.
The ultrasonication-triggered interfacial assembly approach was developed to synthesize magnetic Janus amphiphilic nanoparticles (MJANPs) for cancer theranostic applications, where the biocompatible octadecylamine is used as a molecular linker to mediate the interactions between hydrophobic and hydrophilic nanoparticles across the oil-water interface. The obtained Co cluster-embedded Fe3O4 nanoparticles-graphene oxide (CCIO-GO) MJANPs exhibited superior magnetic heating efficiency and transverse relaxivity, 64 and 4 times higher than that of commercial superparamagnetic iron oxides, respectively. The methodology has been applicable to nanoparticles of various dimensions (5-100 nm), morphologies (sphere, ring, disk, and rod), and composition (metal oxides, noble metal and semiconductor compounds, etc.), thereby greatly enriching the array of MJANPs. In vivo theranostic applications using the tumor-bearing mice model further demonstrated the effectiveness of these MJANPs in high-resolution multimodality imaging and high-efficiency cancer therapeutics. The ubiquitous assembly approach developed in the current study pave the way for on-demand design of high-performance Janus amphiphilic nanoparticles for various clinical diagnoses and therapeutic applications.
Landslides have different topographic and morphological characteristics due to their different triggering mechanisms.However,the differences in the characteristics of earthquake-and rainstorm-induced land-slides remain unclear.In this paper,we collect 12 cases of earthquake-and rainstorm-induced landslides around the world and reveal the differences in characteristics of the two types of landslides.By examin-ing the geometric characteristics and location distribution of the landslides,the results show that earthquake-induced landslides tend to have larger areas,perimeter,lengths,widths,area to perimeter ratios(area/perimeter),major axis(SM),and minor axis(sm)than rainstorm-induced landslides.In addi-tion,earthquake-induced landslides have more complex,rounded,and compact shapes than rainstorm-induced landslides.Earthquake-induced landslides are predominantly clustered near ridges,whereas rainstorm-induced landslides are predominantly clustered near valleys.The results also indicate that earthquake-and rainstorm-induced landslides mostly occur on 30°-50° and 10°-30° slopes,respectively,and both types are more likely to occur on sunny slopes.Moreover,the compactness and major axis are negatively logarithmically correlated for earthquake-induced landslides,while they are negatively expo-nentially correlated for rainstorm-induced landslides.Additional earthquake-and rainstorm-induced landslide events have verified the reliability and extensibility of the research conclusions.This work is beneficial for the management of landslide hazards and the effective implementation of landslide predic-tion and risk assessment.
When determining the location of Coalbed Methane (CBM) wells, pressure variation is a key factor affecting CBM productivity. Understanding the relationship between reservoir pressure and gas production has become a central challenge in CBM development. In this study, drainage data from CBM wells located in a specific area on the eastern margin of the Ordos Basin were analysed to investigate the influence of geological parameters (e.g. coal seam thickness, burial depth) on productivity across different CBM well types. Subsequently, the effect of pressure drop rate on CBM production dynamics was studied. To further explore this relationship, low-field nuclear magnetic resonance (NMR) experiments were conducted to simulate stepwise depressurization and methane desorption. These experiments addressed the impact of pressure drop rate on methane desorption from both primary and fractured coal structures. A numerical simulation model was then developed to evaluate CBM production performance under varying pressure drop rate constraints. This integrated approach considered key factors such as reservoir pressure and gas content to systematically assess the influence of pressure drop rate on CBM productivity. The main findings are as follows: 1) Pressure-production relationship: An exponential relationship was observed between the pressure decline curves of vertical and horizontal wells. The pressure drop stage corresponds to the stages of increasing and stable gas production. A higher pressure drop rate enhances early-stage production in vertical wells but adversely affects their long-term performance. In contrast, horizontal wells exhibit a pronounced decline in productivity under high drawdown rates. 2) Methane behaviour characterization: NMR simulations revealed distinct pressure-dependent behaviours for adsorbed and free methane. Adsorbed methane follows the Langmuir isotherm, and free methane exhibits a linear relationship with pressure. This distinction arises because adsorbed methane is mainly associated with micropore surface area, whereas free methane is governed by pore volume. 3) Desorption dynamics: NMR physical experiments showed that higher pressure drop rates hinder the desorption of adsorbed methane. The desorption efficiency was inversely proportional to the drawdown rate, largely due to the complex pore structures (e.g. ink-bottle-shaped pores) in the coal samples. Rapid pressure changes reduce methane desorption efficiency, and limit methane diffusion and migration. 4) Numerical simulation insights: The simulation results quantitatively characterized the impact of pressure drop rate on CBM productivity. Higher draw-down rates accelerate the vertical expansion of the pressure depletion cone, leading to an earlier production peak and higher initial gas output. However, over time, these wells exhibit weaker lateral expansion of pressure funnel, resulting in lower gas content and pressure gradients, and ultimately lower productivity in the later stages of drainage.
It is crucial to accurately segment organs from abdominal CT images for clinical diagnosis, treatment planning, and surgical guidance, which remains an extremely challenging task due to low contrast between organs and surrounding tissues and the difference of organ size and shape. Previous works mainly focused on complex network architectures or task-specific modules but frequently failed to learn irregular boundaries and did not consider that different slices from the same case might contain targets of different numbers of categories. To tackle these issues, this paper proposes UAMSNet for abdominal multi-organ segmentation. In UAMSNet, a hybrid receptive field extraction (HRFE) module is introduced to adaptively learn the features of irregular targets, which has an adaptive dilation factor containing distance information to facilitate spatial and channel attention. The HRFE module can simultaneously learn multiple scales and deformations of different organs. Furthermore, a multi-organ boundary-enhanced attention (MBA) module in the encoder and decoder is designed to provide effective boundary information for feature extraction based on the large peak of the organ edge. Finally, the difference in the number of organ categories between different slices is first considered using a loss function, which can adjust the loss computation based on organ categories in the image. The loss function mitigates the effect of false positives during training to ensure the model can adapt to small organ segmentation. Experimental results on WORD and Synapse datasets demonstrate that our UAMSNet outperforms the existing state-of-the-art methods. Ablation experiments confirm the effectiveness of our designed modules and loss function. Our code is publicly available on https://github.com/HeyJGJu/UAMSNet.
Medical image segmentation faces two main challenges: first, the difficulty of distinguishing targets from backgrounds due to similar intensity and low contrast, and second, the displacement and deformation of targets and surrounding structures across slices. While existing methods incorporate multi-scale or multi-source data to enhance the network's target recognition, such methods still struggle with subtle boundary differences and interference from adjacent structures, limiting their generalization. To address these challenges, we propose a Collaborative Learning of Dynamic and Static Information Network (CLDSINet), designed to extract both static and dynamic features for precise target segmentation. CLDSINet consists of three key components: the Target-focused Perception (TP) branch, the Deformation-aware Learning (DL) branch, and the Static Dynamic Feature Fusion (SDFF) branch. The TP branch extracts static features via a two-stage strategy, while the DL branch captures inter-slice deformation. The SDFF branch integrates these features through a Cross Feature Aggregation (CFA) module in the encoder and adjusts the contributions of these features with the Dynamic Feature Weight Learning (DFWL) module in the decoder. Experiments on three public and two private datasets show that CLDSINet outperforms existing methods in segmentation accuracy and stability, particularly for low-contrast and highly-deformed images.
