Bauman Moscow State Technical University

Severe acute pancreatitis (SAP) is a life-threatening condition marked by systemic inflammation, organ failure, and high morbidity. Nutritional support plays a critical role in SAP management, with total enteral nutrition (TEN) and total parenteral nutrition (TPN) being the primary approaches. While TEN maintains gut integrity and may reduce complications, TPN bypasses the gastrointestinal tract and may worsen systemic inflammation. This study aimed to perform an updated systematic review and meta-analysis comparing the efficacy and safety of TEN versus TPN in SAP patients. Following PRISMA guidelines, we searched PubMed, Scopus, and Web of Science for randomized controlled trials and cohort studies comparing TEN and TPN in SAP. Outcomes included infection rate, organ failure, mortality, necrosis, surgical intervention, hospital stay, and serum markers (amylase, lipase, albumin, IL-6). Quality assessment was done using ROB2 and NOS tools. GRADE methodology was applied to assess evidence certainty. A total of 23 studies involving 7,674 patients (2,750 TEN and 4,924 TPN) were analyzed. TEN was significantly associated with lower odds of infection (OR = 0.35; 95

Machine learning (ML) is increasingly applied in quantum dot (QD) research to support data analysis, device control, materials optimization, sensing, and theoretical modeling. This review surveys recent ML-based approaches across experimental, applied, and computational QD studies, with emphasis on how data-driven methods are embedded within established physical workflows rather than treated as standalone solutions. ML techniques are examined in the context of automated device tuning, high-throughput characterization, synthesis parameter exploration, chemical and biological sensing, photonic and optoelectronic device analysis, and reduced-order modeling of interacting quantum systems. In most cases, ML serves to replace time-consuming fitting procedures, guide experimental sampling, or approximate computationally intensive simulations. Common methodological patterns include supervised learning on limited datasets, transfer learning across device instances, and hybrid approaches incorporating physical constraints into model design or training objectives. Recurrent limitations are also identified, including dataset bias, restricted cross-laboratory transferability, lack of standardized benchmarks, and limited treatment of uncertainty. Rather than positioning ML as a standalone solution, this work frames it as a complementary tool whose reliability and scientific value depend on integration with physical insight, experimental design, and validation protocols.

Alpha-tantalum on silicon is a promising platform for high-coherence superconducting quantum circuits. However, the growth mechanism of alpha-tantalum on silicon remains poorly understood. We present a comprehensive study on alpha-tantalum film growth on various substrates. The decisive role of a substrate material Debye temperature on the phase selection mechanism in tantalum film growth is experimentally confirmed, contradicting the prior assumptions on substrate temperature influence. Crucially, we confirm that alpha-tantalum starts growing only after a 7-10- nm-thick beta-tantalum sublayer. It results in ranging the critical temperature of alpha-Ta films from 3.77 to 4.39 K for the total thickness from 20 to 150 nm, respectively. Finally, we compared high-quality Al and Ta coplanar resonators on silicon, demonstrating compact tantalum resonators (4/10.5/4 mu m) with an internal quality factor exceeding 10 & times; 10(6) at single-photon excitation powers.

Diamond/Cu composites have attracted considerable attention for thermal management applications due to their outstanding thermal conductivity. Recent studies have demonstrated that interconnected diamond network structures can significantly enhance the heat conduction of diamond/Cu composites, while the underlying microscopic heat transfer mechanisms remain to be fully elucidated. This study employed machine learning molecular dynamics simulations to validate and elucidate the heat conduction enhancement mechanisms of interconnected network structures. We developed a machine learning potential for diamond/Cu heterostructures based on the neuroevolution potential model and calculated the lattice thermal conductivity (LTC) of two series of network structures: diamond network/Cu (DN/Cu) and Cu network/diamond (CuN/D). The results demonstrate that DN/Cu exhibits monotonically increasing LTC with diamond content due to continuous heat transfer channels, while CuN/D shows non-monotonic behavior with a minimum LTC at intermediate diamond fractions, originating from the competitive mechanism between contributions from low-frequency Cu phonons (0-8 THz) and high-frequency diamond phonons (10-40 THz) to thermal transport. Through spectral LTC decomposition and wavelike phonon transmission analysis, we elucidated the suppressive effects of multiple heat transfer mechanisms on LTC, including phonon interfacial scattering, coherent interference effects, and total internal reflection in nano-network structures. This work establishes quantitative design thresholds for lattice thermal transport in DN/Cu, revealing a critical diamond volume fraction of similar to 30% above which DN/Cu consistently outperforms CuN/D, and an unexpected LTC minimum at similar to 50% diamond in CuN/D structures driven by phonon frequency competition. These findings explain why diamond network structures, despite demonstrating significant heat conduction enhancement compared to dispersed particles, still exhibit LTC values substantially lower than pure diamond, thus revealing substantial room for design optimization of network architectures.

Carbon fiber–reinforced aluminum laminates (CARALL) combine the high specific stiffness/strength of CFRP with the ductility and damage tolerance of aluminum, offering strong potential for weight-critical transportation structures. Yet, reported flexural responses remain fragmented across studies that separately address laminate architecture, thermo-assisted forming, and damage mechanisms, limiting unified and transferable design guidance. This review consolidates state-of-the-art knowledge on CARALL bending, covering laminate constituents and stacking design, manufacturing and forming routes, experimental characterization, coupled flexural mechanics and damage evolution, and numerical/multiscale modeling. We emphasize how stacking sequence and ply orientation, metal-to-composite thickness ratio, and interface integrity govern stiffness/strength, springback, and delamination-dominated failure, and we compare cohesive-zone and progressive damage implementations against experimental evidence with consistent verification/validation practices. Emerging directions are highlighted in interface–process–performance co-design, open benchmarking datasets and standardized reporting for reproducible calibration from coupon to component scales, digital-twin and data-driven frameworks for forming and crash-relevant structures, and sustainability-oriented end-of-life considerations, including lessons transferable from non-aluminium FML systems (e.g., Ti-FML interfaces). Overall, this review provides an integrated basis for optimizing CARALL flexural performance and accelerating robust deployment in complex service environments.