Browse Articles

Inpatient mortality and associated clinical factors among people living with HIV with cryptococcal meningitis in Uganda: A retrospective cohort study

Introduction Cryptococcal meningitis remains a leading cause of HIV-related mortality in sub-Saharan Africa despite expanded antiretroviral therapy coverage. Evidence on the burden of disease and inpatient mortality among people living with HIV (PLHIV) in routine care settings in Uganda remains limited. This study assessed the proportion of cryptococcal meningitis among HIV-related admissions and examined clinical factors associated with inpatient mortality. Methods We conducted a retrospective cohort study of adult PLHIV admitted with cryptococcal meningitis between 1 st /01/2017–31 st /12/2022 at a national referral hospital in Uganda. Diagnosis was based on cerebrospinal fluid cryptococcal antigen or India ink positivity. Data were abstracted from medical records and analysed using descriptive statistics and multivariable logistic regression to identify factors associated with inpatient mortality. Specific antifungal treatment regimens were not consistently documented and could not be analysed. Results Of 3,042 HIV-related admissions, cryptococcal meningitis accounted for 21.4% (650/3,042). Medical records for 634 patients were analysed, among whom 39.3% (249/634) died during hospitalization. Factors independently associated with higher odds of inpatient mortality included convulsions, headache, vomiting, cryptococcal meningitis–associated immune reconstitution inflammatory syndrome, concurrent opportunistic infections, chronic kidney disease, anaemia, and severe immunosuppression (low CD4 cell count). Longer duration of hospitalization (≥7 days) and symptom duration of one to two weeks before admission were associated with lower odds of mortality. Conclusion Cryptococcal meningitis continues to account for a substantial proportion of HIV-related hospital admissions and inpatient deaths in Uganda. Mortality is associated with identifiable clinical and health-system factors, underscoring the need for early diagnosis, risk stratification, and optimized inpatient management for PLHIV with cryptococcal meningitis in resource-limited settings.

Plasma heparan sulfate structural variation and phenotypic heterogeneity in pediatric Acute Respiratory Distress Syndrome

Abstract Endothelial glycocalyx (eGCX) shedding contributes to microvascular endotheliopathy in Acute Respiratory Distress Syndrome (ARDS) and may represent an underrecognized source of phenotypic heterogeneity. We examined whether circulating heparan sulfate (HS) signatures, as readouts of eGCX shedding, capture patterns of inter-patient biological variation distinct from other eGCX components and conventional protein biomarkers, whether specific HS structural features are enriched, and whether these signatures are associated with heparanase-1 (HPSE) activity. We retrospectively analyzed prospectively collected plasma samples (2018–2020) from children with and without pediatric ARDS (PARDS). Plasma levels (ng/mL) of sulfated and non-sulfated HS disaccharides (following enzymatic digestion of total [unfractionated] HS), and HPSE activity (U/mL) were measured using mass spectrometry, while protein biomarkers were assessed by multiplex assay. Among 46 children (36 PARDS, 10 no PARDS), principal component analysis identified three components explaining > 60% of the variance. The primary component (PC1) was characterized by an HS-driven endothelial signature and was distinct from an inflammatory protein signature captured by PC2. In PARDS, children with higher PC1 scores had worse organ dysfunction and fewer ventilator-free days after adjustment for PC2 and PC3. Higher total HS levels were further associated with selective enrichment of sulfated HS motifs, whereas the opposite pattern was observed in non-PARDS. HPSE activity positively correlated with circulating HS levels. These preliminary findings suggest that circulating HS signatures capture a distinct and clinically meaningful dimension of PARDS heterogeneity.

Optimizing an ethanol-based fixative for enhanced nucleic acid preservation in cervical samples using a central composite design approach

Background To accurately diagnose cervical cancer, high-quality genetic material (DNA and RNA) from clinical samples is crucial. Current preservation methods often have limitations, including poor RNA stability and safety concerns. This study aimed to develop an optimized ethanol-based fixative to preserve DNA and RNA in cervical samples under ambient conditions. Methods HeLa cells were fixed in ethanol-based fixatives containing polyhydric compounds (Sorbitol and polyethylene glycol [PEG]) and chelating agents. We used a central composite design (CCD) approach to evaluate the effects of pH, Sorbitol concentration, and PEG concentration on nucleic acid preservation. DNA and RNA quality were assessed using agarose gel electrophoresis, PCR, and real-time PCR. Cellular morphology was evaluated using Papanicolaou-stained slides. HPV genotyping of clinical samples was conducted using real-time PCR. Results The optimized fixative, developed in this study consisted of 40% ethanol, 4.3% Sorbitol, 1.2% PEG 8000, with a pH of 5.6. This new formula significantly improved DNA and RNA preservation compared to the commercial solution, PreservCyt. DNA showed high integrity and was successfully amplified in PCR test targeting HPV-18 oncogenes. RNA quality was confirmed through clear 28S and 18S rRNA bands and lower threshold cycle (Ct) values in qRT-PCR. HPV genotyping and morphological analysis revealed excellent preservation, enabling both molecular and cytological evaluations. Conclusion The new ethanol-based fixative represents a promising cost-effective and environmentally friendly solution for preserving nucleic acids and cellular morphology in cervical samples. Its good performance under ambient conditions suggests it may serve as an option for Human Papillomavirus (HPV) testing and cervical cancer prevention, particularly in places with limited resources.

DHCRWOA: adaptive whale optimization algorithm with Cauchy-Rayleigh distribution for numerical and engineering design optimization

Abstract This paper proposes DHCRWOA, a stage-specific extension of the Whale Optimization Algorithm (WOA) designed to improve its exploration–exploitation transition and reduce stagnation around suboptimal leaders. Five modules are introduced, each targeting a single WOA search phase: Good Nodes Set (GNS) initialization, adaptive parameter control, statistical-guided exploration, Cauchy-based local exploitation, and Rayleigh-weighted spiral updating. On the CEC2005 benchmark suite at 30 dimensions, DHCRWOA achieves the best average rank (2.13) among ten algorithms and the best or tied-best mean on 22 of 23 functions, with no baseline outperforming it on more than one function. At 50 and 100 dimensions, DHCRWOA maintains the top Friedman rank (1.69 and 1.62, respectively) at average runtimes of 0.52 s and 1.03 s per run. A five-variant ablation study confirms that weakening the statistical-guided exploration or Cauchy-based exploitation produces the largest rank degradation (from 2.35 to 4.09 and 4.70, respectively), while all five modules contribute to overall performance. Five constrained engineering design problems further confirm competitive results on practical optimization tasks. Theoretical analysis establishes the relationship between DHCRWOA and standard WOA, including a perturbation bound on the adaptive parameter ( $$\le 0.02$$ ), boundedness of the population trajectory, monotonicity of the best-so-far process, and a global reachability theorem.

Adaptive robust sparse representation for face recognition based on weighted and fusion dictionary

We propose a new model for face recognition under insufficient sampling conditions in this paper. In the proposed method, we combine the fusion dictionary with nuclear norm regularization to preserve the details of the restored images, and adopt a Laplacian-uniform mixture function to fit the error distribution. Since the proposed model is convex and separable, we employ the classic alternating direction method of multipliers to solve it by introducing auxiliary variables to transform the original problem into the saddle point problem. Theoretically, we conduct the convergence analysis of the proposed numerical algorithm. Final experimental comparisons are provided to verify the satisfactory performance of the proposed model, which outperforms other related competitive methods in both recognition rate and the robustness.

Risk-aware explainable multi-output workflow for early-stage screening of building heating and cooling load indicators

Abstract Early-stage design decisions largely determine a building’s operational energy demand, but benchmark-trained surrogate models must be used with explicit validity limits. This paper presents a risk-aware, reproducible and explainable multi-output machine-learning workflow for screening simulated heating load (HL) and cooling load (CL) indicators from concept-stage geometric and envelope parameters. Using the Energy Efficiency (ENB2012) dataset (768 simulated building configurations), Ridge regression, Random Forest and Extremely Randomised Trees (Extra Trees) are evaluated under five-fold cross-validation and a held-out test set, with additional revision diagnostics for geometry-blocked validation, split-conformal prediction intervals, compact TreeSHAP attribution, multi-output ablation and fixed-setting XGBoost, LightGBM and MLP comparators. The Extra Trees model achieves test-set RMSE of 0.60 kWh/m2 for HL and 1.45 kWh/m2 for CL, with R2 of 0.996 and 0.977, respectively. The added diagnostics show that these values should be interpreted as benchmark interpolation rather than external generalisation or HVAC plant-sizing evidence. The contribution is therefore not a new learning algorithm, but an auditable workflow combining transparent reporting, residual checks, feature-dependence-aware interpretation, scenario screening, uncertainty communication and conservative deployment gates. Results indicate that overall height, roof area and glazing area dominate the fitted model’s predictions, consistent with building-physics intuition and published studies.

Immunosuppressive regimens and long-term kidney transplant outcomes: A dual survival modeling framework

Optimizing immunosuppressive therapy remains central to improving long-term outcomes after kidney transplantation. Both induction and maintenance therapies are widely used, yet their comparative effectiveness across diverse recipient populations requires further evaluation. This national retrospective cohort study analyzed 228,855 deceased-donor kidney transplant recipients using data from 2000–2024. Multivariable Cox proportional hazards (PH) models were used for clinical inference, and four machine learning (ML) survival models: random survival forest (RSF), support vector machine (SVM), penalized Cox regression (CoxNet), and extreme gradient boosting optimized with the Cox partial likelihood (XGBoost-Cox), were developed to assess predictive performance for death-censored graft failure and all-cause patient mortality. Model performance was evaluated using the concordance index (C-index) and time-dependent area under the curve (tdAUC). Maintenance regimens incorporating calcineurin inhibitors (CNI) and mycophenolate mofetil (MMF) were associated with lower hazards for both graft failure (CNI + MMF: hazard ratio [HR] 0.72, 95% confidence interval [CI] 0.70–0.74; CNI + MMF+steroids: HR 0.84, 95% CI 0.82–0.87) and patient mortality (CNI + MMF: HR 0.78, 95% CI 0.76–0.81; CNI + MMF+steroids: HR 0.90, 95% CI 0.88–0.93). Among induction therapies, antithymocyte globulin (ATG) was associated with lower hazards for both outcomes, whereas interleukin-2 receptor (IL-2R) antagonists and Alemtuzumab demonstrated neutral associations. Combined ATG + IL-2R therapy was associated with higher hazard of graft failure (HR 1.09). Recipient diabetes, dialysis dependence, older age, and higher Kidney Donor Profile Index were strong adverse predictors. Traditional Cox regression achieved robust discrimination (graft failure C-index: 0.685; patient mortality C-index: 0.704), comparable to ML survival models. These findings support the continued association of CNI and MMF maintenance regimens with favorable long-term transplant outcomes while demonstrating variation across induction strategies. The dual analytical framework integrating classical Cox modeling with ML survival methods, suggests that Cox models remain highly competitive for clinical inference, whereas ML approaches provide complementary predictive value to support individualized post-transplant risk stratification.

Enhancing the severity classification of tomato plant epidemic pathogens using adaptive segmentation of mask RCNN and multiscale recurrent MobileNet

Integrated application of meat waste-derived organic fertilizer and chemical fertilizer in sugarcane cultivation

Slaughterhouse waste is a major source of environmental pollution. However, its rich organic content holds considerable potential for enhancing soil fertility. Herein, slaughterhouses meat waste was used to produce organic fertilizer (OF), and its effect on sugarcane growth, productivity and soil fertility were evaluated. The decomposition of the collected meat waste was carried out in a controlled fermentation pit. Initially, the sample was treated by sulfuric acid (H 2 SO 4 ), followed by the addition of potassium hydroxide (KOH) and dipotassium phosphate (K 2 HPO 4 ). The pH of the medium was then adjusted to neutral conditions, after that a bacterial inoculum was introduced to facilitate further biodegradation of the solid waste. The prepared organic fertilizer was characterized for its chemical compositions, physicochemical properties and microbial diversity. The chemical compositions analyses revealed that potentially toxic elements such as mercury (Hg), lead (Pb), cadmium (Cd) and arsenic (As) were detected at concentration of <0.002 mg/kg, 0.411 mg/kg, 0.004 mg/kg and 0.095 mg/kg, respectively, indicating their levels in the fertilizer were very low. The total organic carbon content was determined to be 13.90% that is significantly higher than that of most soil, where organic carbon levels typically fall below 2%. MALDI-TOF (Matrix-Assisted Laser Desorption/Ionization Time-of-Flight) mass spectrometry was used to identify culturable bacterial isolates obtained from the fermented organic fertilizer prior to soil application. A total of 20 morphological distinct isolates were analyzed, of which Lysinibacillus sphaericus was the most frequently detected species, followed by Bacillus simplex, Bacillus cereus, Kocuria varians, and Microascus sp. Identification was accepted based on MALDI-TOF score thresholds (≥2.0 species-level and 1.7–1.99 for genus level identification). Lysinibacillus sphaericus aids nitrogen-fixation, plant growth, and decomposing organic matter, all of which enhance soil fertility and plant health. Additionally, other species such as Bacillus simplex, Bacillus cereus, Kocuria varians , and Microascuswere contribute to phosphate solubilization and pathogen suppression. The organic ferilizer was tested alone and in combination with the conventional chemical fertilizer in a pot experiment for sugarcane. The results showed that applying 5 tonnes ha -1 of organic fertilizer together with 50% of the recommended chemical fertilizer performed comparable to that of 100% recommended chemical fertilizer alone. In addition, combined organic fertilizer and recommended chemical fertilizer resulted in a significantly higher juice phosphate content compared to the sole application of 100% recommended chemical fertilizer, likely due to the contribution of the organic fertilizer. This finding suggests that halving chemical fertilizer use by supplementing with organic alternatives could offer environmental benefits while maintaining productivity.

Development of synthetic bacteriophages with extended host range to overcome resistant Klebsiella pneumoniae

Abstract Klebsiella pneumoniae (Kp) is a leading cause of bacterial nosocomial infections. The rapid emergence of multidrug-resistant (MDR) Kp strains poses a significant global health threat, challenging current antibiotic-based therapies. While bacteriophage therapy offers a promising alternative, its effectiveness is often limited by the narrow host ranges of natural phages and the rapid emergence of phage resistance. Synthetic phages, with their diverse and customizable genomes, have the potential to overcome these limitations. In this study, we engineered synthetic bacteriophage variants using two in-house phage isolates, ɸ115 and ɸ100, which exhibited limited efficacy against a range of clinical Kp strains. In contrast, the synthetic phage libraries derived from the recombination of tail fiber genes and genomes of ɸ115 and ɸ100 exhibited a significantly extended host range and the ability to lyse phage resistant Kp strains. Indirect measurement of bacterial respiration as a growth indicator in presence of phage, using the OmniLog system revealed that no phage resistance Kp emerged against the synthetic phage libraries within 24 h, while other Kp strains developed resistance to the parental phages. Host range analysis, burst size measurements, and genomic DNA comparisons of individual synthetic phage isolates indicated that a short central tail fiber of ɸ100 likely plays a pivotal role in extending the host range and lysing resistant Kp strains. Our findings highlight the potential of synthetic bacteriophages to overcome the dual challenges of narrow host specificity and phage resistance. This represents a significant advancement toward developing viable phage therapeutics against MDR Kp infections.

A curriculum-integrated sun safety intervention to improve adolescent skin health knowledge in an independent girls’ secondary school

Australia has one of the world’s highest melanoma incidences among youth. Adolescents often under-prioritise sun protection due to sociocultural factors, for example, the perceived attractiveness of tanned skin and a sense of invulnerability. This study reports on a curriculum-integrated sun safety intervention implemented in a single independent girls’ secondary school, a context that provides important insights but also limits generalisability to mixed-gender and other school settings. High school students (aged 13–15) participated in sun safety workshops integrated into their Personal Development, Health, and Physical Education (PDHPE) curriculum. Sun protection knowledge, attitudes toward tanning, and self-reported protective behaviours were assessed via surveys administered pre-intervention, immediately post-intervention, and five months later. Students’ sun safety knowledge increased significantly post-intervention (p < 0.001), and remained higher than baseline knowledge at 6 months, indicating substantial knowledge retention. Participants reported more frequent sunscreen use and protective behaviours after the program. The intervention also reduced the appeal of tanning, fewer students agreed that “a suntan looks good” after participating. Notably, all students regardless of ethnicity (Caucasian, Asian, Other), showed knowledge gains, but variations were observed. Students from historically underserved ethnic backgrounds started with lower baseline knowledge and exhibited slightly smaller improvements aligning with prior findings that sun awareness has differed across racial/ethnic group. These findings suggest that an engaging, curriculum-integrated sun safety program can produce immediate and sustained improvements in adolescents’ sun protection knowledge and attitudes within a girls’ secondary school context. While the single-site design limits broader generalisation, the results highlight the potential value of embedding sun safety education within school curricula. Future research should examine the effectiveness of this intervention across multiple school sites, including co-educational and diverse educational settings, to strengthen evidence for scalability and policy implementation aimed at long-term skin cancer prevention.

Direct- versus video laryngoscopy during suction assisted laryngoscopy and airway decontamination (SALAD): A randomized controlled simulation study

Abstract Pulmonary aspiration during airway management is associated with increased morbidity and mortality. First-pass success during tracheal intubation is a key determinant of patient safety and has been shown to improve with videolaryngoscopy (VL). However, comparative evidence regarding hyperangulated VL and direct laryngoscopy (DL) during massive airway contamination remains limited, particularly within a standardized Suction-Assisted Laryngoscopy and Airway Decontamination (SALAD) workflow. This study aimed to compare DL and hyperangulated VL in a standardized SALAD simulation model. In this 1:1 randomized controlled simulation study, physicians performed tracheal intubation using a standardized regurgitation model. Participants were randomized to hyperangulated VL or DL. The primary endpoint was first-pass success (FPS). Secondary endpoints included time to successful tracheal intubation and time to first successful ventilation. Two hundred physicians from various specialties were enrolled and randomized. FPS was significantly higher in the VL group than in the DL group (94.0% vs. 58.0%; OR 11.34, 95% CI 4.54–28.35; p  < 0.001). Time to successful tracheal intubation was significantly shorter in the VL group (median [IQR]: 34 s [29–40] vs. 41 s [35–48]; p  < 0.001). Likewise, time to first ventilation was significantly reduced (median [IQR]: 40 s [34–44] vs. 45 s [40–53]; p  < 0.001). In multivariable logistic regression analysis, VL remained independently associated with FPS (adjusted OR 13.42, 95% CI 5.18–34.78; p  < 0.001), whereas age, sex, professional experience, specialty, and additional qualifications were not significantly associated with the primary outcome. In this simulation model of massive airway contamination, hyperangulated videolaryngoscopy significantly improved first-pass success and reduced both intubation and ventilation times compared with direct laryngoscopy during SALAD-assisted airway management. Whether these procedural advantages translate into improved patient-centered outcomes requires confirmation in future clinical studies. Trial Registration The registration of the clinical trial was initiated prospectively on July 15, 2024, in the German Clinical Trials Register (DRKS00034683).

Understanding smart home automation acceptance through users’ lifestyles and perceived difficulty of use: Evidence from South Korea

Smart home automation (SHA) is an advanced service that automates household tasks with minimal human intervention, enhancing efficiency, convenience, and personalization. SHA serves as the critical bridge that transforms home environments into intelligent spaces that respond adaptively to user needs. Despite these benefits, SHA penetration remains limited due to various challenges, making it essential to understand SHA acceptance. This study examines two user-centered constructs influencing SHA acceptance: users’ lifestyles (UL) and perceived difficulty of use (PDU). Through a systematic process, the dimensions of both constructs were defined and measurement items were developed. A research model extending the Technology Acceptance Model was proposed, in which UL and PDU influence the perceived usefulness (PU) and intention to use (IU) of SHA. The model was tested using data collected from 167 non-users of SHA in South Korea. Results reveal that UL characterized by energy-saving negatively affects IU, whereas UL characterized by lack of time for household tasks positively affects IU. Additionally, PDU from automation configuration negatively affects both PU and IU. Based on these findings, practical strategies to enhance SHA acceptance are proposed. This study contributes by extending the acceptance framework for SHA through conceptualizing UL and PDU and by revealing how these user-centered constructs shape PU and IU, thereby clarifying their role in SHA acceptance and informing managerial strategies that promote broader penetration.

“Molecular insights into the synergistic anticancer effect of stigmasterol and cabazitaxel in colon cancer through ROS-mediated mitochondrial dysfunction and apoptosis”

A mixed-methods analysis of the implementation of a new community long-COVID service during the 2020 pandemic: Learning from practice

Introduction The rapidly increasing prevalence of long-COVID (LC), a condition characterised by multisystem complexity and high patient symptom burden, posed an immediate need to develop new clinics for assessment and management. This article reports on the rapid implementation of a reactive and responsive LC care pathway. We mapped patients’ journeys through this pathway, identifying the services that were activated according to prevalent symptoms, and used the Theoretical Domains Framework (TDF) to assess the barriers and facilitators to its implementation and delivery, from the perspective of health care professionals (HCPs) and LC patients. Methods Mixed methods study, including retrospective quantitative cross-sectional analysis of patient data and semi-structured qualitative interviews. One hundred and sixteen patients who attended the long-COVID clinic in Hertfordshire, UK, in the first 5 months of its existence and consented for their data to be analysed. Six HCPs and five patients participated in semi-structured interviews. Results Patients were referred into the service an average of 5.75 months post initial COVID-19 infection. 82% of patients required onward referral to other HCPs, most commonly pulmonary rehabilitation, chronic fatigue specialists, and a specialist COVID-19 rehab general practitioner embedded within the service. Patients reported having rehabilitation needs, moderate depression and anxiety, and difficulties performing usual activities for daily living. The TDF domains most relevant to the implementation of the LC pathway were beliefs about capabilities, environmental context and resources, knowledge, and reinforcement. Discussion Our study provides novel insight into the development of a reactive multidisciplinary care pathway. Key drivers for successful implementation of LC services were identified, such as leadership, multidisciplinary teamwork, transferable skills, and knowledge exchange. Barriers to rapid set up of the service included funding constraints and the rapid evolution of an emergency context.

Analyzing the impact of power loss reduction in radial distribution network using diverse distributed generators by utilizing ZIP load models

Abstract Distributed generators (DGs) are recent emerging technologies which are executed in distribution system (DS) to reduce power losses (PLs). However, optimal location and sizing of DG (OLSDG) remain a challenging combinatorial optimization problem due to its large search space and load uncertainties. Numerous scholars attempted to effectively handle this optimization problem characterized by a fixed load model (constant power (CP)), which fails to represent the real-world consumer behavior. This assumption causes a research gap, as practical DS is mainly dependent on characteristics of voltage magnitude and ZIP load models better represents this behavior. To overcome these challenges, this work reports a novel metaheuristic Algorithm inspired by quantum principles (Multipartite Adaptive Quantum inspired Evolutionary Algorithm (MAQiEA)) to discover the OLSDG. The originality of the work stands out by using realistic load model (ZIP load model), utilizing multipartite adaptive variation operator and analyzing the influence of diverse DGs with CP and ZIP load models. This study investigates two different scenarios to analyze the effect of various DGs on CP and ZIP load models. In first case, a study was conducted to analyze the influence of diverse DGs with CP load which are tabulated in the statistical analysis results that demonstrate the efficacy and soundness of MAQiEA. For performance assessment, proposed algorithm is compared with some well-known other metaheuristic algorithms. Tabulated results demonstrate that, MAQiEA outperforms other metaheuristic algorithms in PL reduction. Finally, the approach is suitable in terms of overall operational framework which is an important investigation in mitigating the PLs.

Determinants and predictive modelling of barriers to HIV testing uptake among adolescents and young adults in a private higher education institution in Botswana

Background HIV represents a significant public health challenge, contributing to increased mortality and morbidity in developing countries, particularly Sub-Saharan Africa. While HIV testing is crucial for early treatment and prevention, the uptake of the HIV testing among the adolescent and young adults in Botswana remains low. This study aims to predict barriers to HIV testing uptake among adolescents and young adults. Method A quantitative cross-sectional study was conducted. Data was obtained using a questionnaire employing a simple random sampling technique to collect the survey data. Statistical analysis involved descriptive statistics and multivariable logistic regression. Results A total of 353 participants were recruited. The prevalence of HIV testing uptake in the preceding 12 months was (64.9%; n = 229). Determinants associated with non-uptake of HIV testing included participants in casual relationships (OR 1.776,95%CI 0.998–3.160, P < 0.049), transactional relationships (AOR 1.098, 95% CI 0.189–6.382) compared with single participants. Additional determinants included residence beyond 5 km from HIV testing centers (OR 1.825, 95% CI 1.074–3.103, P < 0.026), and cohabiting (AOR 3.45, 95% CI 0.271–43.979). Religious affiliation was also predictive, with Christians (AOR 2.347, 95% CI 0.188–29.295), Muslims (AOR 1.765, 95% CI 0.046–59.506), and adherents of African traditional religions (AOR 1.718, 95% CI 0.084–35.004) exhibiting higher odds of non-testing compared with non-believers. Participants with negative attitudes toward HIV testing were 69% less likely to forgo HIV testing (OR 0.310, 95% CI 0.1704–3.103, P = 0.026) than those with positive attitudes. Conclusion This study provides critical evidence on disparities in HIV testing among adolescents and young adults in Botswana, highlighting persistent gaps in access and utilization. These findings underscore the necessity for context-specific strategies that mitigate access barriers and address behavioral, structural, and socio-cultural determinants to optimize HIV testing uptake, advancing progress toward the UNAIDS 2030 target of zero new infections.

LCNet: balancing representation capacity and computational cost for concrete crack detection in complex backgrounds

Artificial intelligence-based prediction of diseases among homeless populations in Bogotá: Implications for targeted interventions

Background The homeless population in Bogotá exhibits complex social and health vulnerabilities, with a high prevalence of chronic and communicable diseases such as hypertension, diabetes, tuberculosis, HIV/AIDS, and cancer. These conditions hinder epidemiological surveillance and timely public health decision-making, particularly in contexts of social exclusion. Objective To develop and evaluate an artificial intelligence-based predictive model aimed at identifying disease occurrence risk profiles among the homeless population in Bogotá, in order to support public health decision-making. Methods Data from the 2024 Bogotá Homeless Census were analyzed, comprising 10 478 records and 46 demographic, clinical, and socioeconomic variables. Data processing and model development followed the CRISP-DM methodology. The extreme gradient boosting (XGBoost) algorithm was implemented to predict disease occurrence. Model performance was evaluated in terms of accuracy, sensitivity, and the F1-score, while interpretability was assessed through SHAP (SHapley Additive exPlanations) values. Additionally, metrics related to social trust, system response time, and potential health impacts were examined. Results The model achieved an accuracy of 0.91, a sensitivity of 0.57, and an F1-score of 0.70, striking an adequate balance between identifying high-risk individuals and reducing false positives. SHAP analysis identified hypertension, diabetes, and HIV/AIDS as the main predictors of classification. However, complementary metrics revealed limitations in social trust (TAS = 0.49), system response time (SRT = 24.86 hours), and potential health impact (PHIS = 0.24). Conclusions and implications These findings suggest that explainable artificial intelligence (XAI) models may support public health surveillance and intervention prioritization in homeless populations by identifying epidemiological risk profiles. However, the models’ applicability remains context-specific and requires external validation before implementation across other vulnerable populations or healthcare settings.

Curcumin modulates oxidative stress, cortical activity and corticosterone in mice exposed to chronic social stress in the spontaneous aggressiveness model

Abstract Stress constitutes a fundamental physiological mechanism essential for survival and environmental adaptation. Chronic social stress is a major risk factor for emotional and behavioral disturbances, including aggressiveness. Curcumin, a bioactive polyphenol derived from Curcuma longa, exhibits antioxidants, anti-inflammatory, and antidepressant properties, making it a promising neuroprotective agent. Previous studies using the Spontaneous Aggressiveness Model (MSA) demonstrated increased production of reactive oxygen species (ROS) and c-Fos expression in aggressive mice, suggesting a role for oxidative stress. This study investigated the effects of curcumin supplementation on neuroendocrine and oxidative parameters in male Swiss Webster mice subjected to chronic social stress in MSA. Animals ( n  = 60, 3 weeks old) were allocated to control and curcumin-treated groups. The treated group received a 1.5% curcumin-enriched diet for ten days before social regrouping in week 10. Behavioral categorization into non-regrouped (NR), aggressive (AgR), subordinate (Sub), and harmonious (Har) profiles were performed at week 16. Plasma corticosterone was quantified by ELISA, ROS production by DHE assay, and c-Fos expression by immunofluorescence. All procedures were approved by the Ethics Committee on Animal Use of the Oswaldo Cruz Institute (CEUA/IOC L-022/2025). Curcumin supplementation reduced attack intensity, prevented hypercortisolemia in NR mice, and decreased ROS levels in aggressive animals, supporting its antioxidant and neuroendocrine modulatory effects as a potential neuroprotective strategy under chronic social stress.