The common spatial pattern (CSP) is a very effective feature extraction method in motor imagery based brain computer interface (BCI), but its performance depends on the selection of the optimal frequency band. Although a lot of research works have been proposed to improve CSP, most of these works have the problems of large computation costs and long feature extraction time. To this end, three new feature extraction methods based on CSP and a new feature selection method based on non-convex log regularization are proposed in this paper. Firstly, EEG signals are spatially filtered by CSP, and then three new feature extraction methods are proposed. We called them CSP-wavelet, CSP-WPD and CSP-FB, respectively. For CSP-Wavelet and CSP-WPD, the discrete wavelet transform (DWT) or wavelet packet decomposition (WPD) is used to decompose the spatially filtered signals, and then the energy and standard deviation of the wavelet coefficients are extracted as features. For CSP-FB, the spatially filtered signals are filtered into multiple bands by a filter bank (FB), and then the logarithm of variances of each band are extracted as features. Secondly, a sparse optimization method regularized with a non-convex log function is proposed for the feature selection, which we called LOG, and an optimization algorithm for LOG is given. Finally, ensemble learning is used for secondary feature selection and classification model construction. Combing feature extraction and feature selection methods, a total of three new EEG decoding methods are obtained, namely CSP-Wavelet+LOG, CSP-WPD+LOG, and CSP-FB+LOG. Four public motor imagery datasets are used to verify the performance of the proposed methods. Compared to existing methods, the proposed methods achieved the highest average classification accuracy of 88.86, 83.40, 81.53, and 80.83 in datasets 1–4, respectively. The feature extraction time of CSP-FB is the shortest. The experimental results show that the proposed methods can effectively improve the classification accuracy and reduce the feature extraction time. With comprehensive consideration of classification accuracy and feature extraction time, CSP-FB+LOG has the best performance and can be used for the real-time BCI system.

In this study, the characteristic absorption spectra of Vanillin and ortho-Vanillin were investigated by terahertz time-domain spectroscopy (THz-TDS) in the range of 0.42.0 THz. Using quantum chemistry tools, density functional theory (DFT) calculations were performed to obtain the theoretical THz spectra. To better understand the THz absorption spectra on a molecular level and to systematically and comprehensively interpret the origin of the characteristic absorption peaks, the vibrational characteristics and the intermolecular weak interactions of these two compounds were analyzed using a cluster model. We used the vibrational mode automatic relevance determination (VMARD) method to assign the vibrational modes of all absorption peaks and provide the percentage contribution of these modes. The vibrational modes of Vanillin and ortho-Vanillin were mainly dihedral angle torsion. The intermolecular weak interactions were analyzed using the independent gradient model based on Hirshfeld partition of molecular density (IGMH) method. The results indicated that the weak interaction types of Vanillin and ortho-Vanillin were dominated by van der Waals interaction. This work has demonstrated that using THz-TDS combined with quantum chemical calculations is an effective way to identify and investigate isomers. This method has significant potential appli-cations for exploring the relationship between the biomolecular structure and the material performance.

We designed a bilayer-double-H-metamaterials (BDHM) composed of two layers of metal and two layers of dielectric to analog a spectral response of electromagnetically induced transparency (EIT) at terahertz frequency. By changing the incident angle, the BDHM exhibits an EIT-like spectral response. The tunable spectral performances and modulation mechanism of the transparent peak are theoretically investigated using full-wave electromagnetic simulation software. The physical mechanism of the EIT-like effect is based on the constructive and destructive interference between the induced electrical dipoles. Our work provides a new way to realize the EIT-like effect only by changing the incident angles of the metamaterials. The potential applications include tunable filters, sensors, attenuators, switches, and so on.

We experimentally present a mechanically tunable multiband terahertz (THz) bandstop filter based on near field coupling of metamaterials. The unit cell of the device consists of an outside aluminum hexagon ring and an inner aluminum circle on high resistance silicon substrate. Two rings are asymmetrically split at the same position deviating from the center of the unit cell. Initially, the device has four resonance absorption peaks at 0.46 THz, 0.57 THz, 0.66 THz and 0.90 THz, respectively. When the azimuth increases from 0° to 90°, the transmission amplitude at these four absorption peaks can be simultaneously modulated. While, the absorption peak at 0.90 THz shows evident blue shift. The effects of geometric parameters on the tunability of bandstop filter are investigated using the finite-difference time-domain (FDTD) method. The sample is fabricated using a surface micromachining process and characterized using a THz time domain spectroscopy (TDS) system. The experiment is in good agreement with the simulation, and the measured amplitude modulation depth exceeds 12 dB at every resonance frequency. The proposed tunable THz filter can be applied in many fields, such as THz communications, sensing and multi-frequency switching.

A four-peak terahertz metamaterial sensor was used to detect the reaction between different concentrations of vitamin B 6 ( V B 6 ) and bovine serum albumin (BSA), which achieves a concentration range (0.015–0.125 mg/µl) of V B 6 and a maximum binding concentration (0.05 mg/µl) of V B 6 and 0.0875 mg/µl BSA. To understand the combination of V B 6 and BSA, the reactants between VB ( V B 1 , V B 3 , and V B 5 ) with the same concentration (0.05 mg/µl) and a BSA solution with a concentration of 0.0875 mg/µl were carried on the surface of the sensor. Experimental results show that the reactants cause the four resonance peaks of the sensor to produce the coincident redshift, which is the same as the order of their binding coefficients determined by the fluorescence method. The experimental process indicates that it is feasible to use terahertz metamaterials to detect the reaction process of organic matter.