is utilized to improve the selection of time points. Evaluation of both strategies’ efficacy is performed by calculating signal distances at lower and higher sampling rates. First, a collection of synthetic data points that resembled the P-QRS-T wave was used to train the model. Then, the ECG-ID database for real data analysis was used. Filtering to remove baseline wander followed by evaluation and testing were carried out in the real patient data. The results, in particular MSE = 0.0005, RMSE = 0.0216, and Pearson’s CC = 0.9904 for 120 sps in the case of the transformer patient data model, provide viable paths for maintaining the precision and dependability of ECG-based diagnostic systems at much lower sampling rate. Outcomes indicate that both techniques are effective at improving the fidelity between the original and modified ECG signals.

Denoising methods for removal of Baseline Wander, AWGN and power line interface noises in ECG signal: a comparative analysis

ABSTRACT The electrocardiogram (ECG) is one of the most significant signals in the field of biomedicine for the diagnosis of cardiac arrhythmia (CA). Due to non-stationary behaviour of the ECG signal, it is often interrupted with various noises, which cause difficulty in diagnosis. To solve the issues, the image denoising process helps the physicians make the perfect decision. In this paper, Denoising Techniques using Empirical Mode Decomposition (EMD) and Wavelet Transform for the removal of various noises like Additive White Gaussian Noise (AWGN), Baseline Wander (BW) noise, Power Line Interface (PLI) noise at various concentrations. The proposed method has quality and good efficiency in Peak Signal-to-Noise Ratio (PSNR), Root Mean Square Error (RSME), Signal-to-Noise Ratio (SNR), Cross-Correlation (CC) and Percent Root Square Difference (PRD).

Comparative Performance Analysis of Filtering Methods for Removing Baseline Wander Noise from an ECG Signal

ECG signals play a vital role in the diagnosis of cardiovascular conditions. However, they often suffer from the effects of various noise sources, including baseline wandering, respiratory artifact noise, power line interference and electrode motion artifacts. To overcome these challenges, it is imperative to implement low-frequency signal noise reduction strategies. Such strategies aim to significantly improve the quality of ECG signals, thus promoting more accurate and reliable diagnosis of cardiovascular disorders. This paper conducts a comparative analysis to assess the effectiveness of commonly used filtering and wavelet techniques in reducing Baseline Wander (BW) noise within ECG signals generated by the influence of breathing or electrode movements. It is common to observe the selection and evaluation of only one particular technique in the existing literature. In contrast, this study aims to provide a comprehensive comparative analysis, providing insight into the performance and relative merits of different techniques. Our research uses both filtering and Discrete Wavelet Transform (DWT) techniques in baseline noise removal. In this context, a reference point is established utilizing noise-free signals and a meticulous investigation of the wavelet-based approach that most effectively eliminates the resulting noise is provided. Subsequently, we assess the reference input and output signal via Signal-to-Noise Ratio (SNR) and Kolmogorov–Smirnov statistical test measurements. The most important contribution of this work to the scientific community resides in the comprehensive examination of IIR/FIR-based and wavelet method-based filtering methods capable of yielding the highest SNR levels across various ECG signals with various types of BW noise. Additionally, the effectiveness of the Chebychev-II filter in BW noise removal is highlighted. Our study was conducted using the MATLAB platform and code command lines were shared to facilitate the reproduction of our study by other researchers. It is considered that this study will be an important reference in the selection of effective techniques for removing BW noise within ECG signals.

A lightweight hybrid CNN-LSTM explainable model for ECG-based arrhythmia detection

Objective:Electrocardiogram (ECG) is the most frequent and routine diagnostic tool used for monitoring heart electrical signals and evaluating its functionality. The human heart can suffer from a variety of diseases, including cardiac arrhythmias. Arrhythmia is an irregular heart rhythm that in severe cases can lead to stroke and can be diagnosed via ECG recordings. Since early detection of cardiac arrhythmias is of great importance, computerized and automated classification and identification of these abnormal heart signals have received much attention for the past decades. Methods:This paper introduces a light Deep Learning (DL) approach for high accuracy detection of 8 different cardiac arrhythmias and normal rhythms. To employ DL techniques, the ECG signals were preprocessed using resampling and baseline wander removal techniques. The classification was performed using an 11-layer network employing a combination of Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM). Results:In order to evaluate the proposed technique, ECG signals are chosen from the two physionet databases, the MIT-BIH arrhythmia database and the long-term AF database. The proposed DL framework based on the combination of CNN and LSTM showed promising results than most of the state-of-the-art methods. The proposed method reaches the mean diagnostic accuracy of 98.24%. Conclusion:A trained model for arrhythmia classification using diverse ECG signals were successfully developed and tested. Significance:This study presents a lightweight classification technique with high diagnostic accuracy compared to other notable methods, making it a potential candidate for implementation in Holter monitor devices for arrhythmia detection. Finally, we used SHapley Additive exPlanations (SHAP), the most popular Explainable Artificial Intelligence (XAI) method to understand how our model make predictions. The results indicate that those features (ECG samples) that have contributed the most to a prediction are consonant with clinicians’ decisions. Therefore, the use of interpretable models increases the trust of clinicians in AI and thus leads to decreasing the number of misdiagnoses of cardiovascular diseases.

Automated detection of myocardial infarction using binary Harry Hawks feature selection and ensemble KNN classifier

Myocardial infarction (MI), referred to as a heart attack, is a life-threatening condition that happens due to blood clots, typically, blood flow to a portion of the heart muscle is blocked. The cardiac muscle may become permanently damaged if there is insufficient oxygen and blood flow to the affected area. It’s crucial to treat MI as soon as possible because even a small delay might have serious effects. The primary diagnostic tool to track and identify the signs of MI is the electrocardiogram (ECG). The complexity of MI signals combined with noise makes it difficult for clinicians to make a precise and prompt diagnosis. It might be laborious and time-consuming to manually analyse an enormous quantity of ECG data. Therefore, techniques for autonomously diagnosing from the ECG data are required. There have been numerous research on the topic of MI espial, but the majority of the algorithms are cognitively intensive when working with empirical data. The current study suggests a unique method for the efficient and reliable identification of MI. We employed circulant singular spectrum analysis (CSSA) for baseline wander removal, a 4-stage Savitzky–Golay (SG) filter to expunge powerline interference from the ECG signal and segmented in the preprocessing stage. Thus segmented ECG has been decomposed using CSSA, entropy based features are extracted. The best features are selected by using binary Harris hawk optimization (BHHO) and to machine learning (ML) classifiers like Naive Bayes, Decision tree, K-nearest neighbor (KNN), Support vector machine (SVM), and Ensemble subspace KNN. Our suggested method has been examined from both class as well as subject oriented perspectives. While the subject-oriented technique uses data from one patient for testing while using data from the other subjects for training, the class-wise strategy divides data as test data as well as training data regardless of subjects. We succeeded in achieving accuracy ( A c % ) of 99.8, sensitivity ( S e % ) of 99, and 100 specificity (Sp %) under the class-oriented approach. Similarly, for the subject wise strategy we achieved a mean A c % , Se %, and Sp % of 85.2, 83.1, and 84.5, respectively.

Baseline Wander Removal Applied to Smooth Pursuit Eye Movements From Parkinsonian Patients

Prior studies aiming to parametrize the sequences obtained from the Smooth Pursuit Eye Movements (SPEM) of patients with Parkinson’s disease are based on the manual extraction of cues of interest. This is because methods to automatically extract the relevant information are complex to implement and are constrained, in part, by the appearance of a baseline wander (BW). Thus, new methods are required for preprocessing the SPEM sequences to make the potential parameterisation procedures much more robust, removing the aforementioned BW. In this respect, the present study compares different BW removal methods applied to SPEM sequences based on several objective evaluation metrics. At the same time, it proposes a set of guidelines to estimate the ground truth that is required for comparison purposes. Data were collected using a high-speed video-based eye-tracking device. 52 patients and 60 controls and 12 young participants were enrolled in the study. The ground truth required to compare the different BW removal techniques was manually delineated according to a predefined protocol. Seven methods were developed to remove the BW, and four objective metrics were used to evaluate the results. According to the results, a method based on the Empirical Wavelet Transform provided the best performance removing the BW1. Furthermore, the objective and subjective results show that potential asymmetries between left and right eye movements are solved by removing the BW. Regardless of the techniques used, BW removal is revealed to be a crucial step for any autonomous SPEM processing tool.

A Dual-Adaptive Approach Based on Discrete Cosine Transform for Removal of ECG Baseline Wander

Removal of baseline wander (BW) is an important preprocessing step before manually or automatically interpreting electrocardiogram (ECG) records. It is a challenging issue to fully remove BW while preserving original clinical information because BW is usually mingled with low-frequency ECG components. A dual-adaptive approach based on discrete cosine transform (DCT) is presented in this study. Firstly, the cardiac fundamental frequency (CFF) of ECGs is accurately calculated through DCT domain analysis. Secondly, DCT coefficients of ECGs, whose frequencies are below CFF, are used to construct an amplitude vector in which the optimal cut-point between BW and ECGs is distinctly reflected. Finally, a new filtering technique based on DCT is exploited to suppress BW with its cutoff frequency adjusted to the optimal cut-point. The proposed method is applied to both real ECG records and simulated ECGs with its results compared to those of three previous methods published in the literature. The experimental results show that substantial improvements in performance can be achieved when adopting this dual-adaptive approach.

SpO2 Measurement: Non-Idealities and Ways to Improve Estimation Accuracy in Wearable Pulse Oximeters

The blood oxygen saturation level (SpO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ) has become one of the vital body parameters for the early detection, monitoring, and tracking of the symptoms of coronavirus diseases 2019 (COVID-19) and is clinically accepted for patient care and diagnostics. Pulse oximetry provides non-invasive SpO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> monitoring at home and ICUs without the need of a physician/doctor. However, the accuracy of SpO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> estimation in wearable pulse oximeters remains a challenge due to various non-idealities. We propose a method to improve the estimation accuracy by denoising the red and IR signals, detecting the signal quality, and providing feedback to hardware to adjust the signal chain parameters like LED current or transimpedance amplifier gain and enhance the signal quality. SpO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> is calculated using the red and infrared photoplethysmogram (PPG) signals acquired from the wrist using Texas Instruments AFE4950EVM. We introduce the green PPG signal as a reference to obtain the window size of the moving average filter for baseline wander removal and as a timing reference for peak and valley detection in the red and infrared PPG signals. We propose the improved peak and valley detection algorithm based on the incremental merge segmentation algorithm. Kurtosis, entropy, and Signal-to-noise ratio (SNR) are used as signal quality parameters, and SNR is further related to the variance in the SpO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> measurement. A closed-loop implementation is performed to enhance signal quality based on the signal quality parameters of the recorded PPG signals. The proposed algorithm aims to estimate SpO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> with a variance of 1% for the pulse oximetry devices.

Power line interference and baseline wander removal from ECG signals using local characteristic decomposition

Noise detection and removal are crucial in ECG processing. In this study, a method based on local characteristic decomposition (LCD) algorithm for denoising the ECG signals is discussed. In particular, specific techniques for the removal of baseline wandering (BW) and of power line interference (PLI), which are two common noises in ECG, are proposed and tested against real data. Simulation results show that the proposed method is more reliable at various signals to noise ratios. These results also indicate that the proposed method comparing to some conventional denoising methods such as wavelet, Butterworth, and empirical mode decomposition (EMD), leads to a higher quality denoising approach.

Dual phase dependent RLS filtering approach for baseline wander removal in ECG signal acquisition

Electrocardiogram (ECG) signals often get contaminated during physical activity, due to baseline disturbances. Removal of baseline interference also leads to variation in the timing of ECG waves. This paper proposes a dual phase dependent-recursive least square (DPD-RLS) adaptive filter that removes the baseline interference during ECG signal acquisition. This approach initially detects the upper envelope followed by the subtraction of the upper envelope with the original signal. The lower envelope is then detected from the resultant signal. Two phases are extracted from the upper envelope signal and the lower envelope signal that compensates for the phase change effects during the baseline wander removal process. A DPD-RLS adaptive filter structure is proposed that compensates for the phase changes near the P, Q, and T onset of the ECG signal. The evaluation was done using the clean ECG signals from the MIT-BIH arrhythmia dataset and baseline wander signals from MIT-BIH noise stress dataset signals using the metrics such as output signal to noise ratio (SNRo), correlation coefficient (γ), and percent root mean square difference (PRD) for the different input signal to noise ratios (SNRi). The proposed scheme provides an average output SNR, the correlation coefficient of 14.07dB and 0.980 respectively at SNRi=5dB when evaluated on different ECG records. Experimental results reveal that the proposed baseline wander removal scheme outperforms the traditional schemes.

PO-700-08 ELIMINATION OF HARMONIC RINGING IN INTRACARDIAC SIGNALS USING DYNAMIC ALGORITHMIC NOTCH FILTERING

The use of classic notch (CN) and bandpass (BP) filters on intracardiac electrogram (EGM) channels have long been considered the best options for correcting environmental noise in the EP lab. Due to the steep transitions and narrowness of the bandwidth, these filters can alter signal morphology, producing artificial harmonic ringing and additional deflections that may be misinterpreted as physiologic signal fractionation. CN and BP are often automatically preprogrammed on EGM channels without further consideration. While filtering to remove baseline wander is still required to properly interpret signals, an alternative solution that doesn’t generate harmonic ringing and artificial fractionation would improve the confidence and accuracy in identifying abnormal signals. To characterize the impact of CN and BP filters on EGM signals during ablation procedures and evaluate the Algorithmic Notch (AN) filter within the PURE EP™ system (PURE) (BioSig Technologies). Signals of clinical interest were assessed during 20 re-do AF ablation procedures that compared signals between PURE and the GE Cardiolab system. Using PURE the signals from the same intracardiac channel were run simultaneously using different filter settings. A control channel was run with only a high pass filter of 5Hz to remove some baseline wander. The same signals were subsequently compared to the control with three different filter settings: 1) a high pass filter of 10Hz + PURE AN 2) a high pass filter of 10Hz + CN, and 3) a BP 30-500Hz + CN. The AN on PURE was consistently the most successful in removing baseline artifact without significantly altering the morphology or signal characteristics. The appearance of artificial deflections resembling physiologic fractionation were seen in the latter two groups and signal amplitudes were attenuated by 30-40% when compared to the PURE AN group. The AN in PURE can eliminate environmental noise without harmonic ringing, morphological changes, or attenuation seen with CN and BP filtering.

Sparsity-based modified wavelet de-noising autoencoder for ECG signals

Electrocardiogram (ECG) is susceptible to different kinds of noises whose removal is necessary for accurate clinical diagnosis. This paper proposes a hybrid technique that integrates the concepts of sparsity, wavelet transform, and extreme learning machine into a single framework. Initially, the loss function of the sparsity-based method is designed with linear time-variant filtering parameters, and a compound penalty-based Huber function is used for the removal of low-frequency baseline wander. Sparse optimization is carried out by the majorization-minimization (MM) technique ensuring fast and guaranteed convergence irrespective of initialization. The next step involves wavelet-based de-noising with novel thresholding followed by extreme machine learning for remnant noise removal. The comparative analysis of the proposed method is done on the MIT-BIH Arrhythmia database for baseline wander (BW), additive white Gaussian noise (AWGN), muscle artifacts (MA), power-line interference (PLI), and composite noise (CN) both qualitatively and quantitatively. Qualitative analysis is also performed on MIT-BIH NSR and MIT-BIH NST. For AWGN, BW, MA, PLI, and CN, SNRimp is maximum at 27.8670 dB (record 119), 32.5962 dB (record 215), 25.7825 dB (record 119), 31.9277 dB (record 215), 25.5463 dB (record 105) at an SNRin of 10 dB respectively. Significant improvement in terms of SNRimp, RMSE, and PRD is obtained over the state-of-the-art ECG de-noising methods. Feature preservation in the de-noised ECG signal is also investigated with the help of fiducial morphological features.

Power line noise and baseline wander removal from ECG signals using empirical mode decomposition and lifting wavelet transform technique

Power line noise and baseline wander removal from ECG signals using empirical mode decomposition and lifting wavelet transform technique

A Comparison Study for Cardiac Arrhythmias Noise Cancellation Techniques

The electrocardiogram (ECG), which represents the electrical activity of the heart, is always chosen as the basic signal for diagnosing cardiac abnormalities and detecting the patient's states. Generally, the various filters are applied (preprocessing) to denoise the artifacts from ECG signal. This step makes the decision making and diagnosis simpler and faster. The most common noise sources are power line interference and baseline wandering. This article discusses different filtering techniques used to denoise the ECG signals. Electrocardiogram signals were downloaded from the MIT-BIH Arrhythmia Database. The MIT-BIH Database was the first set of standard test materials that is generally available to evaluate arrhythmia detection. To remove baseline wander, 4 filters were designed. The filters are a fourth-order high-pass Butterworth filter, a second-order finite impulse response Hamming window, an finite impulse response rectangular window with length L, and an finite impulse response Hanning window. To extract the power line interference (60 Hz), 4 filters were designed: a second-order digital bandpass Butterworth filter, a second-order finite impulse response filter, a Chebyshev filter, and an finite impulse response notch filter with Hanning window. Finally, 5 parameters were used to evaluate filter performance; the parameters are frequency response, phase response, average filter delay (group delay), phase delay, and signal-to-noise ratio. The results show that, for baseline wander filters, a fourth-order high-pass Butterworth filter removes more noise and has no time delay. Power line interference filters have the same signal-to-noise ratio approximately.

Design of uniform cosine modulated filter bank using [formula omitted] and its application in baseline wander removal from ECG signal

In this paper, a uniform cosine modulated filter bank (CMFB) is designed using the Incremental Ant Colony Optimization with Local Search (IACOR-LS) and utilized for the removal of Baseline Wander (BW) from the ECG signal. Here, for the CMFB design, a prototype filter is designed with window function having shape parameter ‘α’ and its cut-off frequency (ωc) is optimized to reduce the amplitude distortion and aliasing errors. It is observed that the proposed method performed better in terms of aliasing errors with higher values of ‘α’. In the continuation, a simple technique is proposed for BW removal from ECG in which the minima of the first sub-band signal of 8-channel uniform CMFB are used to estimate the BW. Four parameters, i.e., Mean Square Error (MSE), Correlation Coefficient (CC), L_Operator (LO) and Absolute Maximum Distance (AMD) are evaluated and compared with the existing methods. It is observed that the proposed technique for BW removal outperformed the existing methods.

Electrocardiogram signal filtering using circulant singular spectrum analysis and cascaded Savitzky-Golay filter

The electrocardiogram (ECG) is a tool that is used to examine the heart’s electrical activity. The processing of ECG signals is thus critical for recognising anomalies or the start of illnesses. Noises like powerline interference (PLI) and baseline wander (BW) tend to occur in ECG signals, if not eliminated properly they tend to degrade the signal quality considerably. Henceforth, removal of PLI and BW is a crucial task in the analysis of biomedical signals, especially ECG signals. In this paper, we used a newly developed method called circulant singular spectrum analysis (CiSSA) for the removal of BW and four stage cascaded Savitzky-Golay (SG) filter for the elimination of PLI from the ECG signals by retaining the morphological properties of ECG signal. The efficacy of the method proposed is endorsed by using MIT-BIH Arrhythmia database. The simulation results noticeably exhibit that the method performs better when compared with that of other existing state-of-art techniques using performance metrics like output signal-to-noise ratio (SNRout), root mean square error (RMSE), percent root mean square difference (PRD), and correlation coefficient (CC) at various levels of input signal-to-noise ratio power (SNRin). We have achieved an average SNRout of 24.48 and CC of 0.99. Hence, the intended approach can be used in preprocessing stage of ECG signal analysis.

Behavioral control and changes in brain activity of honeybee during flapping.

IntroductionInsect cyborg is a kind of novel robot based on insect–machine interface and principles of neurobiology. The key idea is to stimulate live insects by specific stimuli; thus, the flight trajectory of insects could be controlled as anticipated. However, the neuroregulatory mechanism of insect flight has not been elucidated completely at present.MethodsTo explore the neuro‐mechanism of insect flight behaviors, a series of electrical stimulation was applied on the optic lobes of semi‐constrained honeybees. Times of flight initiation, flapping frequency, and duration were recorded by a high‐speed camera. In addition, flapping and steering initiation experiments of the cyborg honeybee were verified. Moreover, series of local field potential signals of optic lobes during flapping were collected, pre‐processed to remove baseline wander and DC components, then analyzed by power spectrum estimation.ResultsA quantitative optimization method and optimal stimulation parameters of flight initiation were presented. Stimulation results showed that the flapping duration differed greatly while the flapping frequency varied with little difference among different individuals. Moreover, there was always a fluctuation peak around 20–30 Hz in power spectral density (PSD) curves during flapping, distinguishing from calm state, which indicated some brain activity changes during flapping.ConclusionsOur study presented a range of relatively optimal electrical parameters to initiate honeybee flight behavior. Meanwhile, the regularity of flapping duration and flapping frequency under electrical stimulations with different parameters were given. The feasibility of controlling a honeybee's flight behavior by brain electrical stimulation was verified through the flapping and steering initiation experiment of honeybees under semi‐constrained state. PSD fluctuations reflected changes in brain activity during flapping and that those fluctuation characteristics at the specific frequency band could be sensitive determinants to distinguish whether the honeybee was flying or not, which benefits our understanding of honeybee's flapping behavior and furthers the study of honeybee cyborgs.

Quantitatively Recognizing Stimuli Intensity of Primary Taste Based on Surface Electromyography.

A novel approach to quantitatively recognize the intensity of primary taste stimuli was explored based on surface electromyography (sEMG). We captured sEMG samples under stimuli of primary taste with different intensities and quantitatively recognized preprocessed samples with Support Vector Machine (SVM). The feasibility of quantitatively recognizing the intensity of Sour, Bitter, and Salty was verified. The sEMG signals were acquired under the stimuli of citric acid (aq), sucrose (aq), magnesium chloride (aq), sodium chloride (aq), and sodium glutamate (aq) with different concentrations, for five types of primary tastes: Sour, Sweet, Bitter, Salty, and Umami, whose order was fixed in this article. The acquired signals were processed with a method called Quadratic Variation Reduction to remove baseline wandering, and an adaptive notch to remove power frequency interference. After extracting 330 features for each sample, an SVM regressor with five-fold cross-validation was performed and the model reached R2 scores of 0.7277, 0.1963, 0.7450, 0.7642, and 0.5055 for five types of primary tastes, respectively, which manifested the feasibilities of the quantitative recognitions of Sour, Bitter, and Salty. To explore the facial responses to taste stimuli, we summarized and compared the muscle activities under stimuli of different taste types and taste intensities. To further simplify the model, we explored the impact of feature dimensionalities and optimized the feature combination for each taste in a channel-wise manner, and the feature dimensionality was reduced from 330 to 210, 120, 210, 260, 170 for five types of primary tastes, respectively. Lastly, we analyzed the model performance on multiple subjects and the relation between the model’s performance and the number of experiment subjects. This study can provide references for further research and applications on taste stimuli recognition with sEMG.

Comparison of ECG Baseline Wander Removal Techniques and Improvement Based on Moving Average of Wavelet Approximation Coefficients

The baseline wander is among the artifacts that corrupt the ECG signal. This noise can affect some signal features, in particular the ST segment, which is an important marker for the diagnosis of ischemia. This paper presents a study on the effectiveness of several methods and techniques for suppressing the baseline wonder (BW) from the ECG signals. As a result, a new technique called moving average of wavelet approximation coefficients (DWT-MAV) is proposed. The techniques concerned are the moving average, the approximation of the baseline by polynomial fitting, the Savitzky-Golay filtering, and the discrete wavelet transform (DWT). The comparison of this techniques is performed using the main criteria for assessing the BW denoising quality criteria such mean square error (MSE), percent root mean square difference (PRD) and correlation coefficient (COR). In this paper, three other criteria of comparison are proposed namely the number of samples of the ECG signal, the baseline frequency variation and the time processing. Two of these new indices are related to possible real time ECG denoising. To improve the quality of BW suppression including the new indices, a new method is proposed. This technique is a combination of the DWT and the moving average methods. This new technique performs the best compromise in terms of MSE, PRD, coefficient correlation and the time processing. The simulations were performed on ECG recording from MIT-BIH database with synthetic and real baselines.