Search Results (4,296)

Background/Objectives: Glioblastoma (GBM) exhibits heterogeneous, nonlinear invasion patterns that challenge conventional modeling and radiomic prediction. Most deep learning approaches describe the morphology but rarely capture the dynamical stability of tumor evolution. We propose an AI framework that approximates a latent attractor landscape of GBM morphodynamics—stable basins in a continuous manifold that are consistent with reproducible morphologic regimes. Methods: Multimodal MRI scans from BraTS 2020 (n = 494) were standardized and embedded with a 3D autoencoder to obtain 128-D latent representations. Unsupervised clustering identified latent basins (“attractors”). A neural ordinary differential equation (neural-ODE) approximated latent dynamics. All dynamics were inferred from cross-sectional population variability rather than longitudinal follow-up, serving as a proof-of-concept approximation of morphologic continuity. Voxel-level perturbation quantified local morphodynamic sensitivity, and proof-of-concept control was explored by adding small inputs to the neural-ODE using both a deterministic controller and a reinforcement learning agent based on soft actor–critic (SAC). Survival analyses (Kaplan–Meier, log-rank, ridge-regularized Cox) assessed associations with outcomes. Results: The learned latent manifold was smooth and clinically organized. Three dominant attractor basins were identified with significant survival stratification (χ² = 31.8, p = 1.3 × 10⁻⁷) in the static model. Dynamic attractor basins derived from neural-ODE endpoints showed modest and non-significant survival differences, confirming that these dynamic labels primarily encode the morphodynamic structure rather than fixed prognostic strata. Dynamic basins inferred from neural-ODE flows were not independently prognostic, indicating that the inferred morphodynamic field captures geometric organization rather than additional clinical risk information. The latent stability index showed a weak but borderline significant negative association with survival (ρ = −0.13 [−0.26, −0.01]; p = 0.0499). In multivariable Cox models, age remained the dominant covariate (HR = 1.30 [1.16–1.45]; p = 5 × 10⁻⁶), with overall C-indices of 0.61–0.64. Voxel-level sensitivity maps highlighted enhancing rims and peri-necrotic interfaces as influential regions. In simulation, deterministic control redirected trajectories toward lower-risk basins (≈57% success; ≈96% terminal distance reduction), while a soft actor–critic (SAC) agent produced smoother trajectories and modest additional reductions in terminal distance, albeit without matching the deterministic controller’s success rate. The learned attractor classes were internally consistent and clinically distinct. Conclusions: Learning a latent attractor landscape links generative AI, dynamical systems theory, and clinical outcomes in GBM. Although limited by the cross-sectional nature of BraTS and modest prognostic gains beyond age, these results provide a mechanistic, controllable framework for tumor morphology in which inferred dynamic attractor-like flows describe latent organization rather than a clinically predictive temporal model, motivating prospective radiogenomic validation and adaptive therapy studies. Full article

With the increasing power output of aero-engines, combustor hot-gas mass flow rate and temperature continue to rise, posing more severe challenges to combustor structural cooling design. To enhance the film-cooling performance of the Z-profile feature in a reverse-flow combustor, this study performs a multi-parameter numerical optimization by integrating computational fluid dynamics (CFD), a radial basis function neural network (RBFNN), and a genetic algorithm (GA). The hole inclination angle, hole pitch, row spacing, and the distance between the first-row holes and the hot-side wall are selected as design variables, and the area-averaged adiabatic film-cooling effectiveness over a critical downstream region is adopted as the optimization objective. The RBFNN surrogate model trained on 750 CFD samples exhibits high predictive accuracy (correlation coefficient (R > 0.999)). The GA converges after approximately 50 generations and identifies an optimal configuration (Opt C). Numerical results indicate that Opt C produces more favorable vortex organization and near-wall flow characteristics, thereby achieving superior cooling performance in the target region; its average adiabatic film-cooling effectiveness is improved by 7.01% and 9.64% relative to the reference configurations Ref D and Ref E, respectively. Full article

With the accelerating pace of urbanization, cities are increasingly affected by rainstorm and flood disasters, which pose severe threats to the safety of residents’ lives and property. Existing models are increasingly inadequate in meeting the accuracy requirements for flood simulation in highly urbanized regions. Thus, it is urgent to develop a new method for flood inundation simulation based on high-resolution urban hydrological units. The novelty of the model lies in the novel structure of the high-resolution Urban Hydrological Units model (HRGM), which replaces coarse sub-catchments with a fine-grained network of urban hydrological units. The primary innovation is the node-based coupling strategy, in which the HRGM provides precise overflow hydrographs at drainage inlets as point sources for LISFLOOD-FP, rather than relying on diffuse runoff inputs from larger areas. In this paper, a high-resolution hydraulic model (HRGM) based on urban hydrological units coupled with a 2D hydrodynamic model (LISFLOOD-FP) was constructed and successfully applied in the Chebeichong watershed. Results show that the model’s simulations align well with observed data, achieving a Nash efficiency coefficient above 0.8 under typical rainfall events. Compared with the SWMM model, the simulation results of HRGM were significantly improved and more consistent with measured results. Taking the rainstorm event on 10 August 2021 as an example, the Nash coefficient increased from 0.7 to 0.85, while the peak flow error decreased markedly from 15.8% to 3.1%. It should be emphasized that urban waterlogging distribution is not continuous but appears as patchy, discontinuous, and fragmented patterns due to the segmentation and blocking effects of roads and buildings in urban areas. The framework presented in this study shows potential for application in other regions requiring flood risk assessment at urban agglomeration scales, offering a valuable reference for advancing flood prediction methodologies and disaster mitigation strategies. Full article

This paper proposes a physics-informed global path planning framework for underwater vehicles integrating CFD simulation and the genetic algorithm. The CFD simulation models the flow field along the planned path of the underwater vehicle. The current velocity data are incorporated into the following path planning that is based on an improved genetic algorithm (GA), which uses migration operators to share the information about feasible solutions or paths, improving the fitness of the whole population. In the three steps of the GA procedure, an elite selection strategy is adopted to avoid losing excellent solutions. A segmented crossover strategy is adopted to avoid low-quality crossover. An adaptive mutation strategy is used to enhance the ability to escape a local optimal solution. Using the improved GA, single-target and multi-target underwater path planning are investigated. In multi-target path planning, a combined algorithm is proposed to solve the optimal traversal order of target points and plan a feasible path between target points. The simulation results show that the proposed algorithm has good planning ability for both simple and complex underwater scenarios. Compared with the conventional GA and an improved GA, the number of average iterations decreases by 45.3% and 29.9%, respectively, for 2D multi-target path planning. The number of average inflection points decreases by 50.3% and 44.2%, respectively, for 2D multi-target path planning. Full article

The production of a stable and uniform spray is a primary concern in fuel atomization applications, such as in fluid catalytic cracking reactors, directly affecting the process quality and gas emissions. However, depending on nozzle geometry and operating conditions, undesired pulsed spray behavior may occur. This phenomenon originates from the internal multiphase flow interaction in Y-jet nozzles and leads to unstable sprays. Understanding the formation of spray pulsations is challenging due to limited internal flow visualization in the nozzle and the fast dynamics involved. Accordingly, this work elucidates the mechanisms of the pulsed spray formation through 3D transient numerical multiphase simulations inside a mixing chamber. The model is validated against internal pressure measurements and applied to investigate the internal mixing behavior across several operating conditions. Results show that the liquid-to-gas momentum flux ratio governs the internal flow regimes. A higher liquid momentum flux obstructs the gas flow, leading to periodic spray bursts when the gas overcomes the liquid back pressure. The simulations also reveal self-sustained oscillatory flow patterns and cyclic transitions between gas penetration and liquid accumulation, which produce periodic pressure fluctuations and nozzle discharge pulsations. The findings offer valuable guidance for optimizing nozzle operation and geometry to suppress pulsation and improve atomization performance. Full article

Microfluidic technologies have ushered in a new era in cardiac tissue engineering, providing more predictive in vitro models compared to two-dimensional culture studies. This review examines Organ-on-a-Chip (OoC) and Lab-on-a-Chip (LoC) platforms, with a specific focus on cardiovascular applications. OoCs, and particularly Heart-on-a-Chip systems, have advanced biomimicry to a higher level by recreating complex 3D cardiac microenvironments in vitro and dynamic fluid flow. These platforms employ induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs), engineered extracellular matrices, and dynamic mechanical and electrical stimulation to reproduce the structural and functional features of myocardial tissue. LoCs have introduced miniaturization and integration of analytical functions into compact devices, enabling high-throughput screening, advanced diagnostics, and efficient pharmacological testing. They enable the investigation of pathophysiological mechanisms, the assessment of cardiotoxicity, and the development of precision medicine approaches. Furthermore, progress in multi-organ systems expands the potential of microfluidic technologies to simulate heart–liver, heart–kidney, and heart–tumor interactions, providing more comprehensive predictive models. However, challenges remain, including the immaturity of iPSC-derived cells, the lack of standardization, and scalability issues. In general, microfluidic platforms represent strategic tools for advancing cardiovascular research in translation and accelerating therapeutic innovation within precision medicine. Full article

Headspace (HS) in anaerobic batch biodigesters is a critical design parameter that modulates pressure stability, gas–liquid equilibrium, and methanogenic productivity. This systematic review, guided by PRISMA 2020, analyzed 84 studies published between 2015 and 2025, of which 64 were included in the qualitative and quantitative synthesis. The interplay between headspace volume fraction ${\mathrm{V}}_{\mathrm{H}\mathrm{S}}/{\mathrm{V}}_{\mathrm{t}\mathrm{o}\mathrm{t}}$, operating pressure, and normalized methane yield was assessed, explicitly integrating safety and instrumentation requirements. In laboratory settings, maintaining a headspace volume fraction (HSVF) of 0.30–0.50 with continuous pressure monitoring P(t) and gas chromatography reduces volumetric uncertainty to below 5–8% and establishes reference yields of 300–430 NmL CH₄ g⁻¹ VS at 35 °C. At the pilot scale, operation at 3–4 bar absolute increases the CH₄ fraction by 10–20 percentage points relative to ~1 bar, while maintaining yields of 0.28–0.35 L CH₄ g COD⁻¹ and production rates of 0.8–1.5 Nm³ CH₄ m⁻³ d⁻¹ under OLRs of 4–30 kg COD m⁻³ d⁻¹, provided pH stabilizes at 7.2–7.6 and the free NH₃ fraction remains below inhibitory thresholds. At full scale, gas domes sized to buffer pressure peaks and equipped with continuous pressure and flow monitoring feed predictive models (AUC > 0.85) that reduce the incidence of foaming and unplanned shutdowns, while the integration of desulfurization and condensate management keep corrosion at acceptable levels. Rational sizing of HS is essential to standardize BMP tests, correctly interpret the physicochemical effects of HS on CO₂ solubility, and distinguish them from intrinsic methanogenesis. We recommend explicitly reporting standardized metrics (Nm³ CH₄ m⁻³ d⁻¹, NmL CH₄ g⁻¹ VS, L CH₄ g COD⁻¹), absolute or relative pressure, HSVF, and the analytical method as a basis for comparability and coupled thermodynamic modeling. While this review primarily focuses on batch (discontinuous) anaerobic digesters, insights from semi-continuous and continuous systems are cited for context where relevant to scale-up and headspace dynamics, without expanding the main scope beyond batch systems. Full article

This paper presents a method for determining the equivalent circuit parameters of magnetically coupled air-core coils used in wireless power transfer (WPT) systems. The proposed approach enables fast and accurate modeling of inductively coupled energy transfer structures, which is essential for the design and optimization of high-efficiency wireless energy systems. The equivalent circuit of the analyzed system was developed using Cauer circuits, while a two-dimensional (2D) axisymmetric electromagnetic field model was employed to derive the equations. The model was implemented in proprietary software based on the edge-element finite element method (FEM) using the A–V formulation. The A–V formulation combines the magnetic vector potential A and the electric scalar potential V, enabling simultaneous representation of magnetic field distribution and current flow in conducting regions. The eddy currents in the conductors were considered in the electromagnetic field analysis. Simulations were carried out for two operating states: short-circuit and idle. The results were used to determine the parameters of the horizontal and magnetizing branches of the equivalent circuit of considered system and to analyze the frequency dependence of the resistances and inductances of the coupled coil system. The proposed modeling approach provides an effective and energy-oriented tool for the design of wireless power transfer systems with improved efficiency and reduced computational cost. The proposed method reproduces impedance characteristics with an accuracy of 0.2 × 10⁻³% in the idle state and 1.4 × 10⁻³% in the short-circuit state compared to the full FEM model, while significantly reducing the computation time. Full article

The Axi gold deposit, a low-sulfidation epithermal deposit in the Western Tianshan, China, hosts over 50 t of gold resources and is widely regarded as the result of coupled processes of rock deformation, heat transfer, pore fluid flow, and chemical reactions. However, research on the ore-forming processes of this gold deposit from a coupled perspective remains limited, resulting in its ore-forming mechanisms being incompletely understood. In this paper, we use the concept of mineralization rate based on computational modeling to indicate the 3D spatial distribution of mineralization. The simulation results reveal the following: (1) temperature gradients play a key role in influencing mineral precipitation, whereas the effect of pore fluid pressure gradients is relatively negligible; (2) gold precipitation, characterized by a negative mineralization rate, predominantly took place along fault zones that exhibit vertical transitions from steep to gentle slopes or lateral bends, which are further distinguished by the accumulation of fluids and the presence of significant temperature gradients. Notably, this particular distribution pattern of gold precipitation closely mirrors the spatial arrangement of known gold orebodies. These findings suggest that the coupling of multiple physical and chemical processes at specific fault sites plays a critical role in ore formation, providing new insights into the mechanisms governing the development of the Axi gold deposit. Furthermore, based on these observations, it can be inferred that the deeper regions of the Axi gold deposit hold considerable mineralization potential. Full article

This study evaluates energy conversion and heat transfer in a germicidal chamber employing gold nanorods (AuNRs) irradiated with an infrared laser (808 nm, 0.8 W) to generate heat via localized surface plasmon resonance. The investigation focused on the preliminary selection of chamber materials and the geometry of the bottom surface supporting the AuNRs as the heat source in a photothermoablation application. A one-way multiphysics and multiscale approach was applied, integrating nanoscale heating phenomena with a macroscale fluid and heat flow. The validated 2D numerical model shows satisfactory agreement with experimental data and is suitable for further design analyses. Computational Fluid Dynamics (CFD) simulations were conducted to determine temperature and entropy distributions, mean and maximum temperatures, and Nusselt numbers, allowing the assessment of the energy conversion process under different configurations and AuNR dimensions. The results indicate that a configuration with a gradually descending stepped structure enhances interactions between nanoparticles and the fluid, increasing the internal energy and producing elevated temperatures. Under optimal conditions, a temperature rise of approximately 75 °C was achieved. These findings demonstrate that integrating material selection, surface geometry, and nanoparticle absorbance optimization can significantly improve the efficiency of bacterial inactivation in germicidal chambers. This study provides a framework for future investigations on fully three-dimensional multiscale and multiphysical modeling, as well as a targeted AuNR design to maximize the thermal performance. Full article

Reconstructing animatable humans, together with their surrounding static environments, from monocular, motion-blurred videos is still challenging for current neural rendering methods. Existing monocular human reconstruction approaches achieve impressive quality and efficiency, but they are designed for clean intensity inputs and mainly focus on the foreground human, leading to degraded performance under motion blur and incomplete scene modeling. Event cameras provide high temporal resolution and robustness to motion blur, making them a natural complement to standard video sensors. We present E-Sem3DGS, a semantically augmented 3D Gaussian Splatting framework that leverages hybrid event-intensity streams to jointly reconstruct explicit 3D volumetric representations of human avatars and static scenes. E-Sem3DGS maintains a single set of 3D Gaussians in Euclidean space, each endowed with a learnable semantic attribute that softly separates dynamic human and static scene content. We initialize human Gaussians from Skinned Multi-Person Linear (SMPL) model priors with semantic values set to 1 and scene Gaussians by sampling a surrounding cube with semantic values set to 0, then jointly optimize geometry, appearance, and semantics. To mitigate motion blur, we derive optical flow from events and use it to supervise image-based optical flow between rendered frames, enforcing temporal coherence in high-motion regions and sharpening both humans and backgrounds. On the motion-blurred ZJU-MoCap-Blur dataset, E-Sem3DGS improves the average full-frame PSNR from 21.75 to 32.56 (+49.7%) over previous methods. On MMHPSD-Blur, our method improves PSNR from 25.23 to 28.63 (+13.48%). Full article

The Dynamic Electric Vehicle Routing Problem (D-EVRP) presents a framework that requires electric vehicles to meet demand with limited energy capacity. When dynamic demand flows and charging requirements are considered together, traditional methods cannot provide sufficient adaptation for real-time decision-making. Therefore, a learning-based approach was chosen to ensure that decision-making processes respond quickly to changing conditions. The solution utilizes a model with a Double Deep Q-Network (DDQN) architecture and a discrete valuation structure. Prioritized Experience Replay (PER) was implemented to increase model stability, allowing infrequent but effective experiments to contribute more to the learning process. The state representation is constructed using the vehicle’s location, battery level, load status, and current customer demands. Scalability is ensured by dividing customer locations into clusters using the K-means method, with each cluster handled by an independent representative. The approach was tested with real-world road data obtained from the Meşelik Campus of Osmangazi University in Eskişehir. Experiments conducted under different demand levels and data sizes have shown that the PER-assisted DDQN structure produces more stable and shorter route lengths in dynamic scenarios, but random selection, greedy method and genetic algorithm experience significant performance losses as dynamicity increases. Full article

Dragon’s Blood, from the Dracaena cochinchinensis plant, is known for enhancing blood circulation. Its main components are Dragon’s Blood phenolic extracts (DBE). To pinpoint the active DBE constituents that are effective against thrombosis and understand their mechanism of action, the PT-stroke model was employed to assess DBE’s antithrombotic effects on cerebral blood flow and platelet aggregation. This investigation demonstrates that DBE enhances cerebral blood flow and inhibits ADP-induced platelet aggregation in photothrombotic (PT) stroke models. An FeCl₃-induced carotid artery thrombosis model was developed to test the antithrombotic activity of four DBE fractions. Through screening with this model, the ethyl acetate (EA) and methanol fractions were identified as the principal active components that effectively reduced thrombus weight and improved hemodynamics. Furthermore, the EA fraction was found to preserve the integrity of the blood–brain barrier. Phytochemical isolation allowed for the identification of compounds in the EA fractions, and UHPLC-MS was performed to characterize DBE and its active components in the bloodstream. In vitro ADP-induced platelet aggregation assays highlighted the active compounds. Through phytochemical analysis, Loureirin D (compound 17) was identified as a predominant constituent present in plasma. In vitro assays revealed that compounds 1 and 17 possess strong antiplatelet activity, with Loureirin D being confirmed as a selective P2Y12 receptor antagonist via molecular docking and cellular thermal shift assays. These findings substantiate Loureirin D as a pivotal antithrombotic component in DBE and its potential as a P2Y12-targeting therapeutic agent for thrombosis treatment. Full article

This numerical study investigates steady separated flow past a grooved circular cylinder within the Reynolds number range $7.5\le R{e}_D\le 30$, comprising variations in groove depth ($h$) and spacing ($\beta$). The groove width ($w$) is kept constant, while h/w varies across four levels ($0.5\le h/w\le 2$) and $\beta$ across five angles ($10°\le \beta \le 45°$). The results exhibit strong agreement with unbounded flow data, confirming blockage independence across the examined regime. Detailed analysis shows that $\beta$ has a stronger influence than $h/w$ on surface-pressure-dependent variables ($C_{p,0}$, $C_{p,b}$, $C_D$, ${\theta }_{sep}$) and wake-defining parameters ($L_w$, $W_w$, $\xi$, $\eta$), underscoring the dominant role of $\beta$ in rectilinear groove aerodynamics. In this regard, a critical spacing of $\beta =20°$ is observed, beyond which the sensitivity of the parameters toward the cylinder configuration decreases. Thus, significant flow control and drag reduction are attained for $R{e}_D=7.5$ at the lowest spacing $\beta =10°$, regardless of the groove’s $h/w$. Among these, the streamwise-oriented variables, $C_{p,0}$, $C_D$, $L_w$, $\xi$, and $u_{min}$, exhibit monotonic trend with respect to $\beta$ and are modeled using power-law relations. The models for $C_{p,0}$ and $C_D$ exhibit significant accuracy with $R^2\approx 0.999$ across all $\beta$ values considered, while it is 0.89–0.98 for $L_w$, $\xi$, and $u_{min}$, depending on $R{e}_D$. Transverse-oriented parameters ($W_w \mathrm{a}\mathrm{n}\mathrm{d} \eta$) vary non-monotonically. In addition, it is found that the streamwise locations of maximum wake width ($x_{w,max}$) and minimum velocity ($x_{u,min}$) are unaffected by the grooved cylinder configuration. Full article

A central challenge in tight fault-depression reservoirs is understanding how three-dimensional fracture structures control gas storage and flow. This study introduces a data-driven, geologically informed framework that integrates structural-mechanical coupling to decipher fracture networks within the Shahezi Formation. Our model, based on rock failure criteria, achieves quantitative fracture prediction across one-dimensional to three-dimensional scales. This capability overcomes the limitations inherent in single-method approaches for tight, fracture-dominated reservoirs. By synthesizing sedimentary facies-controlled reservoir modeling, sweet-spot inversion, and geo-engineering integration, we establish a predictive system for accurate reservoir assessment. The continental clastic Shahezi Formation is typified by secondary fractures. This study utilizes leverage small-scale data (core, thin section, log) to quantify key parameters (fracture density, aperture), enabling a systematic analysis of fracture typology, heterogeneity, and controls. Building on this foundation, and spatially constrained by large-scale datasets (seismic interpretation, stress-field simulations), we developed a robust fracture development model for deep tight reservoirs. Stress-field modeling delineated fracture-prone zones, where a discrete fracture network (DFN) model was built to characterize 3D fracture geometry and connectivity. Integrating simulated fracture size and aperture-derived permeability allowed us to quantify fracture contribution to total permeability, ultimately mapping favorable targets. The results identify favorable zones primarily in the western sector of the study area, forming an NS-trending, belt-like distribution. They are mainly concentrated around the wells Changshen-4, Changshen-40, and Changshen-41. This distribution is clearly controlled by the Qianshenzijing Fault. Full article