Interest in mechanical computing has surged in recent years, but direct translations from digital to mechanical computation, especially in the context of von Neumann architectures, face scalability challenges. These challenges necessitate alternative strategies beyond miniaturization to enhance information density and facilitate the development of intelligent structures. Neuromorphic architectures that rely on in-memory computing can help overcome these scalability constraints. Additionally, using mechanical properties as a mechanism for information encoding and decoding paves the way to utilize these structures in applications where energy resources are limited or embedded electronics are not feasible. In this paper, we leverage the response of viscoelastic mechanical metamaterial blocks to realize memory augmentation and in-material computation from temporal neural coding. We use an asymmetric bistable viscoelastic mechano-bit abstraction to encode information and a protocol for decoding state information using the global stiffness of the mechanical storage system. We derive fundamental limits on memory capacity for a mechanical memory made from elastic mechano-bits and identify a geometric stiffness design rule to ensure global stiffness uniqueness for each bit elastic memory block combination. The transient memory augmentation is achieved via the combined response of our neuron-inspired viscoelastic mechano-bits (Visco-Bits), showing temporal stiffness modulation that yields augmented mechanical memory storage from the resulting non-abelian, path-dependent mechanical behavior. Subsequently, we examine the information entropy gained by using such viscoelastic mechano-bits and demonstrate that this can exceed the conventional n-bit limit for digital storage. This abstraction of transient memory augmentation can be extended to any mechanical, electrical, or optical system capable of exhibiting temporal modulation in a physical property. We show this via a physical demonstration of a 2-unit viscoelastic memory block and establish the maximum time limit for which the order-dependent memory expansion occurs. Finally, we explore the utility of augmenting the in-memory computational capabilities in viscoelastic metamaterials by temporally encoding two distinct logic gate operations in our Visco-Bit memory block.

The utility of non-functionalized poly(N-isopropylacrylamide) (pNIPAM) microgels in physiological and environmental applications is strictly dependent on their reversible thermoresponsiveness and stability in saline media. Despite their importance, a unified understanding of how network topology, specifically crosslinker concentration and crosslinking distribution regulates ionic sensitivity remains fragmented in the literature. This work systematically investigates the interplay between network topology and ionic strength (0-100 mM NaCl) across eight distinct microgel architectures, ranging from ultra-low crosslinked (ULC) to core-corona and homogeneously crosslinked (HC) variants. Utilizing dynamic light scattering across 22 batches, we analyzed critical thermoresponsive properties, including volume phase transition temperature (VPTT) shifts, salt tolerance thresholds, hysteresis indices, and flocculation kinetics (only at extreme salinity, 1000 mM NaCl and at 25 °C). This comprehensive investigation enables a multi-dimensional analysis of how ionic strength, the presence or absence of crosslinkers (MBA), spatial crosslinking distribution, and thermodynamic states dictate microgel behavior across varying temperatures. Finally, we evaluate the applicability of this experimental dataset to established theoretical frameworks, specifically the Flory-Rehner and Flory-Rehner-Donnan models, making an attempt to address the ongoing debate regarding their validity in describing complex microgels.

The expansion and intensification of agriculture in the Argentine Pampas have increased pesticide use, raising concerns about groundwater contamination in aquifers used for drinking water. In this context, contamination risk assessment is a key tool for environmental management, although results may vary depending on the methodology applied. This study compares two approaches for assessing pesticide-related groundwater contamination risk in the Quequén Grande River watershed (Argentina). Risk was defined as the interaction between aquifer contamination hazard and population vulnerability. Hazard was estimated by combining aquifer vulnerability (DRASTIC-P) with two representations of contaminant load: (i) a land-use-based approach derived from LANDSAT imagery, and (ii) an empirical model based on pesticide leaching experiments relating mobility to soil organic matter. Social vulnerability was assessed using a census-based index. All analyses were conducted within a GIS framework. Results show that the two approaches produce markedly different spatial risk patterns. These differences were evaluated through methodological comparison and preliminary field evidence, but not through a comprehensive independent basin-wide validation. The land-use-based method identified a larger proportion of high-risk areas (20.1%), whereas the laboratory-based approach yielded more restrictive estimates (9.4%) and greater spatial differentiation. These differences reflect the conceptual assumptions underlying each method. The results highlight that method selection strongly influences risk assessment outcomes and should be aligned with management objectives, data availability, and acceptable levels of uncertainty.

Content-based retinal image analysis plays a crucial role in the early diagnosis of ocular diseases. In this study, we proposed a novel approach for efficient content-based retinal image retrieval and Diabetic Retinopathy (DR) detection using variants of local texture features derived from Local Binary Pattern (LBP), Local Ternary Pattern (LTP), and Gray Level Co-occurrence Matrix (GLCM). The methodology begins with meticulous image preprocessing to enhance feature extraction, followed by the extraction of LBP, LTP, and GLCM features, which capture intricate texture patterns and enrich the feature space for robust analysis. Subsequently, we trained machine learning models, including Support Vector Machine (SVM), Decision Tree, and Random Forest, on the extracted features to effectively retrieve retinal images and detect DR. A comparative analysis between preprocessed and raw images highlights the impact of preprocessing techniques on performance. A key innovation of this study lies in the fusion of multiple texture-based features, creating a comprehensive representation that integrates high-level semantic information with fine-grained local patterns. This hybrid approach enhances the system's capability to handle diverse retinal image variations, leading to improved retrieval accuracy and robustness. Further, a metaheuristic approach for feature selection and optimization is employed, comparing Differential Evolution, Genetic Algorithm, and Particle Swarm Optimization to identify the most effective features for retrieval. Differential Evolution achieved the highest precision of 90.67 % for retrieving the top 10 relevant images. The proposed hybrid approach demonstrates the effectiveness of integrating classical image analysis methods with machine learning for DR detection and content-based image retrieval. This research contributes to precision medicine and healthcare innovation by advancing ML-driven retinal image analysis.

Biodegradation is the most sustainable biomass recycling strategy, yet the absence of efficient microbial degraders remains a critical bottleneck. While soil microorganisms can decompose diverse biomass, their functional specificity for distinct agricultural by-products remains inadequately characterized. To address this, five agricultural by-products, including fish skin, soybean meal, shrimp shell, corn straw and chicken feather, were individually or combinatorially incubated in soil. Comparative analysis of 16 S rRNA amplicons and metagenomics from actively degrading microbial communities versus native soil identified key functional degraders. Declines in Chao1 and Shannon indices within biomass-amended soil groups indicated community simplification, driven by the dominance of novel utilizers over indigenous taxa. Genera enriched in native soil were replaced by divergent taxa across biomass types, revealing substrate-dependent community succession. LEfSe analysis identified biomass-specific utilizers at multiple taxonomic levels. Co-occurrence network analysis showed strong positive co-occurrence patterns between significantly enriched operational taxonomic units (OTUs), suggesting potential co-occurrence patterns and shared responses to biomass amendment. FAPROTAX revealed enhanced C/N/S metabolism during biomass utilization. Metagenomic screening identified markedly higher numbers of biomass-degradation genes encoding hydrolases (e.g., proteases, cellulases, chitinases), consistent with significantly elevated enzyme activities in amended soils compared to undetectable levels in controls. Among six candidate OTUs substantially enriched in chicken feather-amended soil, three species demonstrated efficient feather degradation, with some exhibiting multi-substrate capability. This study elucidates substrate-dependent biomass cycling in soil and provides candidate degraders, including Vicinamibacterales-related OTUs, unclassified Enterobacteriaceae, Sphingobacterium paludis, Sphingobacterium griseoflavum, and Lysinibacillus mangiferihumi, as well as enzymatic gene resources for engineered biomass recycling.

This cross-sectional study evaluated the knowledge, skills, attitudes, and perceived barriers of dentists from the Federal District, Brazil, regarding Minimal Intervention Dentistry (MID). A validated online questionnaire was administered using non-probabilistic snowball sampling, and 404 dentists participated (74% female; mean professional experience: 12 ± 10.9 years). Data were analyzed using descriptive and inferential statistics, including multiple linear regression (p < 0.05). The dependent variables were knowledge, skills, attitudes, and barriers, while the independent variables included administrative region, years of professional experience, type of graduation institution, academic qualification, professional role, and information-seeking behavior. Overall, participants demonstrated high levels of competency in MID, particularly in the preventive and diagnostic domains; however, selective carious tissue removal was the procedure performed least frequently (38.6%). Regression analyses showed that dentists involved in teaching or research reported significantly fewer barriers (β = -2.56; 95% CI: -3.90 to -1.23; p < 0.001). Dentists who had received formal training in MID achieved higher scores in knowledge (β = 2.95; 95%CI: 2.31 to 3.59; p < 0.001) and in skills and attitudes (β = 3.40; 95%CI: 2.18 to 4.62; p < 0.001), as well as fewer perceived barriers (β = -1.52; 95%CI: -2.35 to -0.68; p < 0.001). Information-seeking behavior was positively associated with greater knowledge (β = 2.35; 95%CI: 1.56 to 3.14; p < 0.001) and improved skills and attitudes (β = 3.06; 95%CI: 1.55 to 4.56; p < 0.001). Graduates from public universities presented higher scores for skills and attitudes (β = 2.17; 95%CI: 1.15 to 3.18; p < 0.001) and reported fewer barriers (β = -0.88; 95%CI: -1.61 to -0.15; p < 0.01). Brazilian dentists demonstrated a strong theoretical understanding of MID; however, practical implementation remains limited. Academic background, involvement in education and research, and proactive learning behaviors were key determinants of higher competency levels and reduced barriers to the adoption of MID.

Park and colleagues recently reported, in a single-centre Korean cohort of 391 adult patients with myelodysplastic neoplasms (MDS) receiving fludarabine-busulfan conditioning and antithymocyte globulin (ATG)-based prophylaxis, that matched sibling donor (MSD) recipients experienced a substantially higher 5-year cumulative incidence of relapse (34.3%) than recipients of unrelated (UD; 16.4%) or haploidentical (HID; 16.8%) donors, with reduced acute and chronic graft-versus-host disease (GVHD) in the MSD arm. The authors conclude that under ATG-based prophylaxis, MSD may no longer be the optimal donor choice in MDS. We respectfully argue that this conclusion is constrained by two structural features of the cohort. First, ATG dose is severely confounded with donor type: 61.3% of MSD recipients received 10 mg/kg ATG versus 0% of HID recipients, a dose pattern that inverts standard clinical practice and overlaps a range in which independent evidence indicates impaired disease control without offsetting GVHD benefit. Second, the GVHD- non-relapse mortality pattern in the cohort is more consistent with a cumulative immunosuppression-intensity hypothesis than with a primary donor-source effect. We propose that the most clinically actionable inference is to re-examine ATG dosing in MSD transplants for MDS, rather than to displace MSD as a preferred donor.