Deep neural networks have recently been recognized as one of the powerful learning techniques in computer vision and medical image analysis. Trained deep neural networks need to be generalizable to new data that are not seen before. In practice, there is often insufficient training data available, which can be solved via data augmentation. Nevertheless, there is a lack of augmentation methods to generate data on graphs or surfaces, even though graph convolutional neural network (graph-CNN) has been widely used in deep learning. This study proposed two unbiased augmentation methods, Laplace-Beltrami eigenfunction Data Augmentation (LB-eigDA) and Chebyshev polynomial Data Augmentation (C-pDA), to generate new data on surfaces, whose mean was the same as that of observed data. LB-eigDA augmented data via the resampling of the LB coefficients. In parallel with LB-eigDA, we introduced a fast augmentation approach, C-pDA, that employed a polynomial approximation of LB spectral filters on surfaces. We designed LB spectral bandpass filters by Chebyshev polynomial approximation and resampled signals filtered via these filters in order to generate new data on surfaces. We first validated LB-eigDA and C-pDA via simulated data and demonstrated their use for improving classification accuracy. We then employed brain images of the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and extracted cortical thickness that was represented on the cortical surface to illustrate the use of the two augmentation methods. We demonstrated that augmented cortical thickness had a similar pattern to observed data. We also showed that C-pDA was faster than LB-eigDA and can improve the AD classification accuracy of graph-CNN.Despite its success in understanding brain rhythms, the neural mass model, as a low-dimensional mean-field network model, is phenomenological in nature, so that it cannot replicate some of rich repertoire of responses seen in real neuronal tissues. Here, using a colored-synapse population density method, we derived a novel neural mass model, termed density-based neural mass model (dNMM), as the mean-field description of network dynamics of adaptive exponential integrate-and-fire (aEIF) neurons, in which two critical neuronal features, i.e., voltage-dependent conductance-based synaptic interactions and adaptation of firing rate responses, were included. Our results showed that the dNMM was capable of correctly estimating firing rate responses of a neuronal population of aEIF neurons receiving stationary or time-varying excitatory and inhibitory inputs. Finally, it was also able to quantitatively describe the effect of spike-frequency adaptation in the generation of asynchronous irregular activity of excitatory-inhibitory cortical networks. We conclude that in terms of its biological reality and calculation efficiency, the dNMM is a suitable candidate to build significantly large-scale network models involving multiple brain areas, where the neuronal population is the smallest dynamic unit.In this paper, we propose a visual embedding approach to improve embedding aware speech enhancement (EASE) by synchronizing visual lip frames at the phone and place of articulation levels. We first extract visual embedding from lip frames using a pre-trained phone or articulation place recognizer for visual-only EASE (VEASE). Next, we extract audio-visual embedding from noisy speech and lip frames in an information intersection manner, utilizing a complementarity of audio and visual features for multi-modal EASE (MEASE). Experiments on the TCD-TIMIT corpus corrupted by simulated additive noises show that our proposed subword based VEASE approach is more effective than conventional embedding at the word level. Moreover, visual embedding at the articulation place level, leveraging upon a high correlation between place of articulation and lip shapes, demonstrates an even better performance than that at the phone level. Finally the experiments establish that the proposed MEASE framework, incorporating both audio and visual embeddings, yields significantly better speech quality and intelligibility than those obtained with the best visual-only and audio-only EASE systems.Burkholderia gladioli is a Gram-negative bacterium that biosynthesizes a cocktail of potent antimicrobial compounds, including the antifungal phenolic glycoside sinapigladioside. VH298 nmr Herein, we report the total synthesis of the proposed structures of gladiosides I and II, two structurally related phenolic glycosides previously isolated from B. gladioli OR1 cultures. Importantly, the physical and analytical data of the synthetic compounds were in significant discrepancies with the natural products suggesting a misassignment of the originally proposed structures. Furthermore, we have uncovered an acid-catalyzed fragmentation mechanism converting the α,β-unsaturated methyl carbamate-containing gladioside II into the aldehyde-containing gladioside I. Our results lay the foundation for the expeditious synthesis of derivatives of these Burkholderia-derived phenolic glycosides, which would enable to decipher their biological roles and potential pharmacological properties.The chief ROS formed by mitochondria are superoxide (O2·-) and hydrogen peroxide (H2O2). Superoxide is converted rapidly to H2O2 and therefore the latter is the chief ROS emitted by mitochondria into the cell. Once considered an unavoidable by-product of aerobic respiration, H2O2 is now regarded as a central mitokine used in mitochondrial redox signaling. However, it has been postulated that O2·- can also serve as a signal in mammalian cells. Progress in understanding the role of mitochondrial H2O2 in signaling is due to significant advances in the development of methods and technologies for its detection. Unfortunately, the development of techniques to selectively measure basal O2·- changes has been met with more significant hurdles due to its short half-life and the lack of specific probes. The development of sensitive techniques for the selective and real time measure of O2·- and H2O2 has come on two fronts development of genetically encoded fluorescent proteins and small molecule reporters. In 2015, I published a detailed comprehensive review on the state of knowledge for mitochondrial ROS production and how it is controlled, which included an in-depth discussion of the up-to-date methods utilized for the detection of both superoxide (O2·-) and H2O2.