Search Results (271,649)

Background and Objectives: This study aimed to evaluate the safety and feasibility of reduced-port laparoscopic distal gastrectomy (RPLDG) in elderly patients. Materials and Methods: Electronic medical records of 226 patients who underwent RPLDG performed by a single high-volume surgeon at a single institution (Chonnam National University Hospital) between January 2015 and April 2020 were retrospectively analyzed. Among these patients, 60 were aged ≥ 75 years (elderly group), and 166 were younger than 75 years (non-elderly group). Patient characteristics, surgical outcomes, and short-term postoperative outcomes were compared between the two groups. Results: Surgical outcomes were comparable between the age groups. However, medical complication events, assessed using an event-based approach allowing multiple events per patient, were more frequent in patients aged ≥ 75 years. Compared with the non-elderly group, the elderly group demonstrated a higher frequency of overall postoperative complication events (18 [30%] vs. 29 [18%], p = 0.040), primarily attributable to medical complications (9 [15%] vs. 6 [4%], p < 0.01). The elderly group also showed a delayed time to first gas passage (3.5 [3.0–4.0] vs. 3.0 [3.0–3.0] days, p < 0.001). However, no statistically significant differences were observed in length of hospital stay (7.0 [6.0–10.0] vs. 6.0 [6.0–8.0] days, p = 0.262) or intraoperative blood loss (p = 0.831). No significant differences were found in surgical complication events (p = 0.05) or Clavien–Dindo grade ≥ 3 complication events (p = 0.13). In the risk factor analysis for complications, univariate analysis identified age ≥ 75 years as a significant factor. However, in the multivariate analysis, only respiratory comorbidities, liver disease, and poor ECOG performance status remained independent risk factors, whereas age ≥ 75 years was no longer statistically significant (p = 0.193). The finding regarding liver disease should be interpreted with caution because of the extremely small sample size. Conclusions: RPLDG appears to be a viable surgical option for patients aged ≥ 75 years, demonstrating acceptable surgical outcomes and severe complication rates comparable to those observed in non-elderly patients. Full article

Background and Objectives: Hip fracture surgery is considered a major operation due to the risk of complications and increased mortality. COVID-19 is a newly recognized risk factor for increased mortality in regard to various diseases. Materials and Methods: The aim of this retrospective observational study, conducted from January 2018 to April 2022, was to analyze mortality among rehabilitation ward patients after surgical treatment of acetabular or femoral fractures in both the COVID-19 and pre-COVID-19 periods. The association between mortality and age, gender, comorbidity status, and number of complications during hospital stay was also examined. Results: Higher mortality was observed in the COVID-19-period group during all analyzed periods: cumulative three-year mortality was 2.3 times higher (14.2% vs. 6.2%, p = 0.013); two-year mortality was 3.7 times higher (9.2% vs. 2.5%, p = 0.005); and first-year mortality was 8.3 times higher (5.0% vs. 0.6%, p = 0.006). The Charlson Comorbidity Index (CCI) and admission during the COVID-19 period were strong predictors of mortality, while the number of complications, age, and gender did not significantly influence the mortality rate. An increase of one point in CCI resulted in a 42% increase in the likelihood of mortality, while hospitalization during the COVID-19 period was associated with an odds ratio of 2.44 for death compared to the pre-COVID-19 period (p = 0.013, 95% CI [1.19, 4.94]). Conclusions: The excess mortality may be attributed to the COVID-19 pandemic because the groups were comparable in all other aspects (Barthel index, CCI, complications, age, and gender). Additional five-year mortality data will be useful for analyzing mortality dynamics, as pre-COVID-19 patients will enter the COVID-19 period and COVID-19 patients will enter the post-COVID-19 period. Full article

Photovoltaic (PV) performance in desert environments is significantly hindered by soiling and partial shading. To bridge the gap in empirical data for extreme Saharan conditions, this study presents a novel techno-economic assessment of uniform horizontal shading (UHS) specifically conducted in Mauritania. Controlled outdoor experiments were performed using a 250 W crystalline silicon PV module and a PVPM 2540C I–V curve tracer, applying progressive shading levels from 2.5% to 20%. The novelty of this work lies in the integration of high-resolution experimental I–V/P–V characterization with a localized techno-economic model for five pre-commercial PV plants. It was observed that PV modules are exceptionally sensitive to shading; specifically, a mere 10% shaded area leads to a catastrophic 90% drop in power and current, while the voltage remains remarkably stable. Thermographic analysis further validates the thermal gradients and bypass diode functionality. By quantifying the financial impacts, this research highlights that cumulative economic losses across the five real-world sites reached 87.95%, exceeding 55,000 MRU. These findings provide a strategic framework for optimizing PV systems in arid terrains and offer a robust tool for enhancing the design and operation of large-scale solar applications in desert environments. Full article

Recent advances in artificial intelligence have expanded the use of large language models (LLMs) beyond speech-based applications and increased interest in their potential roles in dental education. However, evidence regarding LLM performance in postgraduate dental education, particularly in prosthodontics, remains limited. Therefore, this study aimed to compare the accuracy of responses from prosthodontic residents and LLMs to standardized multiple-choice questions in prosthodontics and to explore the potential role of artificial intelligence in prosthodontic education. Thirty-two prosthodontic residents participated in this cross-sectional study. Participants completed a standardized 30-item multiple-choice test comprising four demographic items and 26 questions assessing basic knowledge, general dentistry, and advanced prosthodontic specialty questions. The same questions were administered to seven large language models (LLMs): ChatGPT-4o, ChatGPT-o1, ChatGPT-o3-mini, Claude Sonnet 3.7, Gemini 2.5 Pro, Microsoft Copilot (web interface, accessed in August 2025), and DeepSeek V3. Response accuracy and consistency were evaluated. Statistical analyses were performed using IBM SPSS Statistics (version 27.0), with statistical significance set at p < 0.05. A statistically significant difference was observed between prosthodontic residents and LLMs in responses to advanced-level prosthodontic specialty questions (p < 0.05), with higher correct response rates recorded for LLMs. No statistically significant differences were identified between the two groups for basic knowledge and general dentistry questions (p > 0.05). In addition, no significant association was found between the duration of prosthodontic residency training and residents’ response accuracy (p > 0.05). LLMs achieved high scores on this structured MCQ-based assessment, particularly in advanced theoretical prosthodontic items. However, these findings should be interpreted with caution within the limits of a written examination format and do not represent overall clinical competence or real-world patient care performance. Accordingly, artificial intelligence may be considered a supportive educational tool in postgraduate prosthodontic education rather than a replacement for clinical training. Full article

Curcumin exhibits potent anti-obesity and anti-inflammatory activities; however, its therapeutic application is limited by poor aqueous solubility and low oral bioavailability. A curcumin-loaded chewable gel was developed to transform into an in situ gastric gel upon contact with gastric fluid after mastication. Curcumin solid dispersions (CUR-SDs) were prepared with Eudragit^® EPO (1:1–1:7, w/w) using the solvent evaporation method. The optimized formulation (1:3) markedly enhanced solubility and dissolution in acidic medium (0.1 N HCl, pH 1.2) compared with crystalline curcumin and physical mixtures. The optimized CUR-SD was subsequently incorporated into chewable gels composed of sodium alginate and κ-carrageenan, with calcium carbonate as a gas-forming agent. The formulations formed buoyant matrices under acidic conditions, exhibiting floating lag times of 21–215 s and sustaining drug release for up to 8 h. Increasing polymer content improved mechanical strength and modulated release kinetics. Among the tested formulations, F7 achieved the optimal balance between texture properties, floating behavior, and controlled-release performance. In LPS-stimulated RAW264.7 macrophages, curcumin, CUR-SD, and F7 showed comparable and potent anti-inflammatory activity (IC₅₀ = 4.12–4.84 µg/mL), outperforming indomethacin. In 3T3-L1 adipocytes, F7 significantly reduced lipid accumulation (~47%) in a concentration-dependent manner. These findings demonstrate that this transformable chewable in situ gelling platform is a promising gastroretentive strategy for improving the oral therapeutic efficacy of poorly soluble bioactive compounds for anti-obesity applications. Full article

Background: Age- and cancer-related sarcopenia and malnutrition are common in older patients with colorectal cancer (CRC) and may negatively influence treatment tolerance and prognosis. However, the comparative prognostic value of post-chemotherapy changes in CT-based body composition parameters at the third lumbar vertebra (L3) and the twelfth thoracic vertebra (T12) levels, and their associations with nutritional indices, remain unclear. This study aimed to examine and compare the prognostic relevance of post-chemotherapy body composition changes at L3 and T12 and to assess their relationship with nutritional indices in older patients with metastatic CRC (mCRC). Methods: This retrospective study included 87 older patients with mCRC. Baseline and ~3-month follow-up CT scans were analyzed at L3 and T12 using 3D Slicer to quantify skeletal muscle index (SMI), subcutaneous adipose tissue index (SATI), visceral adipose tissue index (VATI), visceral-to-subcutaneous fat ratio (VSR), and intramuscular adipose tissue index (IMATI). Changes (Δ) in CT-derived body composition after chemotherapy were calculated as percentage change using ((follow-up − baseline)/baseline) × 100. Prognostic Nutritional Index (PNI) and Geriatric Nutritional Index (GNRI), which are established nutritional assessment tools, were calculated from baseline laboratory/anthropometric data. Agreement between T12 and L3 was assessed, and associations with grade ≥3 toxicity, progression-free survival (PFS), and overall survival (OS) were evaluated using multivariable models and ROC analyses. Results: Mean age was 69.0 ± 4.5 years (59 male/28 female), and 26.4% developed grade ≥3 adverse events. Over 3 months, mean SMI declined significantly at both L3 (46.7 ± 8.8 → 42.8 ± 9.8 cm²/m²) and T12 (34.6 ± 8.2 → 31.6 ± 8.1 cm²/m²) (p < 0.001 for both), accompanied by decreases in VATI and VSR; T12-IMATI increased significantly. Baseline PNI showed a weak positive correlation with L3-SMI (r = 0.302, p = 0.033), whereas GNRI showed moderate correlations with SMI at L3 (r = 0.502, p < 0.001) and T12 (r = 0.317, p = 0.025) and was associated with longitudinal changes in muscle metrics. T12-SMI consistently yielded lower values than L3-SMI, and agreement varied by compartment (best for SATI; weakest for VSR). Lower GNRI and greater L3-SMI loss were independently associated with grade ≥3 toxicity; ΔL3-SMI showed the highest discrimination (AUC = 0.79, 95% CI = 0.69–0.87, p < 0.001; cut-off >5.1% loss). All patients progressed (median PFS 7.6 months); mortality was 82.8% (median follow-up: 25 months). In multivariable analysis, PFS, CRP, GNRI, and ΔL3-SMI remained independently associated with OS. ΔL3-SMI provided the strongest mortality discrimination (AUC = 0.85, 95% CI = 0.74–0.94, p < 0.001; cut-off >10.4% loss), while ΔIMATI was also informative (AUC = 0.71, 95% CI = 0.59–0.82, p = 0.023). Conclusions: In older patients with mCRC, early post-chemotherapy skeletal muscle loss—particularly at the L3 level—showed the strongest prognostic association with severe toxicity and mortality. GNRI provided complementary prognostic information as a marker of baseline immunonutritional reserve. Although T12-derived measurements were correlated with L3-derived values, systematic bias suggests that they should not be interpreted interchangeably for longitudinal risk stratification. Full article

Background/Objectives: Urine samples are the most frequently analyzed specimens in clinical microbiology laboratories. Although urine culture remains the gold standard for diagnosing urinary tract infections, it is time-consuming and resource-intensive. Therefore, reliable screening methods capable of predicting urine culture positivity are needed to optimize laboratory workflow. Automated urine analysis based on flow cytometry enables efficient screening and identification of samples with a low probability of bacterial infection, thereby rationalizing microbiological testing. This study evaluated the usefulness of a multivariable approach to support interpretation of flow cytometry results following the implementation of the Sysmex UF-4000 urine flow cytometer. Methods: Routinely collected urine samples from outpatients and hospitalized patients were analyzed using the UF-4000 flow cytometer, with a positivity threshold of ≥100 leukocytes/µL. Urinary parameters were compared between samples with positive and negative cultures. Multivariable logistic regression was applied to identify independent predictors of a positive urine culture. Urinary sediment parameters, including leukocyte, bacterial, fungal, and squamous epithelial cell counts, were assessed as covariates. Results: Urine samples with positive cultures showed significantly higher leukocyte counts (median 355.0, IQR 146.5–1429.4) and bacterial counts (median 9805.2, IQR 1134.3–45,011.5). Fungal and squamous epithelial cell counts differed only slightly between groups, although the differences were statistically significant (p < 0.001). Leukocyte counts were higher in urine samples from which Gram-negative bacteria were isolated compared with samples containing Gram-positive bacterial isolates (p < 0.001). The multivariable model demonstrated the most favorable overall performance, combining high sensitivity with improved specificity and the highest negative predictive value (AUC = 0.927). Optimal cut-off values were 70 leukocytes/µL and 105 bacteria/µL. Conclusions: Leukocyte and bacterial counts were the strongest predictors of positive urine culture results. A multivariable model including only these two parameters demonstrated high diagnostic accuracy and may serve as a practical screening tool to identify urine samples with a low probability of bacterial infection. The implementation of this approach could support more efficient use of urine cultures and help optimize laboratory workflow. Full article

Temporal monitoring of bioprocesses is inherently complex because process variables do not evolve independently over time, and their interpretation changes as the reaction progresses. In many existing abnormality detection methods, sensor signals are analyzed at isolated time points or temporal characteristics are only weakly reflected through model structures. As a result, such approaches struggle to explain or detect abnormal behavior that emerges differently across reaction states. This study proposes TPP-TimeNet, a time-aware artificial intelligence framework developed to improve abnormality detection in bioprocess monitoring. Unlike conventional methods, the proposed framework explicitly incorporates reaction time as contextual information. Multivariate process signals are reorganized into sliding windows that reflect reaction-state transitions rather than uniform time segmentation. Temporal behavior inside each window is captured using a sequential encoding model, and reaction-state information is subsequently integrated to form state-dependent representations. Through this design, the model can distinguish between temporal patterns that are similar in shape but occur at different points in the reaction timeline. This capability leads to improved sensitivity to abnormal events that may otherwise remain undetected. Abnormality is evaluated at the window level using a probabilistic scoring scheme with a fixed threshold, enabling consistent and reproducible decision-making. The performance of TPP-TimeNet was evaluated using publicly available process control datasets from Kaggle. The datasets were reinterpreted in a bioprocess context by mapping variables such as temperature, pH, and pressure. Experimental results show that the proposed method outperforms traditional machine learning models as well as deep learning approaches that focus only on temporal features, achieving higher accuracy, sensitivity, and F1-score. These findings suggest that incorporating explicit reaction-state awareness is essential for effective abnormality detection in bioprocess monitoring systems. Full article

Background: Adrenocortical carcinoma (ACC) is a rare and aggressive endocrine malignancy, for which population-level evidence regarding prognostic factors and survival conditions is limited. The available data mostly represent single-institution series, limiting their applicability. This study, therefore, assesses clinicopathological features and determines independent predictive variables of overall survival (OS) in patients with ACC using a population-based cohort. Methods: This retrospective observational cohort study used data from the Surveillance, Epidemiology, and End Results (SEER) Program between 2000 and 2022, initially identifying 1176 patients with ACC. Adult patients (≥18 years) with histologically confirmed ACC were identified using ICD-O-3 histology code 8370/3 and primary site code C74.0. Cases with zero-month survival, missing survival data, or identified only through autopsy or death certificate were excluded. To ensure dataset harmonization, patients with missing or indeterminate tumor grade and unknown stage were also excluded. After applying these inclusion and exclusion criteria, the final analytic cohort consisted of 267 patients. Data on demographic factors, stage of the disease, and treatment (surgery, chemotherapy, radiotherapy) were extracted. OS was evaluated using the Kaplan–Meier method, and independent prognostic factors were identified using Cox proportional hazards regression analysis. Results: The median OS was 54 months [95% confidence intervals (CI): 36–85]. The estimated 1-, 3-, and 5-year OS rates were 77%, 57%, and 48%, respectively. Survival differed significantly according to tumor grade, stage, and surgical treatment. In multivariable Cox regression analysis, increasing age [Hazard ratio (HR): 1.03, 95% CI: 1.02–1.04; p < 0.001], high tumor grade (HR: 2.21, 95% CI: 1.43–3.41; p < 0.001), and distant-stage disease (HR: 3.24, 95% CI: 1.95–5.38; p < 0.001) were independently associated with an increased risk of mortality, whereas surgical treatment was associated with improved survival (HR 0.53, 95% CI 0.30–0.93; p = 0.028). Chemotherapy and radiotherapy were not significantly associated with mortality. Conclusion: In this SEER-based cohort of patients with adrenocortical carcinoma, older age, high tumor grade, and distant-stage disease were independently associated with worse OS, whereas documented receipt of surgery was associated with longer OS. Treatment-related associations should be interpreted cautiously in view of the inherent limitations of registry-based data. Further prospective multicenter studies are needed to confirm these findings. Full article

In this work, kyanite-reinforced alumina ceramics were prepared using the prestress reinforcement method and the particle enhancement method. The effects of different preparation methods on the mechanical properties and microstructures of kyanite-reinforced alumina ceramics were investigated. The results showed that, compared to the pure alumina ceramic, the prestressed alumina ceramic (labeled as P-Al₂O₃) prepared by the prestress reinforcement method exhibited a significant improvement (31% higher than that of pure alumina) in flexural strength. This is mainly attributed to the fact that the compressive stress existing on the surface of P-Al₂O₃ inhibited crack propagation; therefore, the fracture energy and strength were increased. However, due to the numerous pores and cracks in the fracture surface caused by the decomposition reaction of kyanite, the alumina composites fabricated through the particle enhancement method (labeled C-Al₂O₃) displayed lower flexural strength and hardness than those with P-Al₂O₃. Additionally, an increase in kyanite content led to a decrease in properties such as flexural strength, Vickers hardness, density, the elastic modulus, and the thermal expansion coefficient, while resulting in an increase in porosity. This work demonstrates the importance of using a suitable preparation method aligned with the specific composite. Full article

The river ecosystems of the Qinghai–Tibet Plateau, recognized as a vital component of the “Asian Water Tower,” possess unique hydrological conditions and extreme environments that have shaped key indicator groups, most notably zooplankton. The community dynamics and structural characteristics of these zooplankton exhibit regular spatio-temporal distribution patterns across elevational gradients and seasonal successions. However, the intrinsic mechanisms underlying community succession and their correlations with environmental factors remain poorly understood, and the primary environmental drivers influencing community structure require further elucidation. Based on systematic zooplankton surveys and environmental data collection conducted across the Lhasa River basin from 2019 to 2021, this study established a comprehensive species inventory comprising 113 taxa across four major groups, alongside a multi-dimensional environmental dataset. We analyzed the spatio-temporal heterogeneities of zooplankton community structures—including abundance, biomass, and diversity indices—across different seasons and river reaches. The results revealed the composition and seasonal turnover of dominant taxa, with rotifers accounting for 39.82% of the total taxonomic richness. Mean zooplankton abundance and biomass across the basin were 1.18 ind./L and 343.60 × 10⁻⁵ mg/L, respectively, with peak values observed during autumn and within the Chabalang Wetland. The zooplankton community structure in the upstream, midstream, and downstream reaches, as well as associated wetlands, was significantly correlated with specific environmental factors (p < 0.05), including ammoniacal nitrogen (NH₄⁺-N), magnesium (Mg²⁺), total hardness (TH), potassium (K⁺), iron (Fe²⁺), sodium (Na⁺), sulfite (SO₃²⁻), nitrate ion (NO₃⁻), chloride ion (Cl⁻), total phosphorus (TP), and sulfide (S²⁻). Cl⁻, TH, Mg²⁺, SO₃²⁻, and elevation (Ele) were the key environmental drivers significantly influencing zooplankton abundance across seasons (p < 0.05). Furthermore, zooplankton abundance decreased significantly with increasing elevation during the winter. This research deepens our understanding of community assembly mechanisms in plateau river ecosystems and provides a scientific foundation for aquatic biodiversity conservation and ecological management in the Lhasa River basin. Full article

Background and Objectives: Severe COVID-19 frequently fulfills Sepsis-3 criteria and is characterized by thrombo-inflammation and endothelial injury. We evaluated whether a bedside endothelial activation index (EAI = D-dimer/fibrinogen) identifies biologically distinct phenotypes and relates to interleukin-6 (IL-6) response after therapeutic plasma exchange (TPE), and whether baseline IL-6 predicts a ≥50% IL-6 reduction. Methods: Retrospective single-center ICU cohort of adults with SARS-CoV-2 infection, sepsis-related organ dysfunction, and ≥1 TPE session (n = 51). Patients were stratified by median EAI (low vs. high). Outcomes included peri-procedural biomarker/physiology changes (post–baseline), IL-6 responder status (≥50% reduction), correlations with IL-6 reduction (%), and multivariable predictors of response. Results: Compared with low EAI (n = 25), high EAI (n = 26) had higher baseline D-dimer (6.2 vs. 2.2 µg/mL) and lower fibrinogen (2.9 vs. 7.1 g/L) (both p < 0.001). Low EAI showed larger CRP decreases (ΔCRP −84.0 vs. −2.3 mg/L; p = 0.001) and larger fibrinogen falls (Δ −3.1 vs. −0.4 g/L; p < 0.001), while high EAI had larger D-dimer decreases (Δ −2.5 vs. −0.6 µg/mL; p = 0.004) and a modest SOFA improvement (Δ −0.3 vs. +0.1; p = 0.026). IL-6 responders (n = 20) had higher baseline IL-6 than non-responders (365.2 vs. 47.1 pg/mL; p < 0.001). Baseline IL-6 independently predicted response (per doubling: OR 1.94, 95% CI 1.27–2.95; p = 0.002), while age reduced odds (OR 0.91/year, 95% CI 0.84–0.99; p = 0.032). IL-6 reduction correlated with ΔCRP (ρ = −0.41; p = 0.003) and ΔPaO₂/FiO₂ (ρ = 0.37; p = 0.01). Conclusions: EAI stratifies distinct thrombo-inflammatory patterns around TPE, while baseline IL-6 is the dominant predictor of achieving large IL-6 reductions. To emphasize the novelty and clarify the study objective, this exploratory analysis used a phenotype-stratified framework to test whether a simple bedside endothelial activation index could enrich biological response assessment to adjunctive TPE. The prespecified primary outcome was achievement of a ≥50% IL-6 reduction after completion of the TPE course; secondary outcomes included peri-procedural biomarker, oxygenation, SOFA, and ICU endpoints. Full article

Background/Objectives: Increased accessibility of suicidal cognitions reflects the cognitive processes underlying the acquisition of suicidal thoughts. Previous research shows that expectation violation reduces the accessibility of life cognitions rather than increasing that of suicidal cognitions, but this may be due to a slowing effect masking an increase in suicidal cognitions. Methods: Beyond the reaction time task, the present study used simultaneous EEG-fNIRS to reveal how expectation violation differentially affects the accessibility of suicidal and life cognitions. In a trial-by-trial cognitive task, participants read sentences that were either semantically consistent (expectation confirmation) or anomalous (expectation violation), followed by a semantic judgment on suicide-related, neutral, and life-related words. Response times for each word type served as a measure of cognitive accessibility for that category. Results: Compared to expectation confirmation, expectation violation reduced the cognitive accessibility of life rather than increasing that of suicide in the reaction time task. However, in neural responses, it led to reduced N1 amplitude, increased P2 amplitude for suicide-related information, and greater hemodynamic response in the left frontopolar region. Conclusions: Expectation violation triggered distinct neural responses to suicidal information, reflecting an attentional bias that may explain how suicidal thoughts emerge within normative cognition. Full article

Background/Objectives: Multimodal depression detection research has traditionally relied on early or hybrid fusion strategies without systematically analyzing how dynamic fusion mechanisms interact with modality-specific pretraining. Although gated and attention-based architectures are increasingly adopted, their behavior is rarely examined within a structured fusion taxonomy framework. Methods: In this study, we conduct a controlled taxonomy-level evaluation of multimodal fusion strategies in a Japanese PHQ-9-annotated depression dataset. We compare four fusion paradigms (concatenation, summation, gated fusion, and cross-attention) across three integration stages, crossed with modality-specific affective pretraining configurations for visual (CMU-MOSI/MOSEI), acoustic (JTES), and textual (WRIME) encoders, yielding 512 experimental conditions. Results: The results reveal strong position-dependent effects of fusion strategy. Cross-attention fusion at the audio integration stage achieved the highest mean AUC (0.774) and PR-AUC (0.606), with statistically significant superiority over gated and concatenation-based fusion (Kruskal–Wallis $H=86.28$, $p<0.001$). In contrast, fusion effects at the text stage were non-significant in AUC but significant in PR-AUC, highlighting metric-sensitive behavior under class imbalance. Pretraining effects were modality-specific: SigLIP initialization produced significant positive transfer $(\Delta =+0.018,p<0.001)$, whereas audio pretraining on JTES resulted in negative transfer $(\Delta =-0.014,p=0.004)$, suggesting domain mismatch effects. Gate analysis further revealed condition-dependent modality dominance, including cases of semantic–geometric reversal under joint auxiliary augmentation. Conclusions: Our findings suggest that multimodal depression detection systems should not be interpreted through static fusion categories alone. Instead, modality contribution appears to be associated with structured interaction effects between fusion strategy, integration position, and affective pretraining. This work provides a controlled empirical bridge between fusion taxonomy and dynamic modality weighting in clinical multimodal modeling. Full article

Water pollution is one of the most critical societal and environmental challenges and remains a persisting problem worldwide. The origin of this pollution is diverse, while organic matter occupies a significant portion, originating from different sources. This creates major environmental and health risks, requiring reliable and sensitive analytical tools for effective monitoring. The permanganate index stands as a conventional assessment method for organic pollution, but it demonstrates compound non-specificity toward compounds and limited sensitivity to various contaminant structures. This research introduces cyclic voltammetry as a standalone electrochemical method that provides sensitive detection and characterization of organic oxidizing compounds. Six organic compounds, including gallic acid, phenol, oxalic acid, ascorbic acid, salicylic acid and p-benzoquinone, were used as model compounds and studied in aqueous media. These compounds were analyzed individually, in single-compound mode, to characterize their redox behavior and to identify the voltammetric peaks. Subsequently, a multi-compound analysis was studied to check for the validity of the concept in a more complex matrix. Notably, a strong linear correlation was observed between the measured charge and the theoretical permanganate index, highlighting the quantitative reliability of the electrochemical method. Comparing the obtained results with the permanganate index method confirmed the superiority of cyclic voltammetry in terms of response time and detection capability. The outcomes demonstrate that cyclic voltammetry functions as a robust alternative to the classical chemical oxidation method for environmental water assessment. Full article