Conectar
by Md. Masud Reza, Hasibul Hasan Shanto, Samira Dishti Irfan, A. K. M. Masud Rana, Mohammad Niaz Morshed Khan, Golam Sarwar, Mohammad Sha Al Imran, Mahbubur Rahman, Md. Safiullah Sarker, Muntasir Alam, Md. Abu Hena Chowdhury, Mustafizur Rahman, Sharful Islam Khan
BackgroundUnintended pregnancy among female sex workers (FSW) is a pressing reproductive health concern attributable to risky sexual behaviors, healthcare inequities and poor negotiation powers with male sex partners. However, evidence is scarce on the prevalence and determinants of unintended pregnancies among FSW, which is crucial for enhancing reproductive healthcare. This analysis aims to measure the prevalence of lifetime unintended pregnancies and their associated factors.
MethodsA cross-sectional study was conducted on 327 FSW in Jashore (a border belt district of Bangladesh) from September 2022 to March 2023. Participants were recruited through take-all sampling. Data were collected on the lifetime history of unintended pregnancies and other relevant variables through face-to-face interviews. Chi-square statistic was used to compare the characteristics of FSW reporting unintended pregnancies. To assess the net association of factors associated with unintended pregnancy, multiple logistic regression was applied.
ResultThe lifetime prevalence of unintended pregnancies was reported at 75.8% (95%CI: 71.0–80.1). Among those who reported unintended pregnancies, 37.1% (95%CI: 30.8–43.8) had no education, 39.9% (95%CI: 32.8–47.5) were 25–34 years old, 49.6% (95%CI: 39.3–59.9) were currently married and 62.9% (95%CI: 49.7–74.4) earned ≤10,000 BDT per month compared to those who did not report lifetime unintended pregnancies. The likelihood of unintended pregnancies was significantly higher among those who reported having sex with non-transactional male sex partners (AOR: 2.4, 95%CI: 1.1–5.3, p = 0.036) than those who never had sex with any non-transactional male sex partner. The likelihood was also higher among those who reported rape in their lifetime (AOR: 2.0, 95%CI: 1.0–3.8, p = 0.037) and who self-reported mental health problems (AOR: 2.1, 95%CI: 1.0–4.2, p = 0.045) within the past year, compared to their counterparts.
ConclusionThis study highlights the considerable prevalence and associated determinants of unintended pregnancies among FSW in Jashore. These determinants need to be considered to strengthen reproductive healthcare interventions and policies for FSW. Reproductive health of FSW cannot be improved unless these factors are addressed in the ongoing interventions.
by Nguyen Thien Duc, Nguyen An Ninh, Nguyen Phi Trinh, Le Quang Tuyen, Nguyen Van Hung, Dinh Hoang Khanh, Nguyen Van Luat, Nguyen Huu Phuc Dai, Tran Duc Huy, Chu Duc Hoa, Tran Vuong The Vinh
PurposesAnatomy is fundamental in medical education, yet cadaveric dissection faces challenges including limited specimens, high costs, and chemical hazards. Interactive anatomy tables such as the Pirogov system offer innovative alternatives, but evidence from Southeast Asia is limited.
MethodsIn a prospective cohort, 188 medical students (139 in Y1 and 49 in Y2) were randomly assigned to the Pirogov table group (Group A, n = 99) or the cadaveric dissection group (Group B, n = 89). Knowledge acquisition was measured using a validated 20-item multiple-choice test before and after the intervention. Student perceptions were evaluated with a 10-item Likert-scale questionnaire covering four domains: knowledge and understanding, spatial visualization and relationships, learning experience and engagement, and effectiveness and practical value. Data were analyzed using paired and independent t-tests and Welch’s t-test.
ResultsBoth groups showed significant knowledge gains (Group A: 4.3 ± 1.65 to 5.2 ± 1.75, p Conclusion
The Pirogov table and cadaveric dissection were associated with similar short-term improvements in anatomy knowledge. Students valued the Pirogov table for visualization and engagement. These findings support integrating digital tools with cadaveric dissection to enhance anatomy education, particularly in resource-limited contexts.
by Nilavro Das Kabya, MD Shaifullah Sharafat, Rahimul Islam Emu, Mehrab Karim Opee, Riasat Khan
Malabar spinach is a nutrient-dense leafy vegetable widely cultivated and consumed in Bangladesh. Its productivity is often compromised by Alternaria leaf spot and straw mite infestations. This work proposes an efficient and interpretable deep learning framework for automatic Malabar spinach leaf disease classification. A curated dataset of Malabar spinach images collected from Habiganj Agricultural University and supplemented with public samples was categorized into three classes: Alternaria, straw mite, and healthy leaves. A lightweight SpinachCNN established a strong baseline, while Spinach-ResSENet, enhanced with squeeze-and-excitation modules, improved channel-wise attention and feature discrimination. A customized Vision Transformer (SpinachViT) and SwinV2-Base were further investigated to assess the benefits of transformer-based architectures under limited data. To mitigate annotation scarcity, we employed SimSiam-based self-supervised pretraining on unlabeled images, followed by supervised fine-tuning with cross-entropy or a hybrid objective combining cross-entropy and supervised contrastive loss. The best-performing domain-optimized model, SimSiam-CBAM-ResNet-50, incorporated Convolutional Block Attention Modules and achieved 97.31% test accuracy, 0.9983 macro ROC-AUC, and low calibration error, while maintaining robustness to Gaussian and salt-and-pepper noise. Although a SwinV2-Base benchmark pretrained on ImageNet-22k reached slightly higher accuracy (97.98%, 98.99% with test-time augmentation), its 86.9M parameters and reliance on large-scale pretraining reduce feasibility for edge deployment. In contrast, the SimSiam-CBAM model offers a more parameter-efficient and deployment-friendly solution for real-world agricultural applications. Model decisions are interpretable via Grad-CAM, Grad-CAM++, and LayerCAM, which consistently highlight biologically relevant lesion regions. The spinach dataset used in this study is publicly available on: https://huggingface.co/datasets/saifullah03/malabar_spinach_leaf_disease_dataset.by Yeahyea Ahmed, Md Abdullah Saeed Khan, Laila Afroz, Mohammad Nurunnabi, Md Golam Abbas
BackgroundStreet adolescents often engage in early sexual activity, have multiple partners, and are at high risk of sexual abuse and exploitation. Despite the significance of this issue, there is a critical gap in understanding the sexual and reproductive health (SRH) needs, practices, and challenges of this marginalized population in Bangladesh, which this study aimed to explore.
MethodsA cross-sectional study was conducted from August to December 2023, involving 311 street adolescents aged 16–19 years in Sylhet City Corporation, Bangladesh. Data were collected through face-to-face interviews using a semi-structured questionnaire. The questionnaire covered sociodemographic characteristics, pubertal changes, SRH status, and SRH-seeking behaviors.
ResultsOf all participants, 62.8% were males and 37.2% were females, with a mean age of approximately 17.3 years for both sexes. Sexual intercourse was reported by 32.56% of participants, with a significant gender disparity (76.79% females vs. 6.35% males, p Conclusion
Street adolescents in Sylhet City face severe SRH challenges, including high rates of sexual abuse, low contraceptive use, and limited STD knowledge, with significant gender disparities, which should be addressed through appropriate and urgent interventions.
by Sarah L. Brown, Barry J. McDonnell, David McRae, Paul Angel, Imtiaz Khan, Rhiannon Phillips, Britt Hallingberg, Delyth H. James
Using visualisation to conceptualise a chronic condition can encourage accurate illness beliefs and support treatment adherence. Hi-BP is a digital visual intervention to support adherence to antihypertensive medication, co-produced with patients. The aim of this study was to investigate the feasibility and acceptability of Hi-BP and explore the preliminary direction of effects on illness and treatment beliefs, medication adherence and blood pressure (BP). A two-phased mixed-methods non-randomised feasibility study was conducted from April 2021 to March 2022 in eight community pharmacies across one Health Board in South-East Wales, UK. Hi-BP was delivered as a single researcher-led consultation to 69 patients in Phase 1 and by pharmacists to three patients in Phase 2. Feasibility was determined using predefined criteria, with acceptability explored qualitatively using semi-structured interviews. Quantitative outcome measures (illness perceptions, medication beliefs, medication-adherence, prescription dispensing and collection data, BP) were recorded at baseline and immediately post-intervention.Follow-up outcome measures were collected at two-weeks (medication-adherence) and three-months (all baseline measures). Hi-BP met feasibility criteria for pharmacist recruitment in both phases, and patient recruitment in Phase 1, but not Phase 2. Hi-BP was acceptable to the sub-sample of 15 patient participants interviewed in Phase 1; insufficient data were available to determine patient acceptability at Phase 2. Hi-BP was acceptable to pharmacists in Phase 1 and partially acceptable at Phase 2, due to competing demands on time for intervention delivery. All outcome measures were considered feasible for use, though a ceiling effect was noted for medication adherence. A potentially positive directional effect was found for illness perceptions (X2(2)=10.83,n=54,p=0.004), medication beliefs (BMQ-Necessity (X2(2)=11.71,n=54,p=0.003) and BP (Systolic BP Z=-3.91,n=51,p=2(2)= 2.4,n=45,p=0.299). In the Community Pharmacy setting, Hi-BP was well-accepted and has the potential for significant reductions in BP; however, further research is needed to explore pharmacist capacity to support implementation.by Lauren Riehm, Keean Nanji, Moiz Lakhani, Evelina Pankiv, Dean Hasanee, Wesla Pfeifer
PurposeLarge language models (LLMs) have the potential to change medical education. Whether LLMs can generate multiple-choice questions (MCQs) that are of similar quality to those created by humans is unclear. This investigation assessed the quality of MCQs generated by LLMs compared to humans.
MethodsThis review was registered with PROSPERO (CRD42025608775). A systematic review and frequentist random-effects network meta-analysis (NMA) or pairwise meta-analysis was performed. Ovid MEDLINE, Ovid EMBASE, and Scopus were searched from inception to November 1, 2024. The quality of MCQs was assessed with seven pre-defined outcomes: question relevance, clarity, accuracy/correctness; distractor quality; item difficulty analysis; and item discrimination analysis (point biserial correlation and item discrimination index). Continuous data were transformed to a 10-point scale to facilitate statistical analysis and reported as mean differences (MD). The MERSQI and the Grade of Recommendations, Assessment, Development and Evaluation (GRADE) NMA guidelines were used to assess risk of bias and certainty of evidence assessments.
ResultsFive LLMs were included. NMA demonstrated that ChatGPT 4 generated similar quality MCQs to humans with regards to question relevance (MD −0.13; 95% CI: −0.44,0.18; GRADE: VERY LOW), question clarity (MD −0.03; 95% CI: −0.15,0.10; GRADE: VERY LOW), and distractor quality (MD −0.10; 95% CI: −0.24,0.04; GRADE: VERY LOW); however, MCQs generated by Llama 2 performed worse than humans with regards to question clarity (MD −1.21; 95% CI: −1.60,-0.82; GRADE: VERY LOW) and distractor quality (MD −1.50; 95% CI: −2.03,-0.97; GRADE: VERY LOW). Exploratory post-hoc t-tests demonstrated that ChatGPT 3.5 performed worse than Llama 2 and ChatGPT 4 with regards to question clarity and distractor quality (p Conclusion
ChatGPT 4 may create similar quality MCQs to humans, whereas ChatGPT 3.5 and Llama 2 may be of worse quality. Further studies that directly compare these LLMs to human-generated questions and administer MCQs to students are required.
by Moe Thi Thi Han, Tay Zar Myo Oo, Busayamas Chewaskulyong, Sakorn Pornprasert, Kanyamas Choocheep, Khanittha Punturee, Warunee Kumsaiyai, Yupanun Wuttiin, Sawitree Chiampanichayakul, Ratchada Cressey
Non-smoking-related lung cancer is increasingly associated with environmental factors such as particulate matter (PM) exposure. Using deep small RNA sequencing, we identified distinct miRNA expression patterns in lung cancer patients compared to non-cancer controls, stratified by smoking status. Notably, hsa-miR-125b-5p and hsa-miR-100-5p were significantly downregulated in non-smoking lung cancer patients. Pathway enrichment analysis revealed smoking amplifies pathways related to glycan biosynthesis, signal transduction, and transcriptional regulation, while non-smoking lung cancer is characterized by immune dysfunction and metabolic alterations, including oxidative phosphorylation and natural killer cell cytotoxicity. Validation in a larger cohort using quantitative RT-PCR confirmed the suppression of miR-125b-5p and miR-100-5p in non-smoking lung cancer patients. Additionally, miR-203a and miR-199a-3p were identified as potential biomarkers for lung cancer, independent of smoking status. Chronic PM exposure in primary bronchial/tracheal epithelial cells initially elevated miR-125b-5p and miR-100-5p expression, but prolonged exposure suppressed these miRNAs while increasing their target genes, TXNRD1 and HOXA1, suggesting stress-induced dysregulation. Functional studies using miRNA mimics demonstrated that miR-125b-5p and miR-100-5p suppress PM-induced cancer cell mobility and colony formation, with miR-125b-5p exhibiting broader effects. These findings underscore the critical roles of miR-125b-5p and miR-100-5p in PM-associated lung cancer progression and their potential as biomarkers and therapeutic targets. This study highlights distinct mechanisms of lung carcinogenesis in smokers and non-smokers, providing a foundation for targeted interventions in PM-associated lung cancer.by Mohamadreza Hatefi, Seyedeh Feriyal Mahdavi, Amirreza Abbasi, Farideh Babakhani
BackgroundUpper extremity (UE) dominance is often defined by self-reported hand preference; however, this may not accurately reflect true functional or strength-based dominance. This study examined the relationship between writing hand, throwing hand, and maximal grip strength to assess how these measures align.
MethodsThirty-four healthy, recreationally active college-aged individuals reported their preferred writing and throwing hands and completed standardized grip strength testing. Associations among the variables were analyzed using Phi coefficients and chi-square tests.
ResultsA moderate, significant relationship was found between writing and throwing hand preference (φ = 0.456; p = 0.008), indicating general consistency across these subjective measures. However, no significant association emerged between grip strength dominance and either writing (φ = 0.027; p = 0.876) or throwing hand (φ = 0.096; p = 0.574).
ConclusionThese results suggest that grip strength dominance may not correlate with commonly used indicators of hand preference, highlighting the need for task-specific definitions of dominance in clinical and athletic contexts. Consequently, employing such task-specific definitions allows for more accurate assessments and enhances the translational relevance of research findings in practical settings.
by Elora Sharmin, Ajmain Ishaat Khan, Sheikh Foyez Ahmed
Understanding the safety and efficacy of COVID-19 vaccines in pregnant women and their neonates is crucial for understanding maternal and fetal outcomes, particularly the extent of passive immunity against SARS-CoV-2 which can be imparted to the neonates. The purpose of this study was to evaluate the transplacental transfer of maternal SARS-CoV-2 IgG antibodies against the spike (S) and nucleocapsid (N) proteins to neonates and understand whether factors like maternal comorbidities, gestational weeks, and neonatal birth weight have an influence on placental transfer ratios (PTR). A total of 57 pregnant women were assessed for SARS-CoV-2-specific IgG antibodies at delivery, and corresponding antibody titers were also measured in their neonates immediately after delivery. The PTRs for anti-S and anti-N IgG were calculated, and statistical analyses were performed for identifying potential influencing factors. The mean PTR for anti-S IgG was 1.38, suggesting effective placental transfer, whereas anti-N IgG had a lower PTR of 1.13, indicating limited transfer. A strong positive correlation was observed between maternal and neonatal anti-S IgG (r = 0.558, p
FreshRSS