Microscale Normal Compression Behaviors of Clay Aggregates: A Coarse-Grained Molecular Dynamics Study.

Given the limitations of micromechanical experiments and molecular dynamics simulations, the normal compression process of clay aggregates was simulated under different vertical pressures (P), numbers of particles, loading methods, and environments by a Gay-Berne potential model. On the basis of the variations of particle orientation and the distribution of stacks, the evolution of deformation and stresses was elucidated. The results showed that the effects of the pressure level and loading environment on the deformation were significant. In the range of 0.1-10 MPa, the changes in the void ratio were essentially the evolution of the distribution of stacks determined by attractive short-range van der Waals interactions. The deformation under constant pressure was larger than that under step loading. Because the interactions between clay particles were mainly controlled by mechanical force when in the range of 40-100 MPa, the void ratios under various loading conditions were consistent. It was also found that changes in three-dimensional stresses during compression were dependent on those of the distribution of stacks. In the vacuum environment, owing to the lateral movement of interlocked small stacks, the horizontal stress decreased. The lateral pressure coefficients (k) were greater in an atmospheric environment because the anisotropic particle orientation was relatively less obvious. In the range of 10-100 MPa, when the loading path became longer, k was similar in vacuum but became smaller in an atmosphere. If the initial loading pressure was increased, the number of large stacks sharply increased and the anisotropy was significant in a vacuum environment, which was less prone to lateral expansion. In contrast, more consistent particle arrangements were maintained in an atmosphere. This work will be conducive to explaining experimental observations of long-term ripening.

Quantitative imaging for predicting hematoma expansion in intracerebral hemorrhage: A multimodel comparison.

BACKGROUND: Several studies report that radiomics provides additional information for predicting hematoma expansion in intracerebral hemorrhage (ICH). However, the comparison of diagnostic performance of radiomics for predicting revised hematoma expansion (RHE) remains unclear. METHODS: The cohort comprised 312 consecutive patients with ICH. A total of 1106 radiomics features from seven categories were extracted using Python software. Support vector machines achieved the best performance in both the training and validation datasets. Clinical factors models were constructed to predict RHE. Receiver operating characteristic curve analysis was used to assess the abilities of non-contrast computed tomography (NCCT) signs, radiomics features, and combined models to predict RHE. RESULTS: We finally selected the top 21 features for predicting RHE. After univariate analysis, 4 clinical factors and 5 NCCT signs were selected for inclusion in the prediction models. In the training and validation dataset, radiomics features had a higher predictive value for RHE (AUC = 0.83) than a single NCCT sign and expansion-prone hematoma. The combined prediction model including radiomics features, clinical factors, and NCCT signs achieved higher predictive performances for RHE (AUC = 0.88) than other combined models. CONCLUSIONS: NCCT radiomics features have a good degree of discrimination for predicting RHE in ICH patients. Combined prediction models that include quantitative imaging significantly improve the prediction of RHE, which may assist in the risk stratification of ICH patients for anti-expansion treatments.

Automated Plasmonic Resonance Scattering Imaging Analysis via Deep Learning.

Plasmonic nanoparticles, which have excellent local surface plasmon resonance (LSPR) optical and chemical properties, have been widely used in biology, chemistry, and photonics. The single-particle light scattering dark-field microscopy (DFM) imaging technique based on a color-coded analytical method is a promising approach for high-throughput plasmonic nanoparticle scatterometry. Due to the interference of high noise levels, accurately extracting real scattering light of plasmonic nanoparticles in living cells is still a challenging task, which hinders its application for intracellular analysis. Herein, we propose an automatic and high-throughput LSPR scatterometry technique using a U-Net convolutional deep learning neural network. We use the deep neural networks to recognize the scattering light of nanoparticles from background interference signals in living cells, which have a dynamic and complicated environment, and construct a DFM image semantic analytical model based on the U-Net convolutional neural network. Compared with traditional methods, this method can achieve higher accuracy, stronger generalization ability, and robustness. As a proof of concept, the change of intracellular cytochrome c in MCF-7 cells under UV light-induced apoptosis was monitored through the fast and high-throughput analysis of the plasmonic nanoparticle scattering light, providing a new strategy for scatterometry study and imaging analysis in chemistry.