Search Results (27,601)

In computer-aided diagnosis, accurate skin lesion segmentation is a key step; however, current transformer-based models often fail with blurred or fragmented boundaries, especially in challenging skin lesion images. In this work, we propose a Swin Transformer framework enhanced with a novel Frequency-Aware Multi-Scale Cross-Scale Attention module, named FreqMSCSAM-SwinSegNet. This module separates features into structural (low-frequency) and edge (high-frequency) information, therefore applying an independent attention mechanism to each form of feature information. It also includes a learnable edge extractor, cross-attention skip connections, a dual-attention decoder, and a hybrid loss function that combines Dice, Focal, and boundary losses, and an auxiliary morphological consistency loss that enforces clinically plausible lesion shapes. We evaluate the model on ISIC2018, where it reaches an IoU of 0.9228, a Dice coefficient of 0.9406, and an Accuracy of 0.9713, beating DeepLabV3++, SegFormer, and UNet++. The model also performs well on the ISIC2017 dataset, showing strong generalization. Overall, FreqMSCSAM-SwinSegNet offers a robust and practical solution for clinical skin lesion segmentation, with the potential to improve diagnostic accuracy and early melanoma detection. Full article

Generative artificial intelligence (AI) is driving a shift in mobile banking from static, transaction-focused systems to dynamic engines that generate synthesized financial guidance, rationale, and suggestions tailored to consumers. Research exploring the factors driving mobile banking continuance usage intention and digital financial well-being remains nascent; however, little is known about the role that artificial intelligence system disclosure transparency, end user trust in automated decision-making (“algorithmic trust”), and perceived privacy risk–benefit trade-offs (referred to as “privacy calculus”) play in shaping those critical behavioral outcomes. This study seeks to fill that gap by empirically testing a model of the relationships among explainable AI (XAI) transparency, algorithmic trust, privacy calculus benefits, privacy calculus risk, mobile banking continuance intention, and digital financial well-being among Generation Z (Gen Z) mobile banking consumers in the United States. A quantitative, cross-sectional survey research design was used to collect data from 377 Generation Z (ages 18–29) respondents using a standardized online questionnaire. The data were analyzed using descriptive statistics, Pearson’s correlation, multivariate analysis of variance (MANOVA), and tests of between-subjects effects using IBM SPSS software version 31.0. The results find that XAI transparency significantly predicts both continuance intention (F = 8.69, p = 0.003) and digital financial well-being (F = 66.22, p < 0.001), though in opposite directions across the two outcomes. Algorithmic trust was a statistically significant predictor of both digital financial well-being (F = 320.13, p < 0.001) and continuance intention (F = 10.78, p = 0.001), while privacy calculus risk was a statistically significant, though positive rather than the hypothesized negative, predictor of both continuance intention (F = 36.04, p < 0.001) and digital financial well-being (F = 165.69, p < 0.001); however, because the privacy calculus risk scale showed weak internal-consistency reliability (α = 0.47) and all four predictors were severely intercorrelated (variance inflation factors of 9.19–26.79), these individual coefficients should be interpreted with caution rather than as evidence of four independently distinguishable psychological mechanisms. Privacy calculus benefits significantly predicted digital financial well-being (F = 14.78, p < 0.001) but not continuance intention (F = 2.97, p = 0.085). A very strong correlation was found between continuance intention and digital financial well-being (r = 0.908, p < 0.001). This study discusses theoretical contributions to and managerial implications for the design of explainable AI systems, the generation of algorithmic trust, and privacy considerations in fintech. These findings are timely given the accelerating deployment of large language model (LLM) banking assistants, agentic financial automation, and the regulatory push toward mandatory AI explainability (e.g., the EU AI Act), all of which make XAI transparency an increasingly central, rather than peripheral, construct in digital financial behavior. Because the sample was recruited online and skewed toward heavy neobank and fintech app users, these findings should be understood as applying most directly to similarly engaged Gen Z mobile banking consumers rather than being generalized to the entire U.S. Gen Z population. Full article

Background/Objectives: Accurate fetal head biometry, obtained via ultrasound imaging, is fundamental to prenatal assessment, as it enables fetal growth to be monitored and developmental abnormalities to be detected early. However, conventional manual measurements are time-consuming and operator-dependent, as well as being prone to inter-observer variability. This retrospective study proposes investigating whether conventional segmentation metrics accurately reflect downstream fetal biometric measurements. This would be achieved by evaluating three deep learning architectures on expert-selected, standard-plane fetal head ultrasound images. Methods: Three deep learning architectures, namely, U-Net, Residual U-Net and TransUNet were implemented and evaluated under identical conditions using a local clinical ultrasound dataset of 1918 images and masks from 206 pregnancies. Following segmentation, ellipse fitting was applied to both predicted and reference masks to derive head circumference (HC), biparietal diameter (BPD), and occipitofrontal diameter (OFD). Image-specific physical scaling was applied to convert pixel-based measurements into millimeters. The performance of models was assessed using conventional segmentation metrics, including the Dice Similarity Coefficient (DSC), the Intersection over Union (IoU), and the Hausdorff Distance (HD) metrics. This was complemented by a clinically meaningful biometric evaluation based on the mean absolute error (MAE) of the HC, BPD, and OFD measurements. Results: All three architectures achieved high segmentation performance, with DSC values above 0.98 and IoU values above 0.96. Residual U-Net produced the highest DSC and IoU values (0.9819 ± 0.0102 and 0.9647 ± 0.0196, respectively), whereas TransUNet produced the lowest HD value (1.8935 ± 0.8570 mm). For biometric measurements, the Residual U-Net produced the lowest MAE for HC (1.8877 ± 1.7689 mm) and OFD (0.8672 ± 0.8342 mm), while the TransUNet resulted in the lowest BPD MAE (0.7504 ± 0.7045 mm). Pairwise statistical analysis revealed significant differences in DSC between U-Net and Residual U-Net (p = 0.0172), as well as between Residual U-Net and TransUNet (p = 0.0119). However, no significant differences were observed between U-Net and TransUNet (p = 0.4726). For the biometric measurements, significant differences were only observed for the BPD MAE between the U-Net and TransUNet models (p = 0.0385), while the HC and OFD errors were statistically comparable across the model pairs. Conclusions: This study shows that high overlap-based segmentation performance does not necessarily lead to statistically significant or superior biometric measurement accuracy. The results emphasize the importance of using both conventional segmentation metrics and downstream, measurement-oriented evaluations when comparing deep learning models for fetal head biometry. However, as this study involved a retrospective technical evaluation of curated, expert-selected standard-plane images, the findings could not be interpreted as evidence of clinical effectiveness or immediate clinical applicability. Further validation is required using pregnancy-level separation, multiple observers, independent clinical measurements, pathological and technically challenging examinations, and multicenter external datasets. Full article

The primary cause of the scale effect in scaled models for flood discharge and energy dissipation lies in the dissimilarity of the air dispersion patterns of the ski-jump nappe. To uncover the scale-effect relationship governing the air dispersion patterns of ski-jump energy dissipation nappes in high dams, a series of scaled physical model tests were conducted at the Baihetan Hydropower Station. The air dispersion patterns were systematically observed, as well as the distribution characteristics of entrained air concentration in the ski-jump nappe across various scales. Based on the experimental observations, a two-dimensional stochastic diffusion numerical model was developed, successfully replicating the dispersion process of the nappe as it gradually transformed from a crescent shape to a circular one. Furthermore, by calibrating the concentration distribution curve, a quantitative relationship was established between the random displacement parameter σ and the Weber number. The study revealed that when the Weber number (We) is below 40,000, σ increases rapidly and approximately linearly with We, indicating a high sensitivity to dispersion degree. However, once We surpasses 40,000, the growth rate significantly decelerates, approaching saturation, suggesting that the dispersion degree closely approximates the prototype condition. Consequently, it is suggested that the Weber number control threshold for the physical model of ski-jump water–air two-phase flow in high dams be set above 40,000, providing a valuable reference for selecting large-scale models. Full article

Combining remote sensing data with field inventory data is a common practice in estimation of forest Aboveground Biomass (AGB). However, in areas with well-established ground monitoring systems, the actual added value of this combination has not been fully quantified. This paper aims to critically assess the marginal contribution of optical remote sensing data to AGB estimation when the forest inventory data are already available, and further investigates error sources and residual distributions. Using Longyan City, Fujian Province as a case study, we systematically tested the effects of data types (inventory only, Landsat 8 only, inventory + Landsat 8, and inventory + Landsat 8 + meteorological data), sampling methods, and statistical models on plot-scale AGB estimation accuracy, along with uncertainty distribution across biomass levels. Inventory data alone (stand age and canopy cover) achieved acceptable accuracy, with ${\mathit{R}}^2$ = 0.60 and RMSE = 35.78 t/ha under the optimal configuration (RandomForest + ShuffleSplit_5). Adding Landsat 8 data yielded only modest improvements: RMSE decreased by 5.3% and ${\mathit{R}}^2$ increased by 6.7% relative to the inventory-only baseline. When the optimal combination for each data type was evaluated on the independent test set, the highest ${\mathit{R}}^2$ reached only 0.56, leaving approximately 44% of the observed variation in plot-level AGB unexplained. ANOVA showed that data type was the dominant factor, accounting for $73.0\%$ of the variation in RMSE, followed by model choice ($15.1\%$) and their interaction ($8.5\%$), whereas the contribution of the validation scheme was less than $3.0\%$. We further examined residual distributions across biomass levels and found that residual skewness shifted systematically: negative skewness (overestimation) prevailed in low-biomass plots, near-zero skewness in medium-biomass plots, and positive skewness (underestimation) in high-biomass plots. Among all data types, the Landsat-only configuration exhibited the most severe bias at both low and high biomass extremes; adding remote sensing and climate data to inventory data did not substantially correct this bias structure. Overall, these findings demonstrate that in regions with high-quality ground inventory, the marginal gains from integrating optical remote sensing are limited—both in terms of overall accuracy and the structure of prediction errors. The unexplained variance and systematic residual biases across biomass gradients highlight the need for better representation of high-biomass stands and suggest that future efforts should prioritize sample augmentation in under-represented biomass classes, rather than relying solely on multisource data fusion for accuracy improvement. Full article

The spectral line density of an optical frequency comb (OFC) generated in a microring resonator is fixed by the free spectral range (FSR), and hence by the resonator size: the dense combs required for spectroscopy, optical clocks, and high-capacity communications conventionally demand centimeter-scale cavities, in direct conflict with photonic integration. Here, we propose a route around this FSR–footprint trade-off using topological ring resonators (TRRs) built on a silicon valley photonic crystal (VPC) platform. Evanescently coupling two identical TRRs, an optical analog of quantum tunneling in a double-well potential, deterministically splits each resonance into a doublet of supermodes (Rabi splitting), doubling the spectral line density within a fixed bandwidth while the parallel two-ring layout occupies orders of magnitude less chip area than a single conventional ring of equivalent effective FSR. A coupled-mode-theory model quantitatively captures the splitting observed in full-wave 3D finite-difference time-domain (FDTD) simulations, and the topological protection of the valley edge states preserves the doublet against lattice disorder; a fabrication-tolerance analysis shows the splitting varies by only a few percent for nanometer-scale gap errors. Nonlinear simulations based on the coupled nonlinear Schrödinger equation indicate that the doubled supermode grid translates directly into a denser comb, increasing the generated line count from 48 to 122 under identical Kerr-only pumping conditions. An explicit nonlinear-loss budget, including two-photon and free-carrier absorption, bounds these results for silicon at 1550 nm and identifies mid-infrared silicon and TPA-free platforms such as silicon nitride as physically realistic implementations. This design study establishes coupled topological resonators as a compact, disorder-tolerant architecture for high-density comb generation, which can potentially be experimentally demonstrated. Full article

Point-wise intensification of daily rainfall extremes over India is well established, yet whether their joint spatial behavior is also changing has not been tested. We address this with 75 years (1951–2025) of 0.25° gridded India Meteorological Department daily rainfall, quantifying June–September (JJAS) extremal dependence through the extremal-coefficient function θ(h) and the dependence range ρ of a Brown–Resnick curve fitted to F-madogram estimates on rank-transformed fields; the rank transform strips out marginal trends, so any surviving signal belongs to the joint structure. The all-India range over the full record is 65.6 km (95% CI 62.0–69.5 km). Across 21 sliding 15-year windows, θ at 50 and 100 km fell significantly (p = 0.033 and 0.010 after serial-correlation correction): extremes co-occur more strongly at the storm scale. The signal is regional: Peninsular India’s dependence range grew from 56 km to 76 km between the 1951–1975 and 2001–2025 blocks, a 36% expansion (p = 0.002), and Central India’s window trend is +1.7 km decade⁻¹ (p = 0.017). Independently, the largest connected daily exceedance cluster expanded in area by roughly 40% (p ≤ 0.0006). A Pettitt test dates the step in the footprint series to 1986 (p ≤ 0.0001), whereas the dependence series show progressive trends without a detectable break. Spatial reorganization is compounding marginal intensification: across Peninsular India, joint flood risk is growing faster than cell-wise return levels indicate. Full article

Temperate native grasslands in southeastern Australia have been extensively cleared for agriculture, and the remaining patches are under growing pressure from further land use change, climate variability, and invasive species. Mapping and monitoring their distribution and the cover of native and exotic grasses are critical for their conservation and management. Field-based methods are not always scalable or time-effective, and this study aimed to develop a scalable method to map and monitor the fractional cover-class maps of native C3 and native C4 grass cover as a component of remnant native grasslands on the western outskirts of Melbourne, Victoria, Australia. Field-based reference data for training and validation of random forest machine learning models were collected across multiple sites in 2021. Sentinel-2 optical spectral bands and vegetation indices were used as the primary input data, and Sentinel-1 Synthetic Aperture Radar (SAR)-derived Grey Level Co-occurrence Matrix (GLCM) texture metrics were assessed for their capacity to improve the model. Results show that random forest models trained on Sentinel-2 data without GLCM texture information derived from Sentinel-1 SAR data provided a moderate overall accuracy (C3: 59.1%, C4: 78.1%). Class-specific metrics showed that reliability was highest for better represented lower-cover classes, particularly the 6–25% native C3 class and the 0–5% native C4 class, while higher-cover classes were less reliable because of the limited number of training and validation samples. Grass cover fractions were modelled well for sparse to moderate grass cover, but dense grass cover was not modelled accurately, probably due to limited high-cover samples in the training dataset. Model performance was not improved by the inclusion of Sentinel-1 SAR-derived GLCM texture metrics, indicating that C-band VH-polarised SAR is not sensitive to the fine-scale structural heterogeneity that characterises native grassland ecosystems. Sparse native C3 and C4 grasses could be mapped most reliably in the lower-cover classes as a component of grasslands with optical remote sensing, and the method developed here can now be applied to enable evidence-based management of grasslands, biodiversity conservation and the monitoring of grassland composition in the WGR and elsewhere. Higher-resolution structural datasets and more sophisticated machine learning approaches may be required to accurately predict native C3 and C4 grass cover fractions in denser grasslands. Full article

In a multi-cultural world, problems of living together are often focused on how different faiths accommodate each other. One the basis that we can take lessons from how inter-religious relations were handled in the past, I compare modern Multifaith Spaces (MFS) and the religiously diverse Pu-Jing culture (铺境文化) in Quanzhou, China. I will try to avoid anachronistic application of modern terms to historic situations, and vice versa by treating both as context-specific responses to the same problem of managing religious diversity. I found that they were similar insofar as there was a perceived need for ‘conviviality’, produced through negotiations. In both cases this accommodation was encouraged by top-down legislation and regulation leading to faith activities to be de-privatised. They differed in that there was some syncretic religious art in Pu-Jing culture as well as a lively urban space shared by several faiths. Nothing like this is found in MFS, in which sharing is a matter of time-sharing which allows religions to avoid each other, at a cost of a bland shared small-scale space. I found both cases resulted in a similar spatial solution: standardised spaces are furnished, perhaps temporarily, with elements reflecting the personal beliefs of the users. Full article

Queueing systems are very well studied in the literature. There are a good collection of mathematical tools for studying a variety of queueing related problems. Occasionally, however, researchers doubt the applicability of some of the existing mathematical models to complex and/or large scale queueing systems they encounter in practical situations, such as in communication networks. This has led to the use of simulations and other types of approximation models for dealing with large scale systems. When approximate solutions are presented, purists often state that such solutions do not quite capture all the nuances of the real system. The question is when does a queueing system become large and require approximations, given the abundant literature on exact mathematical methods for analyzing them, and which approximation methods are appropriate to use? Is there a clear approach to readily determine whether a queuing system is large and/or complex? Thus, there is a need for an understanding of how to determine when a system of queues is deemed large and/or complex, and the need for approximation techniques in their analyses. In this paper, we briefly address this issue by providing an overview of existing tools that can be applied in analyzing any queueing problems that are deemed large and/or complex, while also focusing on approximation techniques for systems that are modeled as Markov Chains. In addition, we show that deriving a specific measure for categorizing a queueing system as large and/or complex can be an arduous task to achieve, and we propose the foundations for investigating such measures in future works. Finally, we conclude with open questions and problems with respect to modeling large scale and/or complex queueing systems and their approximations, to promote further research of effective approximation techniques for practitioners and analysts. Full article

Highly efficient tissue culture and transformation systems are indispensable for micropropagation, the molecular studies of secondary metabolism, the biotechnological enhancement of bioactive compounds, and genome editing in plants. Solanum americanum is a diploid species with both vegetable and medicinal value, yet remains largely underutilized. It produces a various pharmacologically active secondary metabolites, including glycoalkaloids, tropane alkaloids, and resveratrol. However, stable genetic transformation and reliable genome editing protocols for this species remain underdeveloped. In this study, we established an efficient genetic transformation and genome editing method for S. americanum. Our results showed that nearly all leaf explants responded positively to MS medium supplemented with 2 mg L⁻¹ 6-benzylamino purine (6-BA) and 0.1 mg L⁻¹ α-naphthaleneacetic acid (NAA), generating an average of 25.85 adventitious shoots per explants within four weeks. Furthermore, among kanamycin-resistant shoots, approximately 89.7% exhibited normal growth and over 75% showed strong GFP expression as well as inheritance, which was confirmed through molecular analysis and fluorescence assays in flowers, fruits, and T1 progeny. Moreover, we successfully utilized this transformation method to achieve stable genome editing of the phytoene desaturase gene (SaPDS). Our results demonstrate that the tissue culture, transformation, and genome editing methods established for S. americanum provide a valuable toolkit for large-scale propagation, the study of secondary metabolism, and the biotechnological improvement of this species. Full article

Among wood types used in industry, ash wood is one of the most commonly heat-treated species. It is widely applied for outdoor applications such as cladding, terrace boards, and garden furniture, as well as in indoor applications that require high dimensional stability. Although the mechanical properties of thermally modified ash have been extensively studied, its creep behaviour under sustained load remains insufficiently described. This study evaluates the short-term creep behaviour of ash wood thermally modified at 210 °C under bending stress, based on deflection measurements at different stress levels and under controlled humidity conditions. Untreated ash wood with comparable moisture content served as a reference. The results show that thermally modified wood exhibited lower creep compliance than untreated wood at all applied stress levels. After 24 h of loading, the absolute reduction in total creep compliance was 3.7% at 40% stress and 20.3% at 60% stress relative to control wood. The difference between modified and unmodified wood became more pronounced as the load level increased. These findings indicate that thermal modification can reduce the tendency of ash wood to creep under short-term loading conditions. Given the limited test duration and the use of small specimens, the results should be interpreted with caution and cannot be directly extrapolated to structural-scale applications. Nevertheless, the study provides new knowledge and is an interesting basis for further investigation of the creep behaviour of thermally modified wood under long-term loading and variable environmental conditions. Full article

Carbon capture and storage (CCS) is a critical technology for mitigating climate change by reducing atmospheric carbon dioxide concentrations. Effective monitoring of CCS sites is essential to ensure that injected CO${}_2$ remains securely trapped and does not leak into the shallow subsurface or atmosphere. Large-scale CCS in the Gulf of Mexico could be facilitated by extensive existing infrastructure and suitable geologic containment; however, legacy wells and structurally complex geology remain critical challenges for ensuring storage integrity. Traditional monitoring methods, while effective, often lack the temporal resolution and cost-effectiveness needed for comprehensive leak detection, particularly in shallow seafloor environments. This study explores the potential of distributed acoustic sensing (DAS) as a novel monitoring solution for near-surface CO${}_2$ leakage at geologic carbon storage (GCS) sites. We conducted a laboratory-scale controlled nitrogen gas bubble injection experiment to compare DAS responses at varying cable burial depths and evaluated these signals using simultaneous hydrophone measurements. Results indicate that while DAS effectively detects acoustic signals from bubbles in both sediment and water columns, its response amplitude diminishes with increased burial depth. These findings suggest that DAS could serve as a promising technology for long-term monitoring of marine GCS sites, providing insights into near-surface gas flow and leak detection. Full article

Antibacterial coatings are essential for preventing infections and maintaining adequate hygiene standards in high-density and high-risk environments, such as public spaces, public transportation systems, schools, and healthcare facilities. In this context, it is essential that such coatings are easy to apply and compatible with large-scale production processes. Among the various strategies for developing antimicrobial surfaces, the incorporation of antibacterial agents into paints represents a particularly practical and versatile approach. In this work, antimicrobial coatings were developed by incorporating curcumin, a naturally derived antibacterial compound, into a commercially available water-based paint. For comparison, coatings were also developed by incorporating a commercially available silver-ion-based additive, a well-established antibacterial agent, into the same paint. Antimicrobial dispersions were deposited onto polycarbonate (PC) and polymethyl methacrylate (PMMA) substrates using a spray coating technique. The produced coatings were evaluated in terms of adhesion and hardness according to the ASTM D3359 and ASTM D3363 standards, respectively, achieving ratings of 4B for adhesion and 4B/5B for hardness. Moreover, after five days of continuous water immersion, no visible signs of cracking, delamination, or discoloration were observed. The resulting curcumin- and silver-based coatings, under the tested surface-contact assay conditions, markedly reduced viable bacterial recovery against both S. aureus and E. coli, with no colonies recovered at the dilution range used for comparison with the Paint-Blank control. These results support the potential of curcumin as a naturally derived antibacterial additive for the development of water-based antibacterial coatings and provide a basis for further investigation of their long-term stability, durability, and practical applicability. Full article

Background: Valvular heart disease frequently accompanies acute heart failure (AHF), yet most studies address single valve lesions, and the prognostic weight of concurrent multivalvular involvement is less well defined. We examined the prevalence of single- and multi-valve disease and its relationship with one-year mortality. Methods: We analyzed 530 consecutive patients enrolled in a prospective single-center AHF registry between February 2023 and June 2025 and followed for 12 months. Aortic, mitral and tricuspid disease was graded according to European Society of Cardiology/European Association of Cardiovascular Imaging (ESC/EACVI) criteria, a valve being considered involved when it carried at least moderate stenosis and/or regurgitation; multivalvular disease was defined as involvement of two or more valves. The primary endpoint was one-year all-cause mortality. Cox regression was adjusted for admission log N-terminal pro-B-type natriuretic peptide (NT-proBNP), frailty (Clinical Frailty Scale ≥ 5) and age. In-hospital acute kidney injury was examined descriptively but was not entered into the models, because it is ascertained after admission. Results: Overall, 71.3% of patients had at least one significantly diseased valve—mitral in 44.3%, tricuspid in 43.8% and aortic in 27.5%—and 34.9% had multivalvular involvement. One-year mortality rose stepwise with the number of valves involved: 19.7%, 24.9%, 36.3% and 54.0% for zero, one, two and three valves, respectively (log-rank p < 0.001), corresponding to a crude hazard ratio (HR) of 1.51 (95% confidence interval [CI] 1.28–1.78) per additional valve. Multivalvular versus single or no valve involvement carried a crude HR of 2.07 (1.51–2.84) and remained associated with mortality after adjustment (HR 1.48, 95% CI 1.06–2.07, p = 0.020), as did each additional valve (HR 1.22, 1.03–1.45, p = 0.023). Adjustment for clinically manifest right heart failure (HF) attenuated the association with mortality (HR 1.15, 95% CI 0.96–1.37), whereas the association with the triple composite remained significant (HR 1.19, 95% CI 1.03–1.36). Propensity-score adjustment excluding right HF retained the association with mortality (HR 1.43, 95% CI 1.02–2.00), whereas inclusion of right HF attenuated it (HR 1.33, 0.94–1.87). The association was not detectable in the 304 patients without clinically manifest right HF (HR 1.06, 0.80–1.41, p = 0.69). In a post-discharge sensitivity analysis, the adjusted associations were directionally similar but did not reach statistical significance. Conclusions: In hospitalized AHF, multivalvular involvement was common, followed a predominantly mitral–tricuspid pattern, and identified an older, frailer, more congested phenotype with substantial right-sided and chronic renal involvement. Valve burden showed a graded association with one-year mortality that persisted after adjustment for baseline prognostic characteristics, but was attenuated once right-sided involvement was taken into account, whether by direct adjustment or by propensity methods; the association was retained when right HF was omitted from the propensity model. Because right-sided variables lie downstream of significant tricuspid disease, this attenuation is what adjustment for an intermediate is expected to produce and does not establish absence of prognostic value. Cumulative valve burden is therefore best interpreted as a powerful phenotypic marker of a more advanced HF phenotype with prominent right-heart involvement, rather than as an independent causal determinant of mortality. Full article