Search Results (476)

Left ventricular non-compaction cardiomyopathy (LVNC) is an uncommon myocardial phenotype characterized by prominent trabeculae and deep blood-filled recesses. The expanding use of cardiac magnetic resonance (CMR) has increased detection, yet uncertainty persists about whether LVNC is a distinct disease or a phenotype that overlaps with other cardiomyopathies. LVNC expression reflects the interplay among genotype, sex, ancestry, and hemodynamic load and thus serves as a model for precision cardiology. We conducted a narrative review of literature published between January 2000 and April 2025 in major databases. We included clinical studies with at least 10 patients, meta-analyses, reviews, and consensus statements addressing pathophysiology, genetics, diagnosis, prognosis, and treatment. Sarcomeric variants account for a substantial fraction of cases and connect LVNC with dilated and hypertrophic cardiomyopathies. Echocardiographic and CMR criteria identify the phenotype but blur the boundary between physiological and pathological hypertrabeculation. Fibrosis on late gadolinium enhancement and systolic dysfunction are consistently associated with worse outcomes. Current management largely adapts heart-failure strategies, including neurohormonal blockade, SGLT2 inhibitors, and implantable cardioverter-defibrillators in selected high-risk patients. Optimal care integrates clinical, imaging, and genetic information. The lack of universal diagnostic criteria highlights the need for prospective studies and consensus to standardize diagnosis and treatment. Future algorithms that combine multi-omics, quantitative imaging, and AI-based risk prediction could individualize surveillance, pharmacotherapy, and device therapy. Full article

Smaller data environments with expert systems are generally driven by the need for interpretable reasoning frameworks, such as fuzzy rule-based systems (FRBS), which cannot often quantify epistemic uncertainty during decision-making. This study proposes a novel Layered Entropy Model (LEM) comprising three semantic layers: Membership Function Entropy (MFE), Rule Activation Entropy (RAE), and System Output Entropy (SOE). Shannon entropy is applied at each layer to enable granular diagnostic transparency throughout the inference process. The approach was evaluated using both synthetic simulations and a real-world case study on the PIMA Indian Diabetes dataset. In the real data experiment, the system produced sharp, fully confident decisions with zero entropy at all layers, yielding an Epistemic Confidence Index (ECI) of 1.0. The proposed framework maintains full compatibility with conventional Type-1 FRBS design while introducing a computationally efficient and fully interpretable uncertainty quantification capability. The results demonstrate that LEM can serve as a powerful tool for validating expert knowledge, auditing system transparency, and deployment in high-stakes, small-data decision domains, such as healthcare, safety, and finance. The model contributes directly to the goals of explainable artificial intelligence (XAI) by embedding uncertainty traceability within the reasoning process itself. Full article

Background: Electrochemical impedance spectroscopy (EIS) is indispensable for disentangling charge-transfer, capacitive, and diffusive phenomena, yet reproducible parameter estimation and objective model selection remain unsettled. Methods: We derive closed-form impedances and analytical Jacobians for seven equivalent-circuit models (Randles, constant-phase element (CPE), and Warburg impedance (ZW) variants), enforce physical bounds, and fit synthetic spectra with 2.5% and 5.0% Gaussian noise using hybrid optimization (Differential Evolution (DE) → Levenberg–Marquardt (LM)). Uncertainty is quantified via non-parametric bootstrap; parsimony is assessed with root-mean-square error (RMSE), Akaike Information Criterion (AIC), and Bayesian Information Criterion (BIC); physical consistency is checked by Kramers–Kronig (KK) diagnostics. Results: Solution resistance ($R_s$) and charge-transfer resistance ($R_{ct}$) are consistently identifiable across noise levels. CPE parameters $(Q,n)$ and diffusion amplitude $\left(\sigma \right)$ exhibit expected collinearity unless the frequency window excites both processes. Randles suffices for ideal interfaces; Randles+CPE lowers AIC when non-ideality and/or higher noise dominate; adding Warburg reproduces the ${45}^{\circ }$ tail and improves likelihood when diffusion is present. The $(R_{ct}+Z_W)\Vert \mathrm{CPE}$ architecture offers the best trade-off when heterogeneity and diffusion coexist. Conclusions: The framework unifies analytical derivations, hybrid optimization, and rigorous statistics to deliver traceable, reproducible EIS analysis and clear applicability domains, reducing subjective model choice. All code, data, and settings are released to enable exact reproduction. Full article

Structural health monitoring (SHM) plays a critical role in ensuring the safety and performance of aerospace structures throughout their lifecycle. As aircraft and spacecraft systems grow in complexity, the integration of machine learning (ML) into SHM frameworks is revolutionizing how damage is detected, localized, and predicted. This review presents a comprehensive examination of recent advances in ML-based SHM methods tailored to aerospace applications. It covers supervised, unsupervised, deep, and hybrid learning techniques, highlighting their capabilities in processing high-dimensional sensor data, managing uncertainty, and enabling real-time diagnostics. Particular focus is given to the challenges of data scarcity, operational variability, and interpretability in safety-critical environments. The review also explores emerging directions such as digital twins, transfer learning, and federated learning. By mapping current strengths and limitations, this paper provides a roadmap for future research and outlines the key enablers needed to bring ML-based SHM from laboratory development to widespread aerospace deployment. Full article

Cardiorenal syndrome (CRS) reflects the intricate and bidirectional interplay between cardiac and renal dysfunction, commonly resulting in diagnostic uncertainty, therapeutic dilemmas and poor outcomes. While traditional biomarkers like serum creatinine (Cr) and natriuretic peptides remain widely used, their limitations in specificity, timing and contextual interpretation often hinder optimal management. This narrative review synthesizes the current evidence on established and emerging biomarkers in CRS, with emphasis on their clinical relevance, integration into real-world practice, and potential to inform precision therapy. Markers of glomerular filtration rate beyond creatinine—such as cystatin C—offer more accurate assessment in frail or sarcopenic patients, while tubular injury markers such as NGAL, KIM-1, and urinary L-FABP (uL-FABP) provide early signals of structural renal damage. The FDA-approved NephroCheck^® test—based on TIMP-2 and IGFBP7— enables risk stratification for imminent AKI up to 24 h before functional decline. Congestion-related markers such as CA125 and bio-adrenomedullin outperform natriuretic peptides in certain CRS phenotypes, particularly in right-sided heart failure or renally impaired patients. Fibrosis and inflammation markers (galectin-3, sST2, GDF-15) add prognostic insights, especially when combined with NT-proBNP or troponin. Rather than presenting biomarkers in isolation, this review proposes a framework that links them to specific clinical contexts—such as suspected decongestion-related renal worsening or persistent congestion despite therapy—to support actionable interpretation. A tailored, scenario-based, multi-marker strategy may enhance diagnostic precision and treatment safety in CRS. Future research should prioritize prospective biomarker-guided trials and standardized pathways for clinical integration. Full article

Background/Objectives: Atypical cartilaginous tumours (ACTs) are intermediate, locally aggressive chondroid tumours in the appendicular skeleton. Due to the potential for transformation into high-grade chondrosarcomas, management typically consists of regular MRI follow-up and, occasionally, surgery. We primarily aimed to examine the rate of malignant transformation in ACTs in our hospital; secondarily, we aimed to identify the factors influencing management choices and outcomes. Methods: All patients referred between 2013 and 2020 with a long-bone ACT were identified from the unit database. For this retrospective study, we analysed the imaging, management, and outcomes for the patients discussed at our musculoskeletal radiological conference. Results: A total of 59 patients were included; of these, 0 cases of malignant transformation were observed with a mean follow-up time of 8.4 years. Of the presenting cases, the musculoskeletal radiological conference advised that 6 should be biopsied, 40 should receive MRI follow-up, 7 should receive X-ray follow-up, and 6 should be re-examined in clinic. Subsequently, 12 patients underwent surgery due to continued pain, diagnostic uncertainty, and historical practices. Of these, seven experienced continued post-operative pain. Conclusions: None of the encountered ACTs underwent malignant transformation, supporting previous findings that this transformation is a rare phenomenon. Furthermore, of the small sample of patients undergoing surgery, less than half were left pain-free. These findings support a more conservative approach to ACT management, with the potential to discharge after an initial review. Full article

Periprosthetic joint infection (PJI) remains a major complication in orthopedic surgery, with accurate and timely diagnosis being essential for optimal patient management. Traditional culture-based diagnostics are often limited by suboptimal sensitivity, especially in biofilm-associated and low-virulence infections. In recent years, non-culture-based methodologies have gained prominence. Molecular techniques, such as polymerase chain reaction (PCR) and next-generation sequencing (NGS), offer enhanced detection of microbial DNA, even in culture-negative cases, and enable precise pathogen identification. In parallel, extensive research has focused on biomarkers, including systemic (e.g., C-reactive protein, fibrinogen, D-dimer), synovial (e.g., alpha-defensin, calprotectin, interleukins), and pathogen-derived markers (e.g., D-lactate), the latter reflecting metabolic products secreted by microorganisms during infection. The development of multiplex platforms now allows for the simultaneous measurement of multiple synovial biomarkers, improving diagnostic accuracy and turnaround time. Furthermore, the integration of artificial intelligence (AI) and machine learning algorithms into diagnostic workflows has opened new avenues for combining clinical, molecular, and biochemical data. These models can generate probability scores for PJI diagnosis with high accuracy, supporting clinical decision-making. While these technologies are still being validated for routine use, their convergence marks a significant step toward precision diagnostics in PJI, potentially improving early detection, reducing diagnostic uncertainty, and guiding targeted therapy. Full article

The fault diagnosis of complex systems is essential for ensuring operational safety. The belief rule base (BRB), a rule-driven framework based on expert knowledge, is widely applied in fault diagnosis because of its ability to manage uncertainty. However, existing BRB models rely heavily on large amounts of high-quality labeled data, and their performance decreases when labels are scarce or noisy. To address this limitation, a perturbed self-training-based BRB method (PS-BRB) is proposed. In this approach, pseudo-labels for unlabeled samples are first inferred by an initial BRB, and Gaussian noise is introduced into the inputs to simulate perturbations. Samples that produce consistent predictions before and after perturbation are retained through class consistency checking. The Jensen–Shannon (JS) divergence then measures the difference between belief distributions, and high-quality pseudo-labels are selected according to the 90th percentile criterion. These pseudo-labels are incorporated into the training set to optimize BRB rules and parameters. The method is validated on two bearing datasets, and the results show improved diagnostic accuracy and applicability, which indicates potential for use in practical engineering scenarios. Full article

Prognostic uncertainty and missed diagnoses of sepsis remain frequent after cardiopulmonary bypass (CPB) surgery, where systemic inflammatory response (SIRS) arises from surgical trauma, blood activation in the extracorporeal circuit, ischemia/reperfusion injury, and endotoxin release. Among innovative biomarkers, pro-adrenomedullin (pro-ADM), particularly its stable fragment mid-regional pro-adrenomedullin (MR-proADM), has shown promise for detecting endothelial dysfunction and predicting organ failure in sepsis. SeptiCyte^® RAPID (Seattle, WA, USA) also represents a novel diagnostic tool that assesses the host immune response by quantifying PLA2G7 and PLAC8 gene expression in whole blood, offering potential for early differentiation between sepsis and sterile inflammation. We analyzed traditional and innovative biomarkers within 24 h post-CPB in a pilot group of patients admitted to the cardiac Intensive Care Unit of the “Città della Salute e della Scienza” University Hospital (Turin, Italy) between June and November 2023. Data from the following 14 patients were collected: 7 undergoing surgery for infective endocarditis (IE, Group 1) and 7 having standard elective cardiac surgery (Group 2). Procalcitonin (PCT), lactate, and pro-ADM increased in Group 1 but not in Group 2. SeptiCyte^® RAPID showed a moderate, borderline increase in Group 1. The innovative biomarkers had a good performance in patients exhibiting signs of organ dysfunction and in subjects demonstrating at least cardiovascular and/or pulmonary damage and under vasopressor and inotropic support. Although limited by the small sample, our preliminary data suggest no biomarker alterations in patients with standard elective cardiac surgery, unlike in those with IE. Full article

Electrochemical impedance spectroscopy (EIS) is widely used at the laboratory level for monitoring/diagnostics of battery cells, but the design and validation of in situ, online measurement systems based on EIS face challenges due to complex hardware–software interactions and non-idealities. This study aims to develop an integrated co-simulation framework to support the design, debugging, and validation of EIS measurement systems devoted to the online monitoring of battery cells, helping to predict experimental results and identify/correct the non-ideality effects and sources of uncertainty. The proposed framework models both the hardware and software components of an EIS-based system to simulate and analyze the impedance measurement process as a whole. It takes into consideration the effects of physical non-idealities on the hardware–software interactions and how those affect the final impedance estimate, offering a tool to refine designs and interpret test results. For validation purposes, the proposed general framework is applied to a specific EIS-based laboratory prototype, previously designed by the research group. The framework is first used to debug the prototype by uncovering hidden non-idealities, thus refining the measurement system, and then employed as a digital model of the latter for fast development of software algorithms. Finally, the results of the co-simulation framework are compared against a theoretical model, the real prototype, and a benchtop instrument to assess the global accuracy of the framework. Full article

This study examines the dynamic relationship between the gold markets of India and the United States from 2005 to 2025. Recognising gold’s role as a hedge and safe-haven during market uncertainty, we employ the Autoregressive Distributed Lag (ARDL) model to assess long-term co-integration and apply the Toda–Yamamoto causality test to evaluate directional influences. Additionally, the Generalised Autoregressive Conditional Heteroskedasticity (GARCH) (1, 1) model is applied to examine volatility spillovers. Results reveal no long-term co-integration between the two markets, suggesting they function independently over time. However, unidirectional causality is observed from the U.S. to the Indian gold market, and the GARCH model confirms bidirectional volatility transmission, indicating interconnected short-run dynamics. These findings imply that gold market shocks in one country may affect short-term pricing in the other, but not long-term trends. From a portfolio diversification and risk management perspective, investors may benefit from allocating assets across both markets. This study contributes a novel empirical framework by integrating ARDL, Toda–Yamamoto Granger causality, and GARCH(1, 1) models over a two-decade period (2005–2025), incorporating post-COVID market dynamics. The combination of these methods, applied to both an emerging (India) and developed (U.S.) economy, provides a comprehensive understanding of gold market interdependence. In doing this, the paper offers valuable insights into causality, volatility transmission, and diversification potential. The econometric rigour of the study is enhanced through residual diagnostic tests, including tests of normality, autocorrelation, and other heteroscedasticity tests, as well as VAR stability tests. These ensure strong inference and model validity; more specifically, they are pertinent to the analysis of financial time series. Full article

A fundamental limitation of maximum likelihood and Bayesian methods under model misspecification is that the asymptotic covariance matrix of the pseudo-true parameter vector ${\mathsf{\theta }}_{*}$ is not the inverse of the Fisher information, but rather the sandwich covariance matrix ${\displaystyle \frac{1}{n}}{\mathbf{A}}_{*}^{-1}{\mathbf{B}}_{*}^{\phantom{-1}}{\mathbf{A}}_{*}^{-1}$, where ${\mathbf{A}}_{*}$ and ${\mathbf{B}}_{*}$ are the sensitivity and variability matrices, respectively, evaluated at ${\mathsf{\theta }}_{*}$ for training data record ${\omega }_1,\dots ,{\omega }_n$. This paper makes three contributions. First, we review existing approaches to robust posterior sampling, including the open-faced sandwich adjustment and magnitude- and curvature-adjusted Markov chain Monte Carlo (MCMC) simulation. Second, we introduce a new sandwich-adjusted MCMC method. Unlike existing approaches that rely on arbitrary matrix square roots, eigendecompositions or a single scaling factor applied uniformly across the parameter space, our method employs a parameter-dependent learning rate $\lambda \left(\mathsf{\theta }\right)$ that enables direction-specific tempering of the likelihood. This allows the sampler to capture directional asymmetries in the sandwich distribution, particularly under model misspecification or in small-sample regimes, and yields credible regions that remain valid when standard Bayesian inference underestimates uncertainty. Third, we propose information-theoretic diagnostics for quantifying model misspecification, including a strictly proper divergence score and scalar summaries based on the Frobenius norm, Earth mover’s distance, and the Herfindahl index. These principled diagnostics complement residual-based metrics for model evaluation by directly assessing the degree of misalignment between the sensitivity and variability matrices, ${\mathbf{A}}_{*}$ and ${\mathbf{B}}_{*}$. Applications to two parametric distributions and a rainfall-runoff case study with the Xinanjiang watershed model show that conventional Bayesian methods systematically underestimate uncertainty, while the proposed method yields asymptotically valid and robust uncertainty estimates. Together, these findings advocate for sandwich-based adjustments in Bayesian practice and workflows. Full article

Background: Distinguishing between bacterial and viral infections in children remains a significant challenge for clinicians. Traditional biomarkers have limited utility, often leading to antibiotic overprescription due to clinician uncertainty. With rising antimicrobial resistance, novel biomarkers are needed to improve diagnosis. This scoping review examines current host miRNA biomarkers for acute bacterial and viral infections in children (0–18), focusing on study methods, diagnostic metrics, and research gaps to support clinical translation. Results: Of the 1147 articles identified, 36 studies were included. Notably, 72.2% of the studies originated from Asia, and the distribution across the paediatric age groups was relatively even. A total of 17 miRNAs were validated in at least two independent studies. Three miRNAs, hsa-miR-182-5p, hsa-miR-363-3p, and hsa-miR-206, were consistently associated with bacterial infection in children. Meanwhile, nine miRNAs were associated with viral infections: hsa-miR-155, hsa-miR-29a-3p, hsa-miR-155-5p, hsa-miR-150-5p, hsa-miR-140-3p, hsa-miR-142-3p, hsa-miR-149-3p, hsa-miR-210-3p, and hsa-miR-34a-5p. Across the 12 studies reporting diagnostic accuracy metrics, miRNA biomarkers exhibited a sensitivity ranging from 70% to 100%, and a specificity ranging from 72% to 100%. The area under the curve across the studies demonstrated a range from 0.62 to 0.99. Conclusions: This scoping review highlights the potential of miRNA targets for diagnosing paediatric infections when studied rigorously. However, clinical translation is limited by poor adherence to STARD guidelines, lack of robust diagnostic metrics, and study heterogeneity. Many studies were set up with a case–control design, a design that, while highlighting differences, is more likely to identify non-specific biomarkers rather than those that are useful for novel clinical diagnostics. Full article

We propose a general and experimentally accessible framework to quantify transition timing in discrete quantum systems via the time-of-flow (TF) distribution. Defined from the rate of population change in a target state, the TF distribution can be reconstructed through repeated projective measurements at discrete times on independently prepared systems, thus avoiding Zeno inhibition. In monotonic regimes, it admits a clear interpretation as a time-of-arrival (TOA) or time-of-departure (TOD) distribution. We apply this approach to optimize time-dependent Hamiltonians, analyze shortcut-to-adiabaticity (STA) protocols, study non-adiabatic features in the dynamics of a three-level time-dependent detuning model, and derive a transition-based quantum speed limit (TF-QSL) for both closed and open quantum systems. We also establish a lower bound on temporal uncertainty and examine decoherence effects, demonstrating the versatility of the TF framework for quantum control and diagnostics. This method provides both a conceptual tool and an experimental protocol for probing and engineering quantum dynamics in discrete-state platforms. Full article

Early and accurate detection of pneumonia from chest X-ray images is essential for improving treatment and clinical outcomes. Medical imaging datasets often exhibit class imbalance and uncertainty in feature extraction, which complicates conventional classification methods and motivates the use of advanced approaches combining deep learning and fuzzy logic. This study proposes a hybrid approach that combines deep learning architectures (VGG16, EfficientNetV2, MobileNetV2, ResNet50) for feature extraction with fuzzy logic-based classifiers, including Fuzzy C-Means, Fuzzy Decision Tree, Fuzzy KNN, Fuzzy SVM, and ANFIS (Adaptive Neuro-Fuzzy Inference System). Feature selection techniques were also applied to enhance the discriminative power of the extracted features. The best-performing model, ANFIS with MobileNetV2 features and Gaussian membership functions, achieved an overall accuracy of 98.52%, with Normal class precision of 97.07%, recall of 97.48%, and F1-score of 97.27%, and Pneumonia class precision of 99.06%, recall of 98.91%, and F1-score of 98.99%. Among the fuzzy classifiers, Fuzzy SVM and Fuzzy KNN also showed strong performance with accuracy above 96%, while Fuzzy Decision Tree and Fuzzy C-Means achieved moderate results. These findings demonstrate that integrating deep feature extraction with neuro-fuzzy reasoning significantly improves diagnostic accuracy and robustness, providing a reliable tool for clinical decision support. Future research will focus on optimizing model efficiency, interpretability, and real-time applicability. Full article