Search Results (64)

The Riemann Hypothesis (RH) asserts that all non-trivial zeros of the Riemann zeta function lie on the critical line Re(s) = 0.5, yet no general proof exists despite extensive numerical verification. This study introduces a machine learning–based framework that combines classification, explainability, contradiction testing, and generative modeling to provide empirical evidence consistent with RH. First, discriminative models augmented with SHAP analysis reveal that the real and imaginary parts of ζ(s) contribute stable explanatory signals exclusively along the critical line, while off-line regions exhibit negligible attributions. Second, a contradiction-test framework, constructed from systematically sampled off-line points, shows no indication of spurious zero-like behavior. Finally, a mixture-density variational autoencoder (MDN-VAE) trained on 10,000 zero spacings produces synthetic distributions that closely match the empirical spacing law, with a Kolmogorov–Smirnov test (KS = 0.041, p = 0.075) confirming statistical indistinguishability. Together, these findings demonstrate that machine learning and explainable AI not only reproduce the known statistical properties of zeta zeros but also reinforce the absence of contradictions to RH under extended empirical exploration. While this framework does not constitute a formal proof, it offers a falsifiability-oriented, data-driven methodology for exploring deep mathematical conjectures. This empirical evaluation involved a baseline of 12,005 points, including 5 known zeros, and 15 off-line test points. The Random Forest classifier achieved high accuracy in distinguishing critical-line zeros, and consistently rejected off-line points. The generative models further corroborated these findings. Full article

A reliable prediction of the remaining useful life of critical electronic components, such as insulated gate bipolar transistors, is necessary for preventing failures in many industrial applications. Recently, diverse machine-learning techniques have been used for this task. However, they are generally focused on capturing the temporal dependencies or on representing the probabilistic nature of the degradation of the device. This work proposes a neural network architecture that combines long short-term memory and mixture density networks to address both targets simultaneously when modeling the remaining useful life. The proposed model was trained and evaluated using a real dataset of insulated gate bipolar transistors, demonstrating a high capacity for predicting the remaining useful life of the validation devices. The proposed model outperformed the other algorithms considered in the study in terms of root mean squared error and coefficient of determination. In general terms, an average reduction of at least 18% of the root mean squared error was obtained when compared with the second-best model among those considered in this work, but in some specific cases, the root mean squared error during the prediction of remaining useful life decreased up to 21%. In addition to the high performance obtained, the characteristics of the network output also facilitated the creation of confidence intervals, which are more informative than solely exact values for decision-making. Full article

Background and Objectives: Resilience and perceived social support are crucial factors influencing the psychological well-being among breast cancer survivors. Understanding their levels and interrelations can inform psychosocial interventions aimed at improving survivorship outcomes. This study aimed to examine the relationship between resilience and perceived social support, evaluate the psychometric properties, and explore their associations with key sociodemographic factors among breast cancer survivors. Materials and Methods: A total of 253 women in clinical remission, at least six months post-primary treatment, were recruited from the University Clinical Center of Vojvodina. Participants completed sociodemographic and clinical questionnaires, the Connor–Davidson Resilience Scale (CD-RISC-25), and the Multidimensional Scale of Perceived Social Support (MSPSS). Results: Participants reported moderate levels of resilience (Mdn = 75, IQR = 19). Among resilience domains, Hardiness (Mdn = 22, IQR = 7) and Coping (Mdn = 14, IQR = 4) scored highest, while Optimism (Mdn = 6, IQR = 3) was the lowest. A significant positive correlation was found between resilience and fertility-related quality of life (ρ = 0.454, p < 0.01), while a negative correlation was observed between resilience and fertility-related stress (ρ = −0.275, p < 0.01). Adaptive coping strategies, particularly from the Practical Management Branch of the CIQ, were positively associated with resilience and quality of life, while avoidance coping was linked to higher stress and lower well-being. Conclusions: Breast cancer survivors in this Serbian cohort reported moderate resilience and social support, with a strong interrelationship between the two. These findings underscore the importance of strengthening social support networks as a potential pathway to enhance resilience and psychological well-being in cancer survivorship care. Full article

Background: Transradial access may affect health-related quality of life (QoL) in cerebral diagnostic angiography. However, its assessment is methodologically challenging, as repeated measurements can be influenced by response shift. To mitigate this bias, a retrospective cross-sectional study was designed using a then-test approach, allowing patients to reflect on their post procedural status at a single time point. Methods: Quality of life was assessed using the 12-Item Short Form Health Survey (SF-12). A then-test approach was also employed, whereby patients were asked to retrospectively indicate whether they perceived their condition as worse following the procedure. The survey yielded Physical (PCS) and Mental Component Summary (MCS) scores, standardized to a mean of 50 (range of 0–100), with lower values indicating greater health-related limitations. Group differences were analyzed using the Mann–Whitney U test. Associations between PCS and MCS, respectively, and clinical variables were assessed using multiple linear regression models. Results: Forty patients underwent diagnostic cerebral angiography over a 15-month observation period. Applying a then-test design, Group A included the 12.5% (n = 5) of patients who reported feeling worse post-procedure while Group B comprised the remaining 87.5% (n = 35). QoL scores were significantly lower in Group A (Mdn = 28.6) compared to B (Mdn = 46.7) for both PCS scores (p = 0.007) and MCS scores (45.3 vs. 54.6, p = 0.018). In the multiple linear regression analysis, no statistically significant associations were found between the PCS or MCS scores and any clinical variable, including age, sex, body mass index (BMI), procedure duration, dose area product, access site, prior neurosurgical history, and fluoroscopy time (p > 0.05). Conclusions: Transradial access for diagnostic cerebral angiography may affect QoL, as assessed using the SF-12 questionnaire. Applying the then-test approach, the group of patients who reported feeling worse after the procedure (12.5%) showed significantly lower physical and mental health scores. These findings underscore the need for prospective studies to further investigate patient-reported outcomes. Full article

Objectives: This study aimed to investigate the impact of platelet-rich plasma (PRP) and platelet-rich gel (PRG) on tympanic membrane closure rates, hearing improvement, and quality of life following tympanoplasty. Methods: Seventy-two patients with chronic tympanic membrane perforations were enrolled in a double-blinded, randomized controlled trial at a single tertiary referral center. All patients underwent tympanoplasty using a temporalis fascia graft and were randomly assigned to one of two groups: one group received standard tympanoplasty alone, while the other received intraoperative application of autologous PRP and PRG, in addition to the standard procedure. Results: The PRP group demonstrated a significantly higher rate of complete tympanic membrane closure compared to the control group (32/36; 88.9% vs. 24/36; 66.7%; p < 0.05). Bone conduction hearing remained unchanged in both groups, while air conduction hearing improved significantly from pre- to post-treatment in each group. However, the difference in air conduction improvement between the PRP group and the control group was not statistically significant (PRP group: Mdn = −8.25; control group: Mdn = −12.20; U = 618; z = −0.54; p = 0.30). Quality of life improved in both the PRP and control groups; however, the difference between the groups was not statistically significant (PRP group: 10.44 ± 10.46; control group: 10.47 ± 8.22; 95% CI [−4.45; 4.40]; t(66) = −0.01; p = 0.16). Conclusions: Our findings suggest that intraoperative application of autologous PRP and PRG may improve tympanoplasty outcomes, particularly in cases with lower expected success rates or when performing minimally invasive transcanal procedures under local anesthesia. However, variability in PRP preparation, application methods, and graft materials across studies limits direct comparisons. Standardized protocols and further controlled studies are necessary to clarify PRP’s clinical value in tympanoplasty. Full article

Background/Objectives: Cariprazine, a D3/D2 partial agonist, is one of the few recommended treatment options for bipolar 1 disorder (BP1D) in Southeast Asia. This study aims to generate insights from leading experts on the safe and effective use of cariprazine for BP1D, specifically by formulating practical recommendations not thoroughly covered in the existing literature. Methods: A formal consensus methodology using the modified RAND/UCLA Appropriateness Method was employed to develop consensus recommendations. The methodology included a targeted literature search, creation of clinical scenarios, two rounds of rating of the appropriateness of each scenario on a nine-point Likert scale by an expert panel of psychiatrists from Southeast Asia (n = 13), and a face-to-face discussion among the expert panel between the two rounds of rating. In the absence of disagreement, scenarios were classified as appropriate (7–9), equivocal (4–6), or inappropriate (1–3) based on median scores. Clinical scenarios were subsequently converted to consensus recommendations upon approval by the expert panel. Results: Most experts recommended a 4–8-week trial of cariprazine for bipolar depression (85%) and 3–4 weeks for acute mania/mixed (71%). For longer treatment, 61.5% and 69% recommended >1 year for acute mania/mixed and bipolar depression, respectively. Cariprazine was also considered suitable as first-line therapy, including for first-episode bipolar depression (Mdn: 8, IQR: 7–9) and first-episode mania (Mdn: 8; IQR: 8–9). Conclusions: The consensus recommendations may serve as practical guidance for clinicians to make informed decisions regarding the management of adult patients with BP1D, while considering the preferences and circumstances of individual patients. Full article

The reliable and uncertainty-aware prediction of surrounding vehicles remains a key challenge in autonomous driving. However, existing methods often struggle to quantify and incorporate uncertainty effectively. To address these challenges, we propose a hybrid architecture that combines a data-driven neural trajectory predictor with physically grounded constraints to forecast future vehicle occupancy. Specifically, the physical constraints are derived from vehicle kinematic principles and embedded into the network as additional loss terms during training. This integration ensures that predicted trajectories conform to feasible and physically realistic motion boundaries. Furthermore, a mixture density network (MDN) is employed to estimate predictive uncertainty, transforming deterministic trajectory predictions into spatial probability distributions. This enables a probabilistic occupancy representation, offering a richer and more informative description of the potential future positions of surrounding vehicles. The proposed model is trained and evaluated on the Aerial Dataset for China’s Congested Highways and Expressways (AD4CHE), which contains representative driving scenarios in China. Experimental results demonstrate that the model achieves strong fitting performance while maintaining high physical plausibility in its predictions. Full article

Prolonged untreated psychosis worsens outcomes, thus motivating the study of pathways and perceived barriers to care, especially for high-risk age groups like college students. The primary objective of this study was to explore pathways to coordinated specialty care (CSC) and perceived barriers to care in college students at high risk for psychosis or with first-episode psychosis and determine any association between them. Twenty-four college students enrolled in CSC completed the Circumstances of Onset and Relapse Schedule and Barriers to Seeking Psychological Help Scale (BSPHS). Non-parametric tests were used for two-group analyses, and medians and interquartile ranges (IQR) were calculated. The median number of total contacts along the pathway to CSC was 5.0 (IQR = 5.2), with more psychiatric contacts (Mdn. = 3.0, IQR = 2.2) than non-psychiatric contacts (Mdn. = 2.0, IQR = 3.0). Students whose first contact was with non-psychiatric services had longer pathways to care overall, with a higher median number of total psychiatric as well as non-psychiatric contacts relative to students whose first contact was with psychiatric services. With the highest possible total BSPHS score being 5, the median score was 2.7 (IQR = 0.8). Targeted psychosis literacy training for non-mental health professionals and anti-stigma campaigns for college students may help reduce the duration of untreated psychosis. Full article

Carbohydrate intake (CI) has the largest impact on the short-term glycaemia of all nutrients, yet optimal CI management in gestational diabetes remains unclear. Objective: To examine CI behaviours in individuals with recently diagnosed gestational diabetes and their association with self-monitored blood glucose. Methods: Data from 97 individuals were collected using food and blood glucose diaries. CI (including amounts, types, and timing) was manually assessed using 15 g servings over 5–8 days, while a 3-day computerised analysis examined a broader intake of macro- and micronutrients. Results: Elevated fasting glycaemia (EFG) was associated with lower total CI (Mdn 10.8 vs. 12.5 servings/day, p = 0.006), missed meals (Mdn 1.4 vs. 0.0/week, p = 0.007), missed snacks (Mdn 10.5 vs. 7.5/week, p = 0.038), low-carbohydrate meals (<30 g; Mdn 4.3 vs. 2.0/week, p = 0.004), and a higher proportion of energy intake from fat (Mdn 35% vs. 33%, p = 0.047), as compared with in-target fasting glycaemia. In contrast, elevated postprandial glycaemia (EPG) was not significantly associated with total CI, low-carbohydrate meals (<30 g), low-carbohydrate snacks (<15 g), or high-carbohydrate loads (>50 g). EPG was instead associated with high-glycaemic index meals (Mdn 1.6 vs. 0.9 lunch meals/week, p = 0.026; Mdn 0.9 vs. 0.0 dinner meals/week, p = 0.023); and a lower intake of energy (Mdn 7650 kJ vs. 9070 kJ/day, p = 0.031), protein (Mdn 91 g vs. 109 g/day, p = 0.015), fat (Mdn 61 g vs. 84 g/day, p = 0.003), and multiple micronutrients (p <0.05), as compared with in-target postprandial glycaemia. Conclusions: CI profiles differed for individuals with EFG, as compared with EPG, emphasising the need for dietary guidelines tailored for gestational diabetes subtypes. Further interventional studies are needed to explore these findings, particularly the associations between EFG and low CI behaviours. Full article

Background/Objectives: Chronic venous insufficiency (CVI), a chronic vascular dysfunction, is a common health problem that causes serious complications such as painful varicose veins and even skin ulcers. Identifying the underlying genetic and epigenetic factors is important for improving the quality of life of individuals with CVI. In the literature, many genes, variants, and miRNAs associated with CVI have been identified through genomic and transcriptomic studies. Despite molecular pathogenesis studies, how the genes associated with CVI are regulated by miRNAs and the effect of variants in binding regions on expression levels are still not fully understood. In this study, previously identified genes, variants, and miRNAs associated with CVI, common variants in the mRNA-miRNA binding regions, were investigated using in silico analyses. Methods: For this purpose, miRNA research tools, MBS (miRNA binding site) database, genome browsers, and the eQTL Calculator in the GTEx portal were used. Results: We identified SNVs associated with CVI that may play a direct role in the miRNA-mediated regulation of the ZNF664, COL1A2, HFE, MDN, MTHFR, SRPX, TDRD5, TSPYL4, VEGFA, and APOE genes. In addition, when the common SNVs in the mRNA binding region of 75 unique CVI related-miRNAs in five candidate genes associated with CVI were examined, seven miRNAs associated with the expression profiles of ABCA1, PIEZO1, and CASZ1 genes were identified. Conclusions: In conclusion, the relationship between genetic markers identified in the literature that play a role in the pathogenesis of the CVI and the expression profiles was evaluated for the first time in the mRNA-miRNA interaction axis. Full article

Background: Acetabular fractures continue to pose a major challenge in clinical practice, not least because of the growing geriatric population. While the influence of the force vectors on fracture formation is well established, the impact of anatomical factors on fracture morphology remains poorly understood. The aim of this study was to investigate patient-specific hip joint geometry, identify structural risk factors and correlate these with the resulting fracture patterns. Methods: This retrospective cohort analysis included 226 patients (Mdn age = 58 yrs.) with acetabular fracture categorized by Judet/Letournel and the AO/OTA classification. Computed tomography (CT) datasets of the injured and contralateral sides were analyzed using multiplanar reconstruction. Parameters included modified center-edge (CE) angle (Wiberg), rotation angles (Ullmann and Anda), acetabular sector angle (Anda), true caput-collum-diaphyseal (CCD) angle, femoral head diameter and volume, as well as femoral neck length, circumference, and diameter. In addition, intrarater reliability within a subcohort was assessed for the metric measurements and inter-rater analysis for the classification of the entire sample. Results: The primary analysis showed direct effects of femoral head diameter, femoral neck length and femoral head size on the fracture type according to AO/OTA (type A/B/C), whereby this effect was particularly seen between type A and type C fractures (p = 0.001). Ordinal regression identified femoral head diameter as the only significant predictor (p = 0.02), with a 25% increased likelihood of complex fractures per unit of change. Low-energy trauma doubled the risk of severe fractures. Specific findings include a higher acetabular anteversion in anterior column fractures. Age correlated positively with the cause of injury and fracture type. The inter-rater reliability for fracture classification was excellent, as was the intrarater reliability of the measurements. Conclusions: This study suggests that anatomical factors, particularly proximal femoral geometry, have an impact on acetabular fracture morphology—in addition to factors such as trauma type and patient demographics. Full article

Interfacial adhesion is one of the key factors affecting the reliability of micro–nano systems. The adhesion contact mechanism is still unclear as the time-dependent viscoelasticity of soft materials. To clarify the adhesion interaction, the pull-off detachment between the rigid flat punch and viscoelastic substrate is explored considering the viscoelasticity of soft materials and rate-dependent adhesion. Taking the Lennard-Jones (L-J) potential characterizing interfacial adhesion and the Prony series defining the viscoelasticity of materials as references, the bilinear cohesion zone model (CZM) and standard Maxwell model are employed, and an adhesion analysis framework is established by combining finite element technology. The influence laws of the loading and unloading rates, material relaxation coefficients and size effect on adhesion pull-off behavior are revealed. The results show that the pull-off force is independent of the material relaxation effect and related to the unloading rate. When $\widehat{v}$ ≥ 50 or $\widehat{v}$ < 0.01, the pull-off force has nothing to do with the unloading rate, but when 0.01 < $\widehat{v}$ < 50, the pull-off force increases with the increasing unloading rate. Also, it is controlled by the size effect, and the changing trend conforms to the MD-n model proposed by Jiang. The energy required for interfacial separation (i.e., effective adhesion work) is a result of the comprehensive influence of unloading rates, material properties and the relaxation effect, which is consistent with Papangelo1’s research results. In addition, we derive the critical contact radius of the transition from the Kendall solution to the strength control solution. This work not only provides a detailed solution for the interfacial adhesion behavior but also provides guidance for the application of adhesion in Micro-Electro-Mechanical Systems (MEMSs). Full article

Intense rainfall events frequently occur in Brazil, often leading to rapid flooding. Despite their recurrence, there is a notable lack of sub-daily studies in the country. This research aims to assess patterns related to the structure and microphysics of clouds driving intense rainfall in Brazil, resulting in high accumulation within 1 h. Employing a 40 mm/h threshold and validation criteria, 83 events were selected for study, observed by both single and dual-polarization radars. Contoured Frequency by Altitude Diagrams (CFADs) of reflectivity, Vertical Integrated Liquid (VIL), and Vertical Integrated Ice (VII) are employed to scrutinize the vertical cloud characteristics in each region. To address limitations arising from the absence of polarimetric coverage in some events, one case study focusing on polarimetric variables is included. The results reveal that the generating system (synoptic or mesoscale) of intense rain events significantly influences the rainfall pattern, mainly in the South, Southeast, and Midwest regions. Regional CFADs unveil primary convective columns with 40–50 dBZ reflectivity, extending to approximately 6 km. The microphysical analysis highlights the rapid structural intensification, challenging the event predictability and the issuance of timely, specific warnings. Full article

Accurate assessment and prediction of mandible shape are fundamental prerequisites for successful orthognathic surgery. Previous studies have predominantly used linear models to predict lower facial structures from facial landmarks or measurements; the prediction errors for this did not meet clinical tolerances. This paper compared non-linear models, namely a Multilayer Perceptron (MLP), a Mixture Density Network (MDN), and a Random Forest (RF) model, with a Linear Regression (LR) model in an attempt to improve prediction accuracy. The models were fitted to a dataset of measurements from 155 subjects. The test-set mean absolute errors (MAEs) for distance-based target features for the MLP, MDN, RF, and LR models were respectively 2.77 mm, 2.79 mm, 2.95 mm, and 2.91 mm. Similarly, the MAEs for angle-based features were 3.09°, 3.11°, 3.07°, and 3.12° for each model, respectively. All models had comparable performance, with neural network-based methods having marginally fewer errors outside of clinical specifications. Therefore, while non-linear methods have the potential to outperform linear models in the prediction of lower facial measurements from upper facial measurements, current results suggest that further refinement is necessary prior to clinical use. Full article

Globally, cognitive impairment (CI) is the leading cause of disability and dependency among the elderly, presenting a significant public health concern. However, there is currently a deficiency in pharmacological interventions that can effectively cure or significantly reverse the progression of cognitive impairment. Methyl donor nutrients (MDNs), including folic acid, choline, and vitamin B12, have been identified as potential enhancers of cognitive function. Nevertheless, there remains a dearth of comprehensive research investigating the connection between the dietary intake of MDNs and CI. In our study, we comprehensively assessed the relationship between MDNs’ dietary intake and CI in older adults, utilizing 16S rRNA gene sequencing to investigate the potential underlying mechanisms. The results showed an obvious difference in the methyl-donor nutritional quality index (MNQI) between the dementia (D) group and the dementia-free (DF) group. Specifically, there was a lower MNQI in the D group than that in the DF group. For the gut microbiome, the beta diversity of gut flora exhibited higher levels in the high methyl-donor nutritional quality (HQ) group as opposed to the low methyl-donor nutritional quality (LQ) group, and lower levels in the D group in comparison to the DF group. Subsequently, we performed a correlation analysis to examine the relationship between the relative abundance of microbiota, the intake of MDNs, and Montreal Cognitive Assessment (MoCA) scores, ultimately identifying ten genera with potential regulatory functions. Additionally, KEGG pathway analyses suggested that the one-carbon metabolism, chronic inflammation, and DNA synthesis potentially serve as pathways through which MDNs may be promising for influencing cognitive function. These results implied that MDNs might have the potential to enhance cognitive function through the regulation of microbiota homeostasis. This study offers dietary recommendations for the prevention and management of CI in the elderly. Full article