Gastric function is regulated by the gut-brain axis, which integrates vagal and enteric nervous system (ENS) pathways. The parasympathetic circuit within the vagal pathway promotes digestion by stimulating peristaltic activity and relaxing the pyloric sphincter (PS) through motor and sensory neurons. In contrast, the sympathetic pathway inhibits digestion by suppressing peristalsis and constricting the PS, highlighting the complex neural coordination involved in gastric regulation. We introduce a novel mathematical model of the gut-brain axis using a computationally efficient compartmental modeling framework. The model simulates the vagal and ENS pathways and their corresponding effects on gastric function to enhance our understanding of gut-brain axis regulation. We employ the Michaelis-Menten equation with a Hill coefficient to capture neurotransmitter release at neuromuscular junctions by stimulation of motor neurons and its effects on gastric cells. Motor, or efferent, neurons are modeled for three key stomach regions: the fundus, which exhibits tonic activity; the antrum, which exhibits phasic activity; and the PS, which exhibits both tonic and phasic activity. Thus, the stomach is represented as a three-compartment model. The stomach model extends our previous work by incorporating passive stress and dynamic changes in stomach geometry. Sensory, or afferent, inputs are represented through linear equations that account for chemo- and mechanoreceptor activity, while a binary variable captures the sympathetic response. Afferent and efferent firing rates are linked via fitted curves to effectively close the gut-brain axis feedback loop, borrowing from a similar approach used to model cardiovascular regulation. The simulation results align with physiological observations, demonstrating inhibitory digestive activity during sympathetic responses and excitatory activity, such as gastric emptying, during parasympathetic responses. During gastric emptying, the interstitial cells of Cajal activity shows constant amplitude for low to medium gastric volumes but exhibits an increase in amplitude at very high gastric volumes. Furthermore, gastric emptying rates decrease with high-calorie liquids due to PS regulation. The flexibility of the model allows for future enhancements based on newly discovered signaling pathways in gut-brain circuitry. The computational efficiency of the model suggests its potential use in developing vagal stimulation therapies for gastrointestinal disorders using closed-loop model-based control.
Performance fatigability during the 30-s sit-to-stand (STS) test is not well characterized despite its potential to detect early functional decline. Therefore, this study aimed to quantify temporal changes in power, trunk flexion and movement subphase durations during the 30sSTS, and to examine differences by age and sex. 93 middle-aged adults (50 males and 43 females; mean age 60.5 ± 3.0 years) and 102 older adults (48 males and 54 females; mean age 71.5 ± 5.0 years) performed a 30sSTS. Inertial measurement units (IMUs) mounted over the L4/L5 vertebral level were used to capture sit-to-stand power, trunk flexion and subphase durations in the first and last 10 s. Linear mixed-effects models evaluated temporal changes and group effects. Mean power declined (-11.9 W, d = -0.66), trunk flexion increased (+1.35°, d = 0.42), sit-to-stand duration lengthened and stand-to-sit duration decreased throughout the test (all p < 0.001). The within-test decrease in stand-to-sit duration was less pronounced in older compared to middle-aged adults (d = 0.42, p = .039). Older adults generated less power and spent more time in all subphases (p < 0.05). Females produced less power and greater trunk flexion (p < 0.001). The 30sSTS captures modest performance fatigability; but longer protocols may better reveal clinically meaningful decline. Future research should investigate mobility-limited individuals or examine associations with functional outcomes (frailty, mobility, balance) to provide additional insight.
Soft interfaces formed by lipid membranes are fundamental to living cells, synthetic cells, and membrane-based soft materials. However, a quantitative framework linking molecular organization with nonlinear interfacial mechanics remains elusive. Here, we establish an analytical framework that captures the nonlinear elastic response of lipid-membrane-coated synthetic cells under micropipette aspiration. Incorporating both area stretching and curvature bending enables the model to quantitatively reproduce the complete pressure-displacement response within the small-deformation regime. This approach reduces interfacial mechanics to two parameters: the in-plane area-stretching modulus and an out-of-plane bending-related term. Using this unified framework, we experimentally demonstrate that nonlinear interfacial mechanics can be programmed by altering the molecular geometry and effective dimensionality of adsorbed elements. The lipid molecular shape and curvature-dependent packing regulate in-plane stiffness, whereas DNA nanostructures, the other adsorbed element, introduce an orthogonal control axis via dimensionality: three-dimensional network architectures markedly reinforce bending resistance. Together, these results establish a general molecular design principle for programming interfacial mechanics and provide a quantitative foundation for engineering mechanically tunable synthetic cells and soft interfaces.
Atrial fibrillation (AF) is a major complication following embolic stroke of undetermined source (ESUS), elevating the risk of recurrent stroke and mortality. Early identification is clinically important, yet existing tools face limitations in accuracy, scalability, and cost. Machine learning (ML) offers promise but is hindered by small ESUS cohorts and high-dimensional medical features. To address these challenges, we introduce supervised and unsupervised hypergraph-based pre-training strategies to improve AF prediction in ESUS patients. We first pre-train hypergraph-based patient embedding models on a large stroke cohort (7,780 patients) to capture salient features and higher-order interactions. The resulting embeddings are transferred to a smaller ESUS cohort (510 patients), reducing feature dimensionality while preserving clinically meaningful information, enabling effective prediction with lightweight models. Experiments show that both pre-training approaches outperform traditional models trained on raw data, improving accuracy and robustness. This framework offers a scalable and efficient solution for AF risk prediction after stroke.
Postnatal mouse retinal vascular development is a widely used model for studying retinal vascular diseases and evaluating candidate therapies. This is particularly relevant for inherited disorders such as familial exudative vitreoretinopathy (FEVR), in which impaired vascular growth and organization are central to disease pathogenesis. Numerous approaches have been used to assess retinal vasculature in mouse flat mounts, ranging from qualitative descriptions to limited quantitative measurements of vascular growth. However, phenotypic variability across genetic models, including different models of FEVR, complicates comparisons and underscores the need for standardized, comprehensive multi-parameter analyses that are suitable for rapid and cost-effective screening studies. We describe a standardized morphometric protocol using ImageJ software to quantitatively analyze mouse retinal vasculature in a reproducible manner. The protocol begins with measurement of areas of vascular disorganization (meshes) as well as total vascular and retinal area. Two defined regions in the peripheral and midperipheral retina are then selected to quantify cell clusters, followed by image processing, binarization, and skeletonization. From these processed images, vascular density, branch number, branch length and thickness, junction number, triple points, and box-counting fractal dimension and lacunarity are quantified. Overall, this protocol provides a rapid, cost-effective, and standardized framework for quantifying retinal vascular phenotypes across diverse mouse models. By capturing multiple structural features and accommodating phenotypic variability, it is well-suited for comparative studies and therapeutic screening in retinal vascular disease. Key features • Computational method for mouse retina vessel image analysis for multi-parameter vascular quantification for user-selected regions of interest. • Free open-source ImageJ-based workflow combining disorganization mapping, skeletonization, and fractal analysis for reproducible vascular network characterization. • Optimized for rapid, cost-effective screening of structural vascular outcomes across developmental stages, disease states, and therapeutic interventions.
Chronic pain (CP) affects one in five Canadians, representing a major public health issue. Its economic burden is estimated between CAD $38.3-$40.4 billion annually. Despite its prevalence, integration of nursing activities in CP management within primary care remains limited, indicating a need for strategies that facilitate effective implementation. Determine implementation strategies to promote the integration of nursing activities in CP management in primary care, considering nursing practice realities and patient needs. A sequential explanatory mixed-methods study (QUAN→qual) was conducted in collaboration with two patient partners. A Delphi study captured prioritized nursing activities for CP management, while qualitative focus groups explored nurses' barriers and facilitators in CP management. Findings were integrated through a joint display comparing quantitative and qualitative results following Fetters' (2019) recommendations, in order to identify implementation strategies. Findings indicated a substantial gap between prioritized nursing activities for CP management and current practice, highlighting that nursing involvement remains at an early stage of implementation. Two key orientations emerge: (1) Strengthening nurses' capacities to integrate CP-related activities by providing appropriate tools, training, and clinical guidance, (2) Mobilizing other healthcare system interest-holders (patients living with CP, healthcare providers, and managers) to support adoption of new nursing activities and create an environment conducive to change. The study underscores the importance of implementation strategies to foster the adoption of nursing activities in CP management. These strategies provide an implementation blueprint for maximizing nurses' contribution to CP management in primary care through full deployment of their competencies. Contexte: La douleur chronique (DC) touche un Canadien sur cinq avec un fardeau économique annuel estimé entre 38,3 et 40,4 milliards de dollars. Malgré son importance, l’intégration d’activités infirmières pour la gestion de la DC en soins primaires demeure limitée, nécessitant des efforts concertés pour optimizer leur mise en œuvre.Objectif: Déterminer les stratégies d’implantation adaptées pour favoriser l’intégration des activités infirmières, en tenant compte des réalités de la pratique et des besoins des personnes soignées.Méthode: Une étude mixte séquentielle explicative (QUAN→qual) a été réalisée avec deux patientes partenaires: 1) une étude Delphi pour identifier les activités infirmières prioritaires en DC, 2) des groupes de discussion pour explorer les obstacles et les leviers à la mise en œuvre de ces activités, 3) une intégration mixte à l’aide de tableaux croisés afin d’identifier des stratégies de mise en œuvre adaptées.Résultats: Un écart important existe entre les activités infirmières prioritaires en DC et la pratique actuelle, révélant un stade précoce de mise en œuvre. Deux orientations principales émergent: (1) renforcer les capacités du personnel infirmier en DC grâce à des outils, formations et guides cliniques appropriés, et (2) sensibiliser et mobilizer les autres acteurs du système de santé (patients vivant avec la DC, professionnels de santé et gestionnaires) pour soutenir l’adoption des activités et créer un environnement favorable au changement.Conclusion: L’étude montre que des stratégies adaptées sont essentielles et propose une feuille de route pour intégrer les activités infirmières en gestion de la DC et maximizer leur contribution en soins primaires.
Polygenic risk scores (PRS) for dementia are increasingly used to capture genetic susceptibility, yet their predictive utility may depend on coexisting health and social conditions. Edentulism, an extreme marker of oral health deterioration and accumulated disadvantage, may influence how genetic risk manifests in later life. This study aims to investigate the independent and joint associations of edentulism and dementia PRS with incident all-cause dementia and mortality risk. We analyzed longitudinal data from the Health and Retirement Study linked to Medicare claims. The sample included 9,806 dementia-free participants aged ≥67 with information on edentulism and PRS for Alzheimer's disease. Edentulism was self-reported and logically imputed across waves. PRS was categorized as low, intermediate, and high. Outcomes included incident dementia and all-cause mortality. Cox proportional hazards and Fine-Gray competing risk models were used to examine independent and interactive associations. Edentulism was not independently associated with dementia risk after full adjustment. High PRS was associated with increased dementia risk (subdistribution hazard ratio [sHR] 1.20; 95% confidence interval [CI], 1.06-1.39). However, the association between PRS and dementia was weaker among edentulous participants (interaction sHR 0.73; 95% CI, 0.54-0.99). In contrast, mortality risk was higher among individuals with both edentulism and high PRS (interaction sHR 1.36; 95% CI, 1.00-1.84). These findings suggest that oral health status may modify the association between genetic risk and dementia in late life. Mortality may occur before a diagnosis of dementia, underscoring the importance of considering competing health risks when interpreting genetic susceptibility in aging populations.
Per- and poly-fluoroalkyl substances (PFAS) are contained in various consumer products that include nonstick coatings, packaging materials, cosmetics, and firefighting foams due to their combined hydrophobic and oleophobic properties and chemical and thermal stability. These properties also result in human toxicity and have led to their accumulation in the environment. Several methods are being used to remove PFAS contaminants from the environment, and one common technique involves removal through ion exchange. Polymerization-induced microphase separation (PIMS) enables the synthesis of PFAS-capturing anion exchange beads featuring co-continuous morphology, tunable domain spacing, and high surface accessibility within a mechanically robust crosslinked network. Beads were synthesized using a poly(ε-caprolactone)-b-poly(4-vinylbenzyl chloride)-based macro chain transfer agent, styrene, and divinylbenzene. Anion exchange beads were obtained by etching the poly(ε-caprolactone) component and quaternizing the poly(4-vinylbenzyl chloride), and their ion exchange capacity was measured to be 1.00 ± 0.05 mmol g-1. The rates of PFAS removal were evaluated using pseudo-second-order kinetic analysis for both short-chain (trifluoroacetic acid [TFA] and perfluorobutanoic acid [PFBA]) and long-chain (perfluorooctanoic acid [PFOA]) PFAS. The initial sorption rates of TFA and PFBA were 2.9 and 2.3 times higher, respectively, in quaternized beads (PB-Q) compared to Amberlite IRA 900 whole resin. In contrast, PFOA exhibited a 1.7 times higher initial sorption rate to IRA 900 whole resin than to PB-Q. Langmuir isotherm analysis indicated significantly stronger affinities of all PFAS for PB-Q than IRA 900, even though the IRA 900 had greater capacity, suggesting that PB-Q is more effective for removing PFAS at low concentrations. Treating the PFAS loaded PB-Q beads with a 1 : 1 v/v mixture of methanol and 1 M NaCl(aq) resulted in 100% PFAS desorption.
This protocol describes sc‑rDSeq, a scalable, droplet‑based method for full‑length, strand‑specific total RNA sequencing at single‑cell resolution. The protocol uses a refined set of 220 ribosomal‑depleted sequences (rDS) primers that selectively exclude ribosomal RNA during initial reverse transcription, enabling capture of both polyadenylated and non‑polyadenylated RNAs such as histone RNAs, noncoding RNAs, and enhancer RNAs, without requiring costly post‑amplification depletion steps. This method is useful for researchers who would like to detect not only gene expression variations, but also alternative splicing events and single nucleotide variations in complex heterogenous cellular systems, providing a more complete view of cellular heterogeneity and regulatory programs that remain invisible to conventional polyadenylated‑only sequencing approaches. Compared with existing full‑length protocols, which are often limited by high reagent costs or reliance on complex multistep microfluidics, sc-rDSeq provides a simpler, single-step microfluidic workflow compatible with standard inDrops platforms, which may reduce experimental complexity and cost relative to existing full-length total-RNA methods. A key improvement is the 10-fold increase in unique molecular identifiers per cell relative to 3' end‑based methods, at a reported reagent cost of approximately $0.08 per cell, making deep total transcriptome analysis more accessible. The protocol includes three major parts: sc‑rDSeq barcode synthesis, single‑cell co‑encapsulation, and library construction.
The active regulation of tissue material properties via phase transitions is central in morphogenesis. Transitions occur abruptly at critical points in different control parameters, such as cell density, shape or adhesion. Whether these parameters are interdependent, and perform redundant or distinct functions, is unknown. Here we show that depending on the co-regulation of multiple control parameters, a tissue not only tunes its deformability but also its morphogenetic trajectory. We theoretically define a phase diagram capturing the material states of zebrafish pluripotent tissues undergoing epiboly-a tissue movement occurring during gastrulation-and show that they simultaneously cross critical points in cell density, connectivity and adhesion strength. We then combine optogenetics, biophysical measurements and quantitative morphometrics to independently modulate each parameter in vivo, and identify adhesion as the main determinant of tissue rheology. Further decoupling adhesion from density and inducing adhesion-driven rigidification in unjammed pluripotent tissues is sufficient to switch their morphogenetic program and trigger epithelial organization. This switch in tissue reorganization is achieved via tricellular junction formation, followed by lumenogenesis and the initiation of apical polarity. Our work reveals that the nonlinear dynamics of emergent tissue mechanics are mechanisms of tissue organization and morphogenesis.
Tumor heterogeneity and complex microenvironment interactions drive malignant progression and therapeutic resistance. Traditional omics technologies struggle to capture this complexity due to the loss of spatial molecular context. Spatial multi-omics technologies enable the simultaneous in situ analysis of gene expression, metabolic activity, and protein function within tissue, thereby transforming cancer research. This review systematically discusses the principles and applications of spatial transcriptomics, spatial metabolomics, spatial proteomics, and their integration. These approaches have revealed spatially driving mechanisms, such as the niche at the invasive front, dynamic evolution of the immune microenvironment, and cell-cell communication networks in metastatic spread. We also highlight their potential to promote precise tumor treatment including guiding patient stratification, optimizing treatment regimens, and overcoming drug resistance. Finally, we discuss the current challenges and future development directions. This study aims to clarify how spatial multi-omics deepens the understanding of tumor biology and accelerates clinical translation through multi-dimensional information integration.
In light of the growing global trend toward health awareness, wearable technologies like smartwatches have become essential for monitoring physiological indicators such as heart rate (HR). However, their utility faces two critical challenges: a technical disparity between premium and affordable devices and a conceptual gap where objective HR may fail to capture the true subjective strain experienced by diverse populations. Consequently, this study has a dual objective: to evaluate the HR accuracy of four commercially available watches and examine how measurement variations propagate through a standardized HR-derived energy expenditure model; and to investigate the dissociation between objective HR and subjective perceived exertion (RPE) in individuals with elevated central adiposity. Forty healthy adults ( n = 40 ) participated in a 45 min multi-stage exercise protocol consisting of stretching, cycling, and running. Participants were stratified into two groups based on their waist-to-height ratio (WHtR): normal ( ≤ 0.5) and elevated (>0.5). Data were synchronized using a temporal alignment procedure, and calorie expenditure was calculated through a standardized heart-rate-based regression model to ensure fair comparisons across all devices. Premium smartwatches, specifically the Apple Watch and Garmin, demonstrated superior HR precision across all activity phases, maintaining high correlations ( r ≥ 0.98 ) with the clinical reference. While the low-cost Xiaomi and ThaiSook watches exhibited higher HR errors during motion-intensive activities, their derived calorie expenditure estimates remained remarkably stable and consistent with the reference standard. Notably, individuals with elevated WHtR reported significantly higher Ratings of Perceived Exertion (RPE) during running and recovery phases ( p < 0.05 ), despite showing no significant difference in heart rate responsiveness compared to leaner participants. This study confirms that while higher-end sensors offer greater heart rate precision, affordable wearables can provide sufficient HR data to yield consistent energy expenditure estimates when using a standardized mathematical model, supporting their potential utility in large-scale health monitoring. The divergence between objective heart rate and subjective exertion in participants with central adiposity indicates that heart rate alone is an insufficient gauge of exercise intensity. Consequently, personalized weight-management programs should integrate wearable-derived metrics with perceived effort to better account for the unique physiological and psychological strain associated with higher body mass.
Illness trends are typically monitored by reportable disease and syndromic surveillance systems, but unanticipated health issues might not be captured. Using diagnosis codes, the New York City Health Department developed a novel data mining process to detect unusual increases in emergency department (ED) visits for any reason. We applied the tree-temporal scan statistic in TreeScan software to ICD-10-CM diagnosis codes for ED visits. We searched for unusual citywide increases in ED visits or hospital admissions, over any recent time period, and at any part of and level on the ICD-10-CM tree. We conducted proof-of-concept analyses for March 2020 when COVID-19 emerged, then investigated signals detected in daily, automated analyses during April-August 2025. If TreeScan analyses had been in place, then increasing hospital admissions for viral pneumonia (J12) would have triggered a signal on March 13, 2020, two days before widespread COVID-19 community transmission was announced. An extreme heat event in June 2025 triggered a signal for admissions for acute kidney failure (N17), prompting outreach to dialysis networks. A sustained signal for hand, foot, and mouth disease (B08.4) prompted outreach to child care programs. Other signals supported situational awareness, including a seasonal increase for swimmer's ear (H60.33) and burns (T30.0) related to consumer fireworks. TreeScan quickly detected credible increases in various diagnoses without pre-specification, from minor to severe, rare to common, acute to sustained, and foreseen to unforeseen. TreeScan can strengthen surveillance for health issues related to new pathogens, non-notifiable conditions, environmental exposures, and mass gatherings.
For study sponsors, clinical trial efficiency and data accuracy are non-negotiable drivers of success. Despite broad digitization in healthcare, redundant data entry persists, and the current reliance on manual transcription from electronic health records (EHRs) into electronic data capture (EDC) systems creates unnecessary complexity- leading to higher costs, site burden, and data quality risks. With the growing maturity of interoperability standards and frameworks, direct EHR-to-EDC integration offers a transformative opportunity to automate data transfer, improve accuracy, and streamline operations. This paper evaluates the impact of direct EHR-to-EDC integration on data quality compared to traditional workflows. Our findings in a multi-study, multi-site assessment demonstrate that manual transcription resulted in an error rate of 8.23% (95% CI: 7.79-8.70), whereas no transcription errors were detected in the integrated EHR-to-EDC workflow (95% CI: 0.00-0.03). These results underscore the potential of interoperability-based automation to improve data reliability and operational efficiency in clinical research.
Photosensitization of molecular catalysts (MCs) for selective CO2-to-CO conversion offers a sustainable pathway toward renewable-driven circular carbon utilization but hinges on the rapid and directional transfer of redox equivalents from the photosensitizer (PS) to the catalyst. As a part of our quest in this direction, we report the photosensitization of a nonheme iron MC, ([Fe(AAP)2Cl2]Cl) (Fe Cat), by colloidal, heavy-metal-free bare Cu-deficient CuInS2 (CIS) and core/shell CuInS2/ZnS (CIS/ZnS) quantum dots (QDs) for photocatalytic CO2 reduction to CO in water under 440 nm irradiation. The CIS/ZnS/Fe Cat assembly delivers a CO turnover number (TONCat) of ∼4536 over 23 h, substantially outperforming the unshelled CIS/Fe Cat system (TONCat∼1545). Transient absorption spectroscopy over femtosecond-to-microsecond time scales reveals that ZnS shelling decelerates the photoinduced electron transfer from the QDs to Fe Cat but more than compensates for it by suppressing charge recombination via back electron transfer to a greater extent. This and the surface curing brought about by formation of the shell are responsible for the enhanced catalytic output. Spectroelectrochemical and photochemical analyses further elucidate the sequential electron-transfer steps and CO2 activation events associated with Fe Cat under operating conditions. Together, these results establish a tunable QD/MC architecture, enabled by shell engineering, for exclusive aqueous CO2-to-CO conversion and provide mechanistic design rules for maximizing interfacial charge transfer in photosensitized molecular catalysis.
Policy, systems, and environment (PSE) change interventions are an evidence-based approach to take in rural areas of the US to improve physical activity (PA) rates. However, there is a need to understand the unique factors that affect such interventions in rural areas. This study used a descriptive, qualitative design to capture the perspectives of 25 key informants across eight participating counties as part of a federally funded project. After purposive recruiting of participants via community-based coalition meetings and snowball sampling, trained researchers conducted semi-structured interviews via Zoom. Multiple team members developed a hierarchical coding structure of facilitators of and barriers to PA-related PSE change, including policy, systems, environmental, and interpersonal factors. Systems factors, when conceptualized using a resources-based model of public health capacity, were more often mentioned as facilitating or presenting a barrier to PSE change than were interpersonal, policy, or environmental factors. Specifically, the existence (or absence) of organizations and their financial and operational priorities were critical determinants of PA PSE change. The findings from this qualitative descriptive study can be used by rural practitioners to identify key facilitators of community-based PA PSE change to focus on based on the size and scope of the environmental change and may offer guidance for researchers doing similar work.
Wearable accelerometer data capture rich behavioral signals relevant for personalized health, yet the comparative evidence on modern representation-learning approaches remains limited. Using accelerometer data from the National Health and Nutrition Examination Survey (NHANES), we evaluated three representation families for predicting multiple clinical outcomes: simple entropy-based features, pretrained large-language-model (LLM) embeddings, and time-series foundation model embeddings. Outcomes included overweight status, lipid biomarkers, glucose, arthritis, and cancers. Across all endpoints, entropy-based features consistently performed comparably to, and often slightly better than, embedding approaches. LLM-derived embeddings offered only marginal improvements (ΔAUC≈0.01-0.05), and time-series foundation model embeddings provided minimal value across varying sequence lengths. Prompt-based LLM reasoning performed worst (AUC≈0.56-0.65), demonstrating limited ability to infer quantitative physiological states from structured text. These results highlight the strength of simple variability features and underscore the need for domain-aligned pretraining in future time-series foundation models for health sensing.
Replacing fossil feedstocks with renewable, drop-in intermediates offers a rapid defossilization strategy while exploiting existing infrastructure. This study quantifies the cradle-to-gate greenhouse gas emissions and raw material costs for methyl methacrylate (MMA) via a representative ethylene ("C2") route using fossil, biomass, and direct-air-capture CO2 feedstocks. Results suggest that conventional MMA emits 3.43 kg CO2-eq kg-1. Switching to biomass or CO2 feedstocks yields negative emissions of -0.79 and -1.13 kg CO2-eq kg-1, respectively. Uncertainty analysis demonstrates -0.88 to -0.40 kg CO2-eq kg-1 (biomass) and -1.23 to -1.03 kg CO2-eq kg-1 (CO2). Raw-material costs rise from $0.65 kg-1 (fossil) to $1.52 kg-1 (biomass) and $2.66 kg-1 (CO2), driven by renewable ethylene and formaldehyde. The green premium required for cost parity with fossil fuel-based MMA is $0.31-1.04 kg-1 (biomass) and $1.27-2.12 kg-1 (CO2), equivalent to a 16-52% and 63-106% markup. Propagated to consumer products, the highest premium inflates prices by <15% for an acrylic sheet and <1% for higher-value items (rear car lamps; LCD televisions). Eliminating the premium requires carbon prices of $72-296 tn-1 CO2 (biomass) and $269-512 tn-1 CO2 (CO2). These results position biomass drop-ins as a near-term strategy, while CO2-derived options need further improvements to reach cost parity.
Breast cancer remains a leading cause of cancer mortality worldwide, underscoring the urgent need for advanced preclinical models that faithfully recapitulate tumor complexity and improve therapeutic prediction. Traditional two-dimensional cell cultures and animal models often fail to capture the heterogeneous and dynamic nature of the disease, limiting their translational relevance. This review addresses a critical gap in the literature by providing a comprehensive and integrated comparison of three contemporary, complementary platforms-patient-derived organoids, zebrafish xenografts, and organ-on-chip systems-within a unified translational framework for breast cancer research. We evaluate each model's capacity to elucidate key signaling pathways, model molecular and genetic alterations, and recapitulate the biology of major breast cancer subtypes. Furthermore, we critically assess their roles in drug screening, personalized treatment prediction, and the study of resistance mechanisms, explicitly distinguishing between early-stage proof-of-concept studies, validated preclinical applications, and emerging clinical predictability. A key contribution of this work is a practical, decision-oriented synthesis, supported by comparative matrices, that guides researchers in model selection based on biological relevance, throughput, cost, and clinical predictability. We also highlight current challenges in reproducibility, scalability, and clinical validation, and discuss how emerging technologies such as CRISPR/Cas9 gene editing, artificial intelligence, and bioprinting are poised to address these limitations. By synthesizing the distinct advantages and applications of each platform, this review aims to accelerate the rational use of these advanced models in mechanistic discovery, drug development, and the advancement of precision oncology for breast cancer.
Existing convolutional neural networks and Transformers cannot effectively capture fine-grained local lesion features and long-range contextual dependencies simultaneously in field-collected plant images. To address this research limitation, we aim to design an effective lightweight model suitable for plant disease identification in complex field scenarios. This work proposes an improved Vision Mamba network for plant disease classification based on the challenging PlantDoc dataset. Three dedicated modules are embedded into the framework, including the Multi-Scale Feature Fusion Module (MFFM), Adaptive Channel Attention Mechanism (ACAM) and Lightweight Residual Connection (LRC). The MFFM fuses multi-scale texture, shape and semantic lesion features extracted from shallow, medium and deep network layers. The ACAM adaptively highlights disease-related feature channels and suppresses irrelevant background interference. The LRC structure is adopted to relieve the gradient vanishing problem existing in deep selective state space model (SSM) networks. Experimental results on the filtered PlantDoc dataset show that the presented model obtains an overall accuracy of 92.67%, macro precision of 91.83%, macro recall of 91.56% and macro F1-score of 91.70% on independent test samples, which outperforms the original Vision Mamba baseline by 5.33% in accuracy. Five-fold stratified cross-validation achieves stable accuracy at 92.41 ± 0.24%, and paired t-tests prove that the performance improvement is statistically significant with p<0.05. Ablation experiments confirm the combined contribution of the three designed modules. Error analysis and confusion matrix visualization reveal that the main classification errors are derived from high similarity among different plant disease categories. This study fully verifies the application potential of state space models in agricultural computer vision tasks. The proposed method can serve as an efficient technical scheme for intelligent identification of crop diseases and is well applicable to edge device deployment in precision agriculture practice.