Machine to brain: facial expression recognition using brain machine generative adversarial networks.

The human brain can effectively perform Facial Expression Recognition (FER) with a few samples by utilizing its cognitive ability. However, unlike the human brain, even the well-trained deep neural network is data-dependent and lacks cognitive ability. To tackle this challenge, this paper proposes a novel framework, Brain Machine Generative Adversarial Networks (BM-GAN), which utilizes the concept of brain's cognitive ability to guide a Convolutional Neural Network to generate LIKE-electroencephalograph (EEG) features. More specifically, we firstly obtain EEG signals triggered from facial emotion images, then we adopt BM-GAN to carry out the mutual generation of image visual features and EEG cognitive features. BM-GAN intends to use the cognitive knowledge learnt from EEG signals to instruct the model to perceive LIKE-EEG features. Thereby, BM-GAN has a superior performance for FER like the human brain. The proposed model consists of VisualNet, EEGNet, and BM-GAN. More specifically, VisualNet can obtain image visual features from facial emotion images and EEGNet can obtain EEG cognitive features from EEG signals. Subsequently, the BM-GAN completes the mutual generation of image visual features and EEG cognitive features. Finally, the predicted LIKE-EEG features of test images are used for FER. After learning, without the participation of the EEG signals, an average classification accuracy of 96.6 % is obtained on Chinese Facial Affective Picture System dataset using LIKE-EEG features for FER. Experiments demonstrate that the proposed method can produce an excellent performance for FER.

Fabrication of Eco-Friendly Hydrolyzed Ethylene-Maleic Anhydride Copolymer-Avermectin Nanoemulsion with High Stability, Adhesion Property, pH, and Temperature-Responsive Releasing Behaviors.

In this study, novel amphiphilic polymer emulsifiers for avermectin (Avm) were synthesized facilely via the hydrolysis of ethylene-maleic anhydride copolymer (EMA) with different agents, and their structures were confirmed by various techniques. Then, water-based Avm-nanoemulsions were fabricated with the emulsifiers via phase inversion emulsification process, and superior emulsifier was selected via the emulsification effects. Using the superior emulsifier, an optimal Avm-nanoemulsion (defined as Avm@HEMA) with satisfying particle size of 156.8 ± 4.9 nm, encapsulation efficiency (EE) of 69.72 ± 4.01% and drug loading capacity (DLC) of 54.93 ± 1.12% was constructed based on response surface methodology (RSM). Owing to the emulsifier, the Avm@HEMA showed a series of advantages, including high stability, ultraviolet resistance, low surface tension, good spreading and high affinity to different leaves. Additionally, compared to pure Avm and Avm-emulsifiable concentrate (Avm-EC), Avm@HEMA displayed a controlled releasing feature. The encapsulated Avm was released quite slowly at normal conditions (pH 7.0, 25 °C or 15 °C) but could be released at an accelerated rate in weak acid (pH 5.5) or weak alkali (pH 8.5) media or at high temperature (40 °C). The drug releasing profiles of Avm@HEMA fit the Korsmeyer-Peppas model quite well at pH 7.0 and 25 °C (controlled by Fickian diffusion) and at pH 7.0 and 10 °C (controlled by non-Fickian diffusion), while it fits the logistic model under other conditions (pH 5.5 and 25 °C, pH 8.5 and 25 °C, pH 7.0 and 40 °C).

Repurposing of Antidiarrheal Loperamide for Treating Melanoma by Inducing Cell Apoptosis and Cell Metastasis Suppression In vitro and In vivo.

BACKGROUND: Melanoma is the most common skin tumor worldwide and still lacks effective therapeutic agents in clinical practice. Repurposing of existing drugs for clinical tumor treatment is an attractive and effective strategy. Loperamide is a commonly used anti-diarrheal drug with excellent safety profiles. However, the affection and mechanism of loperamide in melanoma remain unknown. Herein, the potential anti-melanoma effects and mechanism of loperamide were investigated in vitro and in vivo. METHODS: In the present study, we demonstrated that loperamide possessed a strong inhibition in cell viability and proliferation in melanoma using MTT, colony formation and EUD incorporation assays. Meanwhile, xenograft tumor models were established to investigate the anti-melanoma activity of loperamide in vivo. Moreover, the effects of loperamide on apoptosis in melanoma cells and potential mechanisms were explored by Annexin V-FITC apoptosis detection, cell cycle, mitochondrial membrane potential assay, reactive oxygen species level detection, and apoptosis-correlation proteins analysis. Furthermore, loperamide-suppressed melanoma metastasis was studied by migration and invasion assays. What's more, immunohistochemical and immunofluorescence staining assays were applied to demonstrate the mechanism of loperamide against melanoma in vivo. Finally, we performed the analysis of routine blood and blood biochemical, as well as hematoxylin- eosin (H&E) staining, in order to investigate the safety properties of loperamide. RESULTS: Loperamide could observably inhibit melanoma cell proliferation in vitro and in vivo. Meanwhile, loperamide induced melanoma cell apoptosis by accumulation of the sub-G1 cells population, enhancement of reactive oxygen species level, depletion of mitochondrial membrane potential, and apoptosis-related protein activation in vitro. Of note, apoptosis-inducing effects were also observed in vivo. Subsequently, loperamide markedly restrained melanoma cell migration and invasion in vitro and in vivo. Ultimately, loperamide was witnessed to have an amicable safety profile. CONCLUSION: These findings suggested that repurposing of loperamide might have great potential as a novel and safe alternative strategy to cure melanoma via inhibiting proliferation, inducing apoptosis and cell cycle arrest, and suppressing migration and invasion.

Evaluation of Satisfaction with the Built Environment of University Buildings under the Epidemic and Its Impact on Student Anxiety.

Anxiety on college campuses has increased due to the COVID-19 epidemic's profound effects on society. Much research has been conducted on how the built environment influences mental health; however, little has been undertaken on how it affects student mental health in the context of the epidemic from the architectural scale perspective of academic buildings. Based on online survey data, this study develops multiple linear regression and binary logistic regression models to investigate students' satisfaction ratings of the academic buildings' physical environments during the epidemic and how these satisfaction ratings affect students' anxiety tendencies. According to the study's findings regarding the natural exposure perspective, students who perceived the academic building's poor semi-open space view (p = 0.004, OR = 3.22) as unsatisfactory factors were more likely to show anxiety tendencies. In terms of the physical conditions, students who were dissatisfied with the noise level in the classroom (p = 0.038, OR = 0.616) and the summer heat in semi-open spaces (p = 0.031, OR = 2.38) were more likely to exhibit anxiety tendencies. Additionally, even after controlling for confusing distractions, the general satisfaction rating of the academic building's physical environments (p = 0.047, OR = 0.572) was still able to significantly and negatively affect students' anxiety tendencies. The study's findings can be used in the architectural and environmental planning of academic buildings focusing on mental health.

Marginal Effects and Spatial Variations of the Impact of the Built Environment on Taxis' Pollutant Emissions in Chengdu, China.

Understanding the impact of the urban built environment on taxis' emissions is crucial for sustainable transportation. However, the marginal effects and spatial heterogeneity of this impact is worth noting. To this end, we calculated the taxis' emissions on weekdays and weekends in Chengdu, China, and investigated the impact of the built environment on taxis' emissions by applying multi-source data and several spatial regression models. The results showed that the taxis' daily emissions on weekdays were higher than the emissions on weekends. The time heterogeneity of hourly taxis' emissions was not significant, while the spatial heterogeneity of taxis' emissions was significant. Except the HHI, the selected built environment variables both had a significant positive effect on taxis' emissions on the global scale. There was a marginal effect of some built environment variables on taxis' emissions, such as the density of bus stops and population density. The former exhibited an inhibitory effect on taxis' emissions only when it was greater than 9 stops/km2, while the latter showed an inhibitory effect only in the range between 16,000 people/km2 and 22,000 people/km2. There were some spatial variations in the effects of built environment factors on taxis' emissions, with HHI, road density, and accommodation service facilities density showing the most significant variation. The marginal effect and spatial variation of the impact needs to be considered when developing strategies to reduce taxis' pollutant emissions.

Highly sensitive gas sensor based on a parity-time-symmetric system.

Achieving extremely high sensitivity is an important indicator in the development of novel and stable gas concentration sensors. In this paper, we present a gas concentration sensor with parity-time symmetry for high sensitivity at low concentrations. The proposed sensor can detect toxic gases, such as benzene, bromine, and acetone, by probing the faint changing of the permittivity. Furthermore, the level of the sensitivity can be adjusted by the resistance segment, which is realized by various metallic formations. Our proposed structure provides a novel idea for the development of future gas concentration sensors, showing an exciting prospect for gas sensing technologies.

One-dimensional photonic crystal slot waveguide for silicon-organic hybrid electro-optic modulators.

In an on-chip silicon-organic hybrid electro-optic (EO) modulator, the mode overlap with EO materials, in-device effective r33, and propagation loss are among the most critical factors that determine the performance of the modulator. Various waveguide structures have been proposed to optimize these factors, yet there is a lack of comprehensive consideration on all of them. In this Letter, a one-dimensional (1D) photonic crystal (PC) slot waveguide structure is proposed that takes all these factors into consideration. The proposed structure takes advantage of the strong mode confinement within a low-index region in a conventional slot waveguide and the slow-light enhancement from the 1D PC structure. Its simple geometry makes it robust to resist fabrication imperfections and helps reduce the propagation loss. Using it as a phase shifter in a Mach-Zehnder interferometer structure, an integrated silicon-organic hybrid EO modulator was experimentally demonstrated. The observed effective EO coefficient is as high as 490 pm/V. The measured half-wave voltage and length product is less than 1 V·cm and can be further improved. A potential bandwidth of 61 GHz can be achieved and further improved by tailoring the doping profile. The proposed structure offers a competitive novel phase-shifter design, which is simple, highly efficient, and with low optical loss, for on-chip silicon-organic hybrid EO modulators.

Enhanced optical absorption and electric field resonance in diabolo metal bar optical antennas.

Resonance behaviors of the fundamental resonance mode of diabolo metal bar optical antennas are investigated by using finite-difference time-domain (FDTD) numerical simulations and a dipole oscillator model. It is found that as the waist of the diabolo metal bar optical antenna is reduced, optical energy absorption cross section and near field enhancement at resonance increase significantly. Also reduction of the diabolo waist width causes red-shift of the resonant wavelengths in the spectra of absorption cross-section, scattering cross-section, and the near electric field. A dipole oscillator model including the self-inductance force is used to fit the FDTD numerical simulation results. The dipole oscillator model characterizes well the resonance behaviors of narrow waist diabolo metal bar optical antennas.