Wyniki wyszukiwania - BazTech

1

A combination of statistical parameters for epileptic seizure detection and classification using VMD and NLTWSVM

Zhang Shang, Liu Guangda, Xiao Ruolan, Cui Wenjie, Cai Jing, Hu Xinlei, Sun Yubing, Qiu Jiqing, Qi Yuan

Biocybernetics and Biomedical Engineering

|

2022

|

Vol. 42, no. 1

258--272

EN

The epileptic seizure detection and classification is of great significance for clinical diagnosis and treatment. To realize the detection and classification of epileptic seizure, this paper proposes a method based on the combination of signal decomposition and statistical methods. First, the algorithm of variational mode decomposition (VMD) is applied to extract the components of intrinsic mode functions (IMFs) by decomposing the EEG signals. Then the statistical method is utilized to calculate the eight features of maximum, minimum, average, variance, skewness, kurtosis, coefficient of variation and volatility index for each extracted IMF component. Finally, the best combinations of extracted features are fed into the non-linear twin support vector machine (NLTWSVM) to classify the epileptic signals. The EEG database from University of Bonn is used to confirm the effectiveness of the proposed method for epileptic seizure detection. The final experimental results demonstrate that the classification accuracy can reach 98.86%, 98.37%, 99.02%, 99.41% and 99.57% for the database of C-E, D-E, CD-E, ABCD-E and AB-CD-E, respectively. The TUSZ corpus in the TUH EEG corpus is also used to classify epileptic seizure types using the method in this article. The result is expressed by the confusion matrix and the weighted F 1 score is 0.923, which shows this method has potential to help experienced neurophysiologists classify epileptic seizure types in the clinic.

2

Grey wolf optimization-based variational mode decomposition for magnetotelluric data combined with detrended fluctuation analysis

Zhang Xian, Li Diquan, Li Jin, Li Yong

Acta Geophysica

|

2022

|

Vol. 70, no 1

111--120

EN

We present a magnetotelluric data denoising method that uses grey wolf optimization to optimize variational mode decomposition and combines it with detrended fluctuation analysis. First, envelope entropy is selected as the fitness function for grey wolf optimization and is used to determine the number of modes K and the penalty factor, which are the key parameters of the variational mode decomposition method. Then, the optimized variational mode decomposition method is used to decompose magnetotelluric data. Finally, the scaling exponent in detrended fluctuation analysis is used to determine the corresponding intrinsic mode function components to superimpose and reconstruct the useful magnetotelluric data. Extensive experiments and thorough analyses are performed on the synthetic data and field data. The results of the proposed method are compared with the results of the remote reference, variational mode decomposition, variational mode decomposition and matching pursuit, variational mode decomposition and detrended fluctuation analysis methods; the proposed method can improve the denoising performance and reliability of low-frequency magnetotelluric data. The reconstructed data are closer to the natural magnetotelluric data. The satisfactory performance in the results verifies the effectiveness of the design and optimization method.

3

A combined iCEEMDAN and VMD method for mitigating the impact of ionospheric scintillation on GNSS signals

Dey Abhijit, Chhibba Rohan, Ratnam Devananboyina Venkata, Sharma Nitin

Acta Geophysica

|

2021

|

Vol. 69, no. 5

1933--1948

EN

Severe amplitude and phase scintillation induced by the ionospheric plasma density irregularities degrades the performance of global navigation satellite system (GNSS) receivers. Scintillation typically has adverse effects at the tracking process and thus adversely affects the raw GNSS measurements used in a number of applications. Hence, it is important to develop robust methodologies for detecting and mitigating ionospheric effects on the GNSS signals. In this paper, we propose a novel method based on the combination of improved complete ensemble empirical mode decomposition with adaptive noise (iCEEMDAN) and variational mode decomposition (VMD) methods. The proposed method employs a detrended fuctuation analysis (DFA)-based metric for robust thresholding between the scintillation-free and amplitude scintillated GNSS signals. The major contribution of the proposed method is development of novel approaches for selection of intrinsic mode functions (IMFs) based on DFA and optimised selection of [K, 훼] parameters of the VMD. The performance of the proposed method was evaluated and was observed that it is better than existing ionospheric scintillation effects mitigation algorithms for both simulated and real-time GPS scintillation datasets. The proposed method can denoise approximately 9.23–15.30 dB scintillation noise from the synthetic and 0.2–0.48 from the real scintillation index (S4) values. Therefore, the proposed (iCEEMDAN-VMD) method is appropriate for mitigating the ionospheric scintillation effects on the GNSS signals.

4

Desert seismic noise suppression based on multimodal residual convolutional neural network

Wang Shengnan, Li Yue, Zhao Yuxing

Acta Geophysica

|

2020

|

Vol. 68, no. 2

389--401

EN

Seismic exploration is an important means of oil and gas detection, but afected by complex surface and near-surface conditions, and the seismic records are polluted by noise seriously. Particularly in the desert areas, due to the infuence of wind and human activities, the complex desert noise with low-frequency, nonstationary and non-Gaussian characteristics is produced. It is difcult to extract efective signals from strong noise using existing denoising methods. To address this issue, the paper proposes a new denoising method, called multimodal residual convolutional neural network (MRCNN). MRCNN combines convolutional neural network (CNN) with variational modal decomposition (VMD) and adopts residual learning method to suppress desert noise. Since CNN-based denoisers can extract data features based on massive training set, the impact of noise types and intensity on the denoised results can be ignored. In addition, VMD algorithm can sparsely decompose signal, which will facilitate the feature extraction of CNN. Therefore, using VMD algorithm to optimize the input data will conducive to the performance of the network denoising. Moreover, MRCNN adopts reversible downsampling operator to improve running speed, achieving a good trade-of between denoising results and efciency. Extensive experiments on synthetic and real noisy records are conducted to evaluate MRCNN in comparison with existing denoisers. The extensive experiments demonstrate that the MRCNN can exhibit good efectiveness in seismic denoising tasks.