A Student Facial Expression Recognition Model Based on Multi-Scale and Deep Fine-Grained Feature Attention Enhancement.

In smart classroom environments, accurately recognizing students' facial expressions is crucial for teachers to efficiently assess students' learning states, timely adjust teaching strategies, and enhance teaching quality and effectiveness. In this paper, we propose a student facial expression recognition model based on multi-scale and deep fine-grained feature attention enhancement (SFER-MDFAE) to address the issues of inaccurate facial feature extraction and poor robustness of facial expression recognition in smart classroom scenarios. Firstly, we construct a novel multi-scale dual-pooling feature aggregation module to capture and fuse facial information at different scales, thereby obtaining a comprehensive representation of key facial features; secondly, we design a key region-oriented attention mechanism to focus more on the nuances of facial expressions, further enhancing the representation of multi-scale deep fine-grained feature; finally, the fusion of multi-scale and deep fine-grained attention-enhanced features is used to obtain richer and more accurate facial key information and realize accurate facial expression recognition. The experimental results demonstrate that the proposed SFER-MDFAE outperforms the existing state-of-the-art methods, achieving an accuracy of 76.18% on FER2013, 92.75% on FERPlus, 92.93% on RAF-DB, 67.86% on AffectNet, and 93.74% on the real smart classroom facial expression dataset (SCFED). These results validate the effectiveness of the proposed method.

Elevated levels of hydroxylated phosphocholine lipids in the blood serum of breast cancer patients.

The difference in serum phospholipid content between stage-IV breast cancer patients and disease-free individuals was studied by employing a combination of chemometric statistical analysis tools and mass spectrometry. Chloroform-extracted serum samples were profiled for their lipid class composition and structure using precursor ion, neutral loss, and product ion tandem mass spectrometric (MS/MS) scanning experiments. Changes in the relative abundance of phospholipids in serum as a consequence of cancer progression, measured through electrospray ionization (ESI) mass spectrometry of flow-injected serum samples collected from 25 disease-free individuals and 50 patients diagnosed with stage-IV breast cancer, were statistically evaluated using principal component analysis (PCA), analysis of variance (ANOVA) and receiver operating characteristic (ROC) analysis. Lipids whose abundance changed significantly as a consequence of cancer progression were structurally characterized using product ion spectra, and independently quantified using precursor ion scan experiments against an internal standard of known concentration. Phosphocholine lipids that displayed a statistically significant change as a consequence of cancer progression were found to contain an oxidized fatty acid moiety as determined by MS3 experiments.

Quantitative chiral analysis of salsolinol in different brain regions of rats genetically predisposed to alcoholism.

A method to determine the catecholamine content in putamen (CPU) and midbrain (MB) regions of the brain of alcohol-preferring rats (P) is presented with a focus on the low-level detection of S,R-salsolinol, a metabolite of dopamine and a putative alcoholism marker. The developed strategy allows both quantitative profiling of related catecholamines and the enantiomeric separation and quantification of the S- and R-salsolinol isomers and their ratios. The described LC/MS strategy simplifies the current methodology that typically employs GC-MS by eliminating the need for derivatization. The data also suggest an increase in the non-enzymatic formation of salsolinol as a consequence of ethanol exposure.

Synthesis and characterization of covalent mimics of phosphatidylinositol-4,5-bisphosphate micelles.

There is increasing evidence that multivalency plays an important role in protein-lipid recognition and membrane targeting in biological systems. We describe here the preparation and characterization of multivalent analogues of the signaling lipid phosphatidylinositol-4,5-bisphosphate (PIP2). Tetherable analogues of the PIP2 headgroup were appended to polyamidoamine dendrimers via a squarate linker to afford polymers displaying four or eight headgroup moieties. This class of molecules should provide a powerful tool for the study of protein-lipid interactions.

Ion trajectory simulation for electrode configurations with arbitrary geometries.

A multi-particle ion trajectory simulation program ITSIM 6.0 is described, which is capable of ion trajectory simulations for electrode configurations with arbitrary geometries. The electrode structures are input from a 3D drawing program AutoCAD and the electric field is calculated using a 3D field solver COMSOL. The program CreatePot acts as interface between the field solver and ITSIM 6.0. It converts the calculated electric field into a field array file readable by ITSIM 6.0 and ion trajectories are calculated by solving Newton's equation using Runge-Kutta integration methods. The accuracy of the field calculation is discussed for the ideal quadrupole ion trap in terms of applied mesh density. Electric fields of several different types of devices with 3D geometry are simulated, including ion transport through an ion optical system as a function of pressure. Ion spatial distributions, including the storage of positively charged ions only and simultaneous storage of positively/negatively charged ions in commercial linear ion traps with various geometries, are investigated using different trapping modes. Inelastic collisions and collision induced dissociation modeled using RRKM theory are studied, with emphasis on the fragmentation of n-butylbenzene inside an ideal quadrupole ion trap. The mass spectrum of 1,3-dichlorobenzene is simulated for the rectilinear ion trap device and good agreement is observed between the simulated and the experimental mass spectra. Collisional cooling using helium at different pressures is found to affect mass resolution in the rectilinear ion trap.

Novel linear ion trap mass analyzer composed of four planar electrodes.

A novel linear ion trap mass analyzer was developed using just four elongated planar electrodes, mounted in parallel, and employing an RF potential for ion trapping in the radial and axial directions. Mass analysis was achieved using the mass-selective instability scan with ion ejection in the radial direction. The performance of this new device was characterized in comparison with the 6-electrode rectilinear ion trap (RIT) from which it is derived. The 4-electrode trap gives optimum performance in an asymmetric geometry, just like the original optimized 6-electrode RIT. The strong RF fringing fields at the ends of the RF rods account for axial ion trapping without use of extra electrodes or an axial DC voltage. Field calculations and simulations have been carried out to study the trapping potential inside RITs with various configurations. Demonstrated capabilities include analysis of externally injected ions with mass resolution in excess of 1000 and a mass/charge range of 650 Th as well as tandem mass spectrometry capabilities. The geometric simplicity and performance characteristics of the 4-electrode RIT make it particularly attractive in the development of next generation miniaturized mass spectrometers.

Rectilinear ion trap: concepts, calculations, and analytical performance of a new mass analyzer.

A mass analyzer based on a rectilinear geometry ion trap (RIT) has been built, and its performance has been characterized. Design concepts for this type of ion trap are delineated with emphasis on the effects of electrode geometry on the calculated electric field. The Mathieu stability region was mapped experimentally. The instrument can be operated using mass-selective instability scans in both the boundary and resonance ejection versions. Comparisons of performance between different versions of the device having different dimensions allowed selection of an optimized geometry with an appropriate distribution of higher-order electric fields. Comparisons made under the same conditions between the performance of a conventional cylindrical ion trap and a RIT of 4 times greater volume show an improvement of 40 times in the signal-to-noise ratio resulting from the higher ion trapping capacity of the RIT. The demonstrated capabilities of the RIT include tandem mass spectrometry, a mass resolution in excess of 1000, and a mass/charge range of 650 Th, all in a simple structure that is only 3.5 cm(3) in internal volume.