Browse Articles

Discover research articles across all indexed journals

Down-staging depth score could be a survival predictor for locally advanced breast cancer patients after neoadjuvant chemotherapy

Scientific Reports Miaoli Zheng, Yu Wang, Na Jing et al. Jun 30, 2026 DOI: 10.1038/s41598-026-56400-0

AI-driven molecular diversification and ligand-based optimization of macitentan derivatives targeting VEGFR1 and endothelin signaling pathways

PLoS ONE Sümeyra Koç Sahin, Nouman Ali Jun 30, 2026 DOI: 10.1371/journal.pone.0352451

Preeclampsia is a hypertensive disorder of pregnancy marked by angiogenic imbalance and endothelial dysfunction, driven largely by over expression of soluble fms-like tyrosine kinase-1 (sFlt-1) and up-regulation of endothelin-1 (ET-1). Targeting both VEGFR1 and ET-1 receptor could offer a dual-action therapeutic approach. In this study, macitentan, a dual ETA/ ETB antagonist, was used as a scaffold for AI-assisted derivative generation aimed at dual inhibition of these receptors. A comprehensive computational pipeline was applied including molecular docking, pharmacophore modeling, molecular dynamics (MD) simulation, DFT, MM/GBSA, and ADMET analysis. Derivative 24 showed high binding affinity toward VEGFR1 (–7.7 kcal/mol), while Derivative 15 exhibited superior interaction with ET-1 receptor (–9.2 kcal/mol), both outperforming macitentan. MD simulations over 500 ns confirmed complex stability with RMSD stabilization around 0.2–0.3 nm and consistent hydrogen bonds. MM/GBSA binding energies further supported strong receptor interactions (–17.15 kcal/mol for Derivative 24–VEGFR1; –28.72 kcal/mol for Derivative 15–ET-1 receptor). DFT results showed reduced HOMO–LUMO gaps (3.55 eV for Derivative 24; 3.30 eV for Derivative 15), while MEP analysis indicated favorable electrostatic potential for target interaction. Pharmacophore and ADMET profiling revealed improved drug-likeness, high GI absorption, and reduced predicted toxicity. This study presents Derivatives 15 and 24 as promising dual-target leads against preeclampsia-induced endothelial dysfunction. However, experimental validation remains necessary to confirm efficacy and safety before translational application.

Structural-budgeted QUBO learning of Bayesian networks with spectral and credibility diagnostics

Scientific Reports Po-Chien Huang, Vittaldas Prabhu Jun 30, 2026 DOI: 10.1038/s41598-026-58207-5

Invaders taking over—Mollusc faunal change in volcanic barrier lakes of the Albertine Rift biodiversity hotspot

PLoS ONE Francis Ssenkuba, Marcellin Rwibutso, Marie Claire Dusabe et al. Jun 30, 2026 DOI: 10.1371/journal.pone.0352648

East African freshwater are renowned biodiversity hotspots, particularly within the Rift Valley lakes, which exhibit exceptional species richness and endemism. However, research in this region has predominantly focused on large Rift lakes, with smaller volcanic barrier lakes remaining underexplored. This study integrated morphological, DNA barcoding, and ecological analyses to assess mollusc diversity, biogeographical affinities, environmental correlates of distribution, and community differences in lakes Bunyonyi, Mutanda, and Mulehe in Uganda, and Lakes Ruhondo and Burera in Rwanda. Results revealed a relatively high malacofaunal diversity in these lakes compared to some Great Lakes in the region, although these comparisons should be interpreted cautiously because sampling scope differs among studies. Assemblages largely exhibited Nilotic affinities and were dominated by widespread mollusc species. Our findings suggest ongoing homogenisation of mollusc communities, as indicated by reduced dispersion among invaded lakes. While this pattern is consistent with potential influences of anthropogenic activities and invasive species such as the North American crayfish ( Procambarus clarkii ), Physella acuta , and the Asian lineage of Melanoides tuberculata , our analyses do not allow us to disentangle invasion effects from other confounding factors, including lake-specific characteristics and environmental gradients . Therefore, invasion-driven homogenisation should be interpreted with caution. Nevertheless, this trend, coupled with habitat degradation from agriculture, pollution, and infrastructure development, may pose significant threats to the mollusc diversity in these lakes. This research underscores the importance of prioritising these overlooked, fragile ecosystems in the regional biodiversity conservation strategies to safeguard habitats for molluscs and other faunal elements. A comprehensive regional freshwater survey, including metabarcoding, is needed to document molluscan diversity patterns under ongoing environmental change and increasing anthropogenic pressure.

Introducing Actigraphic Data Analyzer (ADA), an open source software for sleep/wake scoring and circadian rhythm analysis with graphical interface

Scientific Reports Piotr Biegański, Monika Tutaj, Anna Duszyk-Bogorodzka et al. Jun 30, 2026 DOI: 10.1038/s41598-026-59007-7

Abstract We introduce Actigraphic Data Analyser (ADA), a user-friendly open source Python package with a graphical user interface (GUI). It reads raw data from GENEActiv and ActiGraph devices, and the MESA dataset, supports multiple methods for collapsing data into epochs, provides sleep/wake scoring via several classical algorithms and the recently published Universal Filter Approach, offers comprehensive circadian rhythms analysis, and outputs common sleep plots and metrics like daily profiles, sleep onset, sleep fragmentation index and more. Apart from the interactive GUI, ADA can be used as a Python library. As an example of its application, we present a simple Python script revealing correlations between different descriptors of circadian rhythms: DFA, variants of cosinor, AR model spectrum, IS, IV, M10 and L5. Correlation matrix produced by this script (a figure included also in this paper) applied to a dataset of 87 weekly recordings (created for this study and freely available) reveals division of these commonly applied descriptors into two groups: one related mostly to the (length of) circadian period, and the other to the strength of the 24-h rhythm. Both the above-mentioned ADA package and set of weekly actigraphic recordings are freely available on GPLv3 and CC-BY licences, respectively.

Dosimetric characterization of CdZnTe radiation detectors under electron-beam irradiation

PLoS ONE Chansun Park, Jiwon Seo, Sangsu Kim et al. Jun 30, 2026 DOI: 10.1371/journal.pone.0349698

With the growing complexity of modern radiotherapy, the demand for dosimeters capable of delivering real-time, high-resolution, and reliable measurements has grown significantly. Although cadmium–zinc–tellurite (CZT) detectors have emerged as potential solutions to this growing demand, the lack of comprehensive characterization data for these detectors deters their application in clinical electron-beam dosimetry, especially for high-precision tasks such as small-field and point-dose dosimetry. In this study, we systematically characterized CZT detectors, emphasizing the performance trade-offs associated with detector thickness for therapeutic electron-beam dosimetry. The performance of two CZT detectors (6 × 6 mm²), with thicknesses of 5 and 8 mm, was evaluated using a clinical linear accelerator (LINAC) over an electron energy range of 6–18 MeV. Both detectors demonstrated stable baseline performance, including proportional dose linearity (R² > 0.997), dose-rate independence, and excellent long-term stability (<1.3% variation over two months). Furthermore, a distinct thickness-dependent trade-off was identified: while both configurations exhibited highly comparable dose responses at the 6-MeV baseline, the 8-mm detector exhibited better energy capture efficiency for higher-energy electrons. By contrast, the 5‑mm detector offered enhanced angular isotropy with respect to the gantry angle as well as reduced signal variation (±3.5% vs. ± 6.4%), emerging as more suitable for rotational therapies and geometrically demanding scenarios. Additionally, a nonmonotonic energy dependence was observed, characterized by a pronounced response peak at 15 MeV, which was likely attributable to bremsstrahlung photon contamination from the LINAC head. These findings demonstrate that detector thickness serves as an application-specific optimization factor for sensitivity and angular fidelity, offering a practical framework for deploying CZT detectors in both initial clinical evaluations and challenging extra-clinical radiation environments.

A spatially reconfigured polarization converting metasurface for enhanced radar cross section reduction

Scientific Reports Nandhitha Pauly, Juhaina Razack, M. Gopikrishna Jun 30, 2026 DOI: 10.1038/s41598-026-58502-1

Effects of transcutaneous auricular vagus nerve stimulation combined with repetitive facilitative exercise on lower limb motor function in patients with intracerebral hemorrhage: A pilot randomized controlled trial

PLoS ONE Hanji Chen, Hanbo Chen, Chongrui Feng et al. Jun 30, 2026 DOI: 10.1371/journal.pone.0352321

Objective To evaluate the clinical efficacy of transcutaneous auricular vagus nerve stimulation (taVNS) combined with repetitive facilitative exercise (RFE) in improving lower limb function in patients with intracerebral hemorrhage. Methods The research protocol was registered on the Chinese Clinical Trial Registry website (ChiCTR2500106064). Sixty patients with intracerebral hemorrhage were randomly assigned to either the experimental group (n = 30) or the control group (n = 30). The experimental group received real taVNS combined with RFE, while the control group received sham taVNS combined with RFE. Before and after 6 weeks of treatment, functional assessments were conducted using the Fugl-Meyer Assessment-Lower Extremities (FMA-LE), the Functional Ambulation Category (FAC) scale, and a three-dimensional gait analysis system to obtain kinematic parameters of the pelvis during the walking cycle. Results Baseline demographic and clinical characteristics showed no significant differences between the two groups (all P > 0.05). After 6 weeks of intervention, both groups exhibited significant improvements in FMA-LE, FAC, stride length, peak hip extension angle, and peak hip joint moments (all P < 0.05). However, the experimental group demonstrated superior outcomes compared to the control group in FMA-LE (t = 3.233, P = 0.002), FAC (t = 2.868, P = 0.006), stride length (t = 3.077, P = 0.003), and peak hip extension angle (t = 2.682, P = 0.010). No significant difference was observed in peak hip joint moments (t = −0.212, P = 0.833) between the two groups after treatment. No adverse effects or dropouts were reported during the study. Conclusion Compared to sham taVNS combined with RFE, taVNS combined with RFE can significantly enhances lower limb motor function in patients with post-stroke hemiplegia following intracerebral hemorrhage. The research protocol has been registered on the Chinese Clinical Trial Registry website ( ChiCTR2500106064 ).

A comprehensive urine workflow enables robust methylation and fragmentation analysis of cell-free DNA

Scientific Reports Florence Mauger, Nouara Oussada, Caroline Horgues et al. Jun 30, 2026 DOI: 10.1038/s41598-026-60021-y

Impact of kidney function on the metabolome in the general population

PLoS ONE Johan Ljunggren, Mario Delgado-Velandia, Johan Sundström et al. Jun 30, 2026 DOI: 10.1371/journal.pone.0347652

Background A reduction in kidney function cause metabolites that normally are cleared by the kidney to accumulate. These accumulated metabolites could potentially link kidney function to an increased risk of cardiovascular disease. Aim To assess the relationships between creatinine-based estimated glomerular filtration rate (eGFR) and the metabolome (791 endogenous metabolites) in the general population, as well as the cortisol to cortisone ratio. Patients and methods Three population-based cohorts, Epihealth, PIVUS and POEM (total n = 3,444), were used with a discovery/validation approach. Metabolomics was measured by mass spectroscopy. eGFR was calculated using the 2021 CKD-EPI creatinine-based equation. Multivariable linear regression was used to assess the correlations between each metabolite concentration and eGFR, adjusting for sex, % fat mass, cardiovascular risk factors and medications. The metabolites significant in the linear regression were analyzed using MultiVariable Mendelian Randomization (MVMR), adjusting for diabetes and BMI. Results Nighty-six metabolites were significantly associated to eGFR and had the same direction of association in both linear regression and MVMR. The vast majority of these associations were negative. Apart from creatinine, N-acetylalanine, N,N-dimethyl-pro-pro, N,N,N-trimethyl-alanylproline betaine (TMAP) were the top findings. Pathway enrichment analysis (Metaboloanalyst 6.0) found five significantly enriched pathways, two of which involved amino acid metabolism. Conclusions In a general population with only mild to moderately reduced kidney function, several metabolites were associated with kidney function, including the cortisol to cortisone ratio. This finding is of interest since several of these metabolites might link reduced kidney function to an increased risk of cardiovascular disease.

The effect of amniotic membrane application on post-cesarean wound healing and cosmetic outcomes

Scientific Reports Yusuf Başkıran, Kazım Uçkan, İzzet Çeleğen et al. Jun 30, 2026 DOI: 10.1038/s41598-025-23623-6

Abstract Post-cesarean wound complications—such as surgical site infection (SSI), dehiscence, and hypertrophic scarring—remain a challenge in obstetric surgery. We evaluated whether intraoperative autologous amniotic membrane application improves postoperative outcomes. In a single-blind, randomized controlled trial, 372 women undergoing elective cesarean section were allocated to an amniotic membrane group ( n  = 174) or control group ( n  = 198). SSI incidence was lower with amniotic membrane (4.3% vs. 10.8%; p  < 0.05), and wound dehiscence was reduced (2.1% vs. 6.4%; p  < 0.05). Postoperative pain scores (Visual Analog Scale) were significantly lower in the amniotic membrane group on postoperative day 1 (3.38 ± 1.36 vs. 4.05 ± 1.53; p  = 0.01) and day 2 (1.25 ± 0.71 vs. 2.01 ± 0.98; p  = 0.001). Clinician-rated cosmetic appearance (Modified Hollander Wound Evaluation Scale) did not differ between groups ( p  = 0.863), whereas patient-reported satisfaction was higher with amniotic membrane ( p  = 0.001). No predefined adverse events were observed. These findings suggest that amniotic membrane application may reduce wound-related complications and postoperative pain and improve patient satisfaction after cesarean delivery; confirmatory studies in broader populations are warranted.

Time to maternal death and its predictors among obstetric ICU patients in a resource-limited setting: A 10-year survival analysis

PLoS ONE Tigist Nega Alemu, Wondimagegn Genaneh Shiferaw, Workneh Elias Orsongo et al. Jun 30, 2026 DOI: 10.1371/journal.pone.0352904

Background Maternal mortality remains a major concern in resource-limited settings, particularly among critically ill obstetric patients requiring intensive care unit (ICU) admission. Objective To assess the time to maternal death and identify its predictors among obstetric patients admitted to the intensive care unit (ICU) in a resource-limited tertiary hospital in southern Ethiopia, over a ten-year period. Methods A retrospective cohort study was conducted among 378 obstetric patients admitted to the ICU between 2014 and 2023. Kaplan-Meier analysis estimated survival probability, multivariable Cox proportional hazards regression identified independent predictors of time to maternal death reported as adjusted hazard ratios (AHR) with 95% confidence intervals, and a Fine-Gray competing-risks model was additionally conducted with discharge alive as the competing event. Results Of 378 obstetric ICU admissions, 126 resulted in maternal death, with a median time to death of 2.71 days (95% CI: 2.13–3.44); 71.4% of deaths occurred within the first five days, and survival probability declined rapidly before stabilizing after ten days. Rural residency (AHR 1.56, 95% CI 1.03–2.37), shock (AHR 2.27, 95% CI 1.44–3.56), multi-organ failure (AHR 1.75, 95% CI 1.09–2.78), mechanical ventilation (AHR 1.82, 95% CI 1.11–2.98), and impaired consciousness (moderate GCS: AHR 2.94; severe GCS: AHR 5.43, both p < 0.001) were independently associated with higher hazard of death, with findings consistent across the Fine–Gray competing-risks model. Conclusion Maternal ICU deaths occurred early, with most fatalities within the first week of admission and survival probability declining sharply in the first ten days. Shock, multi-organ failure, invasive mechanical ventilation, impaired consciousness at admission, and rural residency were independent predictors of shortened time to maternal death. Interventions targeting these conditions must be initiated early and urgently, particularly within the critical first days of ICU admission, to improve survival in resource-limited settings.

A hybrid DenseNet121-random vector functional link (RVFL) approach for plant leaf classification

Scientific Reports Upendra Mishra, Himanshi Chaudhary Jun 30, 2026 DOI: 10.1038/s41598-026-58360-x

Abstract The Random Vector Functional Link (RVFL) network provides an efficient and quick method of training feedforward single hidden neural networks. It solves major disadvantages of classical neural networks, i.e. dealing with slow convergence and overfitting by exploiting fixed random weights and closed-form output computation. DenseNet121 architecture has huge potential in the artificial intelligence field and more specifically in object recognition tasks. The major drawback, however, is its inability to provide insights into its intermediate levels of analysis of classification, which is why this algorithm offers limited flexibility to analyze the training procedure. To resolve this problem, we offer a combined solution by integrating DenseNet121 with RVFL. Within the framework described, deep features are obtained in relation to the input leaf pictures with the pre-trained DenseNet121 model that captures important texture and contour structure. To, firstly, minimize the dimension of these features and, secondly, remove superfluous redundancies, Principal Component Analysis (PCA) is used, and only the most informative components are retained. This smaller feature set is then marked to the RVFL classification, whose responsibility is a final classification. Understanding how the proposed DenseNet121-RVFL approach performs, we have conducted some comparative experiments against several baseline classifiers, such as DenseNet121- Support Vector Machine, DesnseNet121-Twin Support Vector Machine, DenseNet121-Extreme Learning Machine, DenseNet121 and DenseNet 121-Kernel Ridge Regression. The findings of the experiment reveal that the hybrid DenseNet121-RVFL model has the best effect in comparison to the other methods, as it provided the highest accuracy of 94.45, F1-Score is 0.955, Geometeric Mean(G-Mean) of 0.898 and Area Under Curve (AUC) of 0.961 on the test dataset.

Characterization of entomological drivers of malaria transmission in five villages, Keerom Regency, Papua, Indonesia

PLoS ONE Ismail E. Rozi, Lepa Syahrani, Dendi H. Permana et al. Jun 30, 2026 DOI: 10.1371/journal.pone.0339951

Background Keerom Regency remains one of the high-malaria-endemic regencies in Indonesia. Despite the accelerated malaria elimination strategy underway in Papua, malaria transmission in this regency has not declined. To characterize the entomological and human behavioral factors sustaining and driving malaria transmission, a rapid entomological assessment paired with human behavior observations (HBOs) and household surveys was conducted in five villages of Keerom Regency. Methods Entomological surveys were conducted on three occasions in 2022 and 2023. Human landing catches (HLC), night indoor resting collection, and mosquito larval site surveys were conducted alongside human behavior observations (HBO) and structured household surveys. Results Six species of Anopheles were identified including Anopheles koliensis , An. punctulatus , An. hinesorum , An. kochi , An. bancroftii and An. peditaeniatus . The dominant species were An. koliensis and An. punctulatus . Outdoor Anopheles human biting rates (HBR) averaged 1.4 ± 1.7 bites per person per hour (bph), higher than indoor HBRs of 0.8 ± 1.2 bph. Vector incrimination detected Plasmodium DNA in ten An. koliensis mosquitoes in 2023, including Plasmodium falciparum (n = 2, 20%), P. vivax (n = 7, 70%) and P. ovale (n = 1, 10%). Blood meal analysis showed mixed feeding on humans and dogs, with human blood indices of 58.3% for An. koliensis and 66.7% for An. punctulatus . HBOs highlighted a substantial gap in indoor protection during the early evening before sleeping. Household surveys identified several drivers of exposure, including the absence of window and door screens, limited indoor residual spraying coverage, low usage of insecticide-treated nets, and unprotected outdoor activity. Conclusions Members of the An. punctulatus group were abundant across all study villages, with An. koliensis and An. punctulatus being the primary species. Vector incrimination confirmed active malaria transmission. These vectors demonstrated opportunistic feeding behavior, primarily on humans but also on dogs, which may offer an avenue for targeted interventions. Multiple gaps in personal and household protection were identified, both indoors and outdoors. Strengthening indoor protection through increased ITN use, installation of house screens, and evaluation of complementary tools such as spatial repellents may reduce indoor transmission. Community-led larval source management in and around households may help reduce outdoor transmission.

Examining the physiological and molecular impact of biochar on polyol synthesis and drought resistance in tomato plants

Scientific Reports Afaf Almaghamsi, Aisha A. M. Alayaf Jun 30, 2026 DOI: 10.1038/s41598-026-57460-y

Abstract Biochar is widely reported to buffer drought by improving soil water-holding capacity and nutrient availability, but whether it also modulates polyol metabolism linked to osmotic adjustment in tomato remains underexplored. Here we integrate physiological (polyol pools, SDH enzyme) and molecular (SDH transcripts) readouts with explicit soil and material characterisation to test how a wood-derived biochar alters the sorbitol/ribitol–SDH axis during water limitation. Tomatoes grown in biochar-amended soil (WB, 20 g kg −1 ) versus unamended soil (OB) were exposed to 100%, 75% and 45% soil water content (SWC). Under drought, OB plants accumulated more ribitol and sorbitol, with concomitant increases in SDH activity and SDH mRNA, whereas WB plants showed significantly attenuated polyol accumulation and SDH upregulation. At 75% and 45% SWC, SDH transcripts in OB were ~ 2 × and 4 × the 100% baseline, respectively, versus ~ 1 × and ~ 2 × in WB (two-way ANOVA, irrigation × biochar interaction p  < 0.01). Novelty: The data indicate that biochar attenuates the polyol-associated stress response (lower osmolyte demand and SDH induction) while maintaining growth, consistent with improved plant water status. These findings suggest that biochar may contribute to reducing osmotic stress responses, complementing its agronomic benefits in arid systems and supporting sustainable tomato production.

Validation of deep learning enabled software MetronMind to measure vertebral heart size and vertebral left atrial size in dogs

PLoS ONE Katharine Tess Sykes, Sonya G. Gordon, John J. Craig et al. Jun 30, 2026 DOI: 10.1371/journal.pone.0339253

Background Vertebral heart size (VHS) and vertebral left atrial size (VLAS) are objective radiographic measurements of heart and left atrial size, respectively and are associated with inter- and intraobserver variability when measured by humans. Artificial intelligence (AI) tools to determine VHS and VLAS have been developed and may reduce variability and save time. Objectives To compare two manual methods for measuring VHS and VLAS on right and left lateral canine thoracic radiographs and to compare measurements of VHS and VLAS made by deep learning-enabled program, MetronMind, with those made by a trained observer on right and left lateral radiographs. Animals Client-owned dogs (n = 1058) including 80 breeds with a variety of heart sizes, thoracic conformations and radiographic quality. Methods Retrospective, single-center, method-comparison study. Pearson’s correlation, Bland-Altman plots and Passing-Bablok regression were used to assess agreement. Results Correlation between traditional and modified manual measurements for VHS and VLAS was strong (r = 0.994 and r = 0.974, respectively), with minimal bias (−0.10 and 0.04 vertebrae, respectively) indicating that the modified methods closely approximate traditional measurements obtained from right lateral views. MetronMind measurements of VHS and VLAS from right lateral radiographs also correlated well with the human observer’s modified measurements (r = 0.947 and r = 0.811 respectively), showing small mean biases (0.08 and 0.07 vertebrae, respectively). Correlation between left and right lateral radiographic measurements of VHS (0.87 and 0.91) was higher than for VLAS (0.73 and 0.64) and bias was larger for VHS (0.26 and 0.31 vertebrae) than for VLAS (−0.13 and −0.10 vertebrae) for humans and MetronMind, respectively.

Comparison of techniques for the evaluation of taste sensitivity

Scientific Reports Jerry Hadi Juratli, I. Kaeb, A. Haehner et al. Jun 30, 2026 DOI: 10.1038/s41598-026-59454-2

Abstract Taste is essential for proper nutritional intake and thus has a significant effect on quality of life, yet clinical assessments of gustatory function are often underutilized or poorly standardized. This study aimed to compare two contrasting methodologies for evaluating taste sensitivity: Bayesian adaptive taste thresholds versus taste strip identification scores. One hundred participants underwent two sessions of gustatory testing, with a series of up to 20 trials of spray-based dilutions of sweet, sour, salty, and bitter tastants (QUEST) and the 16-item tastant-impregnated paper strips test. The log-transformed QUEST scores showed similar test–retest reliability to the taste strips score, with the bitter tastant showing the greatest correlation. There was no significant correlation between QUEST thresholds and taste strips identification scores. The QUEST spray test was rated more favorably by the participants for comprehensibility and pleasantness but took 12.38 min (± 33 s) on average, compared to 4.19 min (± 25 s) for the taste strips test. Tastant misidentifications on the taste strips assessment revealed that sour was disproportionately perceived as bitter at low concentrations but as salt at higher concentrations. QUEST logarithmic taste thresholds and tastant misidentifications on the taste strips test may serve as independent clinical measures of hypogeusia progression alongside taste strips identification score.

Retraction: Robust policy evaluation from large-scale observational studies

PLoS ONE Jun 30, 2026 DOI: 10.1371/journal.pone.0352001

Risk factors for abscess formation during or following endodontic treatment in an outpatient setting

Scientific Reports Thomas Gerhard Wolf, Janika Wollmann, James Deschner Jun 30, 2026 DOI: 10.1038/s41598-026-59391-0

Abstract This retrospective matched case-control study aimed to identify potential risk factors associated with the development of dental abscesses during or after endodontic treatment. The study analyzed data from 133 patients (74 males, 58 females; mean age 43.1 ± 16.5 years) with abscess formation of endodontic origin and compared them with a 1:1 matched control group without abscess. Matching was performed based on age and the specifically treated tooth. Collected variables included preoperative symptoms, radiographic findings, type and stage of endodontic intervention, and systemic medical conditions. Statistical analysis used chi-square or Fisher exact tests to compare between groups. STROBE and PROBE guidelines were followed. Among 2637 maxillofacial infections of odontogenic origin, 526 cases (19.9%) were related to endodontic treatment. No statistically significant associations were found regarding age, sex, or underlying systemic diseases ( p  > 0.05). Teeth exhibiting apical radiolucency were at significantly higher risk for abscess development ( p  < 0.001). Teeth that had undergone definitive root canal filling were significantly less likely to develop abscesses compared to those left without temporary filling after trepanation ( p  < 0.001) or treated with intracanal medication and temporary filling ( p  = 0.009). The highest incidence of abscess formation was observed within the first four days post-treatment ( p  < 0.001). Within the limitations of this retrospective study, abscess formation during or after endodontic treatment was linked to local procedural and radiographic factors, whereas systemic diseases did not demonstrate a measurable influence. The absence of temporary sealing and the presence of apical radiolucency were identified as significant risk indicators and may warrant increased clinical attention, particularly in the early postoperative period. Given the retrospective design and limited sample size, these findings should be interpreted with caution and require confirmation in prospective studies.

Dual cross-attentive mutual teaching for semi-supervised 3D medical segmentation

PLoS ONE Weiping Ma Jun 30, 2026 DOI: 10.1371/journal.pone.0352358

Semi-supervised learning can reduce the dependence on large-scale labeled data in 3D medical image segmentation.In this work, we propose a new Dual Crossed Attention Mutual Teaching (DCA-MT) framework that effectively utilizes both labeled and unlabeled data by integrating high-dimensional feature alignment, semantic-level crossed attention, and bidirectional knowledge distillation. Specifically, we employ a two-branch VNet architecture where the teacher-student network co-evolves through mutual mentoring and collaborative learning.To enhance representation consistency, we introduce maximum mean difference (MMD) loss and inter-class and intra-class contrast constraints to achieve global feature distribution alignment and class-level separability. A multi-head cross-attention module is designed to facilitate fine-grained semantic interaction between the two networks, allowing the two branches to dynamically exchange complementary features.In addition, the two-way mutual distillation strategy ensures that teacher and student networks benefit from each other's knowledge. Numerous experiments on the left atrial and pancreatic nih datasets show that our proposed approach has better performance and verifies the effectiveness and robustness of DCA-MT.