Smallholder dairy farms (SDFs) dominate the milk supply in Kenya and provide livelihoods for more than 2 million households. The majority of SDFs keep fewer than 6 cows, often on pieces of land less than 5 acres in size and produce approximately 5-15 liters of milk per cow per day. Due to resource constraints, inadequate knowledge of proper stall design and management and varying motivations in SDFs, cow housing and management conditions may remain suboptimal, affecting their welfare and farm productivity. Previous interventions recommending cow comfort adjustments in such farms have reported varied levels of compliance with minimal scaling opportunities. The aim of this study was therefore to estimate the relative importance of farmer-reported motivation attributes associated with cow-comfort-related dairy management decisions. Using cross-sectional data from 304 zero-grazing SDFs in Uasin Gishu county of Kenya, this study employed the best-worst scaling approach, a robust methodology in behavioral economics, to explore the stated relative priorities among motivation attributes regarding adopting cow comfort practices. Latent class analysis was used to explore farmer preference differences among relatively homogenous farm groups, and farm characteristics were used to understand their preferences. Results indicated that the highest-ranked attribute reflected intrinsic satisfaction /utility they obtain when knowing their cows were well-kept and healthy. Several economic motivations also ranked highly; including maximizing milk production and farm profitability, sustaining the dairy business, and facilitating time to engage in other activities and ensuring prompt diagnosis and treatment. These preferences differed across farm groups based on farmer, farm practice and herd characteristics, such as sex of the household head, proportion of income obtained from dairy, and education levels. These findings indicate that motivations for cow-comfort-related management decisions are heterogeneous and involve both use-value and non-use-value considerations. Extension and policy strategies may therefore be more effective when they are tailored to farmer segments and when they emphasize immediate farm-level economic and welfare-relevant benefits, rather than relying only on broad public-good messaging about animal welfare.
Colour constancy supports stable object colour perception across changes in illumination. Illuminant colour can be inferred from white surfaces or specular highlights, and many models adopt a 'brightest is white' heuristic to identify illuminant colour. We tested an alternative: observers use structured changes in the proximal image to locate highlights even when they are not the brightest elements. Using computer-rendered scenes, we manipulated the reliability of two cues-brightest element and highlight geometry. Each scene showed a single textured sphere lit by multiple point lights with identical spectra. The sphere had uniform spectral reflectance, while a noise texture attenuated the reflectance by a variable scale factor. We tested three specularity levels (matte, low and mid). Observers viewed a 1.5 s animation and judged whether chromatic changes represented an illuminant or material change. Performance for matte surfaces was near chance but improved sharply with increasing specularity. Observers outperformed an ideal observer relying solely on the brightest element. When highlights landed on darker texture regions, performance improved further, contrary to the brightest-element prediction. Phase scrambling, which disrupted highlight geometry while preserving global statistics, substantially reduced performance. Overall, observers do not rely solely on the brightest element; they exploit diffuse-specular regularities to resolve surface-illuminant ambiguity.
Choosing between endovascular and open (bypass) surgical revascularization for chronic limb-threatening ischemia has major implications for clinical practice. To evaluate the cost-effectiveness of endovascular vs bypass surgical revascularization. This economic evaluation used individual-level data from the Best Endovascular vs Best Surgical Therapy for Patients With Chronic Limb Ischemia (BEST-CLI) trial, with the first patient enrolled August 28, 2014, and the final October 18, 2019, for a median follow-up of 2.7 years. An individual-level, continuous-time Markov model with health states based on adjudicated clinical events from BEST-CLI was developed. Rates of clinical outcomes, health utilities, and health care resource use were derived from trial data. Unit costs came from Medicare insurance claims data and the physician fee schedule. The data were analyzed between June 27, 2025, and May 28, 2026. The main outcomes were incremental cost per life-years gained, incremental quality-adjusted life-years (QALYs) gained, incremental net monetary benefit, and cost per major events of revascularization and amputation avoided over a 5- and 10-year time horizon. One-way and probabilistic sensitivity analyses were performed to quantify uncertainty. The cohort included 1434 patients from BEST-CLI (mean [SD] age, 67 [10] years; 1026 male [71.5%]). In the base case analysis, over a 5-year time horizon, the mean per-person direct medical costs were $118 559 (95% credible interval [CrI], $86 978-$157 152) for bypass surgery and $125 535 (95% CrI, $96 205-$160 357) for endovascular surgery. The mean survival per person was 3.84 years (95% CrI, 3.74-3.93 years) and 3.78 years (95% CrI, 3.61-3.94 years) for bypass and endovascular surgery, respectively. The mean QALYs per person were 2.53 (95% CrI, 2.34-2.70) for bypass surgery and 2.48 (95% CrI, 2.28-2.67) for endovascular surgery. Bypass surgery dominated endovascular surgery with respect to both costs per life-year and per QALY gained. The results over 10 years were consistent with those of the 5-year BEST-CLI follow-up. In the Monte Carlo simulation, there was a 93% chance that bypass surgery was more cost-effective than endovascular surgery. This cost-effectiveness study found that bypass surgery was more cost-effective in most probabilistic simulations; however, uncertainty remained, highlighting the need for future research to identify subgroups in whom each approach may be cost-effective.
Understanding dengue dynamics from a spatiotemporal perspective has become increasingly relevant in recent years, as it allows the identification of factors influencing disease transmission, including climatic conditions, human behavior, and the distribution of Aedes aegypti, the primary vector responsible for dengue virus transmission in tropical and subtropical regions. This knowledge is essential for supporting public health decision-making and reducing the impact of dengue on vulnerable populations. This study proposes a spatio-temporal deep learning framework based on the U-Net++ architecture to generate high-resolution dengue risk maps in Colombia. The model integrates climatic, environmental, demographic, and socioeconomic information derived from satellite and census sources. Two spatial approaches were evaluated: a high-dimensional geography (HDG) approach at the national scale and a low-dimensional geography (LDG) approach applied to five departments with high dengue incidence. The model integrating climatic, social, and environmental information achieved the best performance in both approaches. In the HDG configuration, the best model (M9) reached a test mIoU of 0.6646 (validation mIoU = 0.7361). In the LDG configuration, the corresponding model achieved an average test mIoU of approximately 0.73 across departments. These results highlight the potential of integrating heterogeneous climatic, environmental, and socioeconomic data within a spatiotemporal deep learning framework to characterize dengue risk patterns and support high-resolution surveillance.
Environmental DNA (eDNA) analysis enables non-invasive detection of aquatic species, but no established framework integrates eDNA best practices with forensic validation standards. The critically endangered freshwater pearl mussel (Margaritifera margaritifera) exemplifies the need for forensically validated eDNA assays as traditional survey methods are invasive and time-consuming. This research provides a framework that bridges both ecological monitoring standards and forensic legal admissibility requirements in wildlife crime investigations. Here we describe the validation of a multiplex quantitative PCR (qPCR) assay targeting the mitochondrial cytochrome c oxidase subunit I (COI) gene, using the highest forensic standards and eDNA best practice. In silico and in vitro testing confirmed species specificity, with alignment against reference sequences showing 100% identity. The multiplex assay LOD and LOQ were calculated at 2.5 and 3.14 copies/µL respectively, determined by the performance of the short fragment. The assay demonstrated robust repeatability (R², 0.986-1.000), precision (coefficient of variation, 6-28%), and accuracy (recovery, 62-137%), meeting forensic thresholds. Field validation detected a previously undetected Welsh population through intelligence-led sampling, while blind proficiency testing in the Lake District achieved 95.8-100% detection rates across juvenile release sites and successfully detected distribution patterns later confirmed by stakeholders. We establish the first forensic eDNA validation of a protected species, providing a transferable methodology for conservation, regulatory enforcement, and wildlife crime investigations, ready for deployment in an operational environment. We believe that this integrated approach can be transferred to similar forensic and conservation scenarios across a broad range of species and geographical contexts.
To determine the prevalence of retinal pigment epithelium (RPE) microrips within serous pigment epithelial detachments (PEDs) in first-episode acute central serous chorioretinopathy (CSCR) and to assess their spatial relationship with focal angiographic leakage using multimodal imaging. Retrospective observational case series. Thirty-nine eyes of 39 patients with acute CSCR (symptom duration <3 months). All eyes underwent spectral-domain OCT, fluorescein angiography, and indocyanine green angiography (ICGA). Imaging biomarkers at leakage sites included PEDs, RPE microrips, outer retinal erosion and outer retinal pinching, pachyvessels, midphase ICGA hyperfluorescent plaques, and subfoveal choroidal thickness. Clinical outcomes included best-corrected visual acuity, subretinal fluid (SRF) resolution, recurrence, and progression to chronic disease. Mean age was 44.5 ± 10.0 years, and 61.5% of patients were male. PEDs colocalized with the focal leakage site in all eyes, whereas RPE microrips colocalized with the site of focal angiographic leakage in 66.7% of eyes, with good interobserver agreement (κ = 0.79) observed. Outer retinal erosion and outer retinal pinching were observed in 38.5% and 12.8% of eyes, respectively. Midphase ICGA hyperfluorescent plaques were observed in 56.4% and pachyvessels in 46.2% of cases. Mean best-corrected visual acuity improved from 0.24 ± 0.25 (20/34 Snellen) to 0.10 ± 0.17 logMAR (20/25 Snellen) at final follow-up, with complete SRF resolution in 68.4%. Baseline subfoveal choroidal thickness correlated with SRF height (P = 0.018), but no baseline imaging biomarkers predicted final visual or anatomical outcomes. RPE microrips within serous PEDs frequently are observed and precisely colocalise with focal leakage in acute CSCR, supporting a mechanical model of focal RPE dehiscence driven by increased choroidal pressure. These findings provide insight into CSCR pathophysiology and may inform therapeutic strategies.
Objective: To compare the subjective and objective visual outcomes after implantation of monofocal, bifocal, and extended depth-of-focus (EDOF) intraocular lenses (IOLs). Methods: This non-randomized controlled clinical trial enrolled cataract patients scheduled for phacoemulsification and IOL implantation at Beijing Tongren Eye Center, Capital Medical University, between April and December 2025. According to the implanted IOL type, patients were assigned to monofocal, bifocal, diffractive EDOF, and non-diffractive EDOF groups. Visual acuity, defocus curves, objective visual quality, and the Visual Function Index-14 (VF-14) questionnaire were evaluated at 3 months postoperatively. Uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), distance-corrected intermediate visual acuity (DCIVA), and distance-corrected near visual acuity (DCNVA) were measured and recorded in logarithm of the minimum angle of resolution (logMAR). Defocus curves were obtained from +1.00 to -4.00 D in 0.50-D steps under best distance correction. Intraocular wavefront aberrations and modulation transfer function (MTF) were measured at a 4.0-mm pupil diameter using an optical path difference-based aberrometer. Objective scatter index, MTF cut-off frequency, and Strehl ratio (SR) were assessed using a double-pass optical quality analysis system. Intergroup comparisons were performed using one-way analysis of variance or the Kruskal-Wallis H test with Bonferroni post hoc correction. Pearson correlation analysis and linear regression models were used to identify factors associated with intraocular higher-order aberrations (HOAs). Results: A total of 67 patients (67 eyes) were enrolled. Except three patients lost to follow-up and one patient incomplete data, 63 patients (63 eyes) were included in the analysis, with 15 eyes in the monofocal group, 17 eyes in the bifocal group, 14 eyes in the diffractive EDOF group, and 17 eyes in the non-diffractive EDOF group. The mean age was (67.24±6.85) years. There were 17 men and 46 women. Baseline characteristics were comparable among the four groups (all P>0.05). At 3 months postoperatively, UDVA did not differ significantly among the groups (P=0.054). The bifocal group achieved the best CDVA (-0.02±0.07), which was significantly better than that of the monofocal group (0.07±0.09) and the diffractive EDOF group (0.05±0.05) (P=0.033). The bifocal group also demonstrated significantly higher MTF cut-off frequency and SR than the monofocal and non-diffractive EDOF groups (both P<0.05). Both EDOF groups showed significantly better DCIVA than the monofocal and bifocal groups (both P<0.001). On defocus curve analysis, the diffractive EDOF group maintained visual acuity of 0.2 logMAR or better within a defocus range of up to -2.00 D, whereas the non-diffractive EDOF group maintained this level within -1.50 D. The monofocal group exhibited the poorest DCNVA (0.60±0.23), while no significant difference was observed among the other three groups. Postoperative spherical equivalent refraction was close to emmetropia in the monofocal [(-0.15±0.44) D] and bifocal [(0.12±0.37) D] groups, whereas mild myopia was observed in the diffractive EDOF [(-0.73±0.41) D] and non-diffractive EDOF [(-0.63±0.51) D] groups (P<0.001). The spherical aberration was significantly higher in the non-diffractive EDOF group than in the other three groups (all P<0.05). Nevertheless, this group achieved the highest VF-14 score and showed better performance in night driving and fine handwork tasks. Correlation and regression analyses demonstrated that the MTF cut-off and SR were independently associated with intraocular HOAs. Conclusions: EDOF IOLs provided superior intermediate visual acuity and a broader continuous functional range of vision. The non-diffractive EDOF group showed a trend toward better subjective visual function, although VF-14 scores did not differ significantly among the four groups. Bifocal IOLs demonstrated superior objective optical performance. Selection of IOLs should be individualized according to the functional visual range, optical quality, and patient-specific visual demands. 目的： 比较单焦点、双焦点及景深延长型（EDOF）人工晶状体（IOL）植入术后主观与客观视觉结局。 方法： 非随机对照临床试验。纳入2025年4月至12月在首都医科大学北京同仁眼科中心拟行超声乳化白内障吸除联合IOL植入术的白内障患者，根据植入IOL类型分为单焦点组、双焦点组、衍射型EDOF组和非衍射型EDOF组。术后3个月评估视力、离焦曲线、客观视觉质量和视觉功能指数14量表（VF-14）。采用标准对数视力表检测未矫正远视力（UDVA）、矫正远视力（CDVA）、远矫正中间视力（DCIVA）和远矫正近视力（DCNVA），以最小分辨角的对数记录；离焦曲线在最佳远矫正基础上以0.50 D为间隔从-4.00~+1.00 D测定；眼内波前像差及调制传递函数（MTF）采用基于光程差的波前像差仪在4.0 mm瞳孔直径下测量；客观散射指数（OSI）、MTF截止频率及斯特列尔比（SR）采用双通道客观光学质量分析仪进行系统评估。组间比较行单因素方差分析或Kruskal-Wallis H检验，事后行Bonferroni多重比较；变量相关性采用Pearson相关分析，影响因素采用线性回归模型。 结果： 本研究共纳入白内障患者67例，其中3例未完成术后3个月随访，1例资料缺失，未纳入最终分析。最终63例（63只眼）完成研究并纳入统计分析，年龄为（67.24±6.85）岁；男性17例，女性46例。其中单焦点组15例、双焦点组17例、衍射型EDOF组14例及非衍射型EDOF组17例。研究对象4个组间年龄、性别分布等差异均无统计学意义（均P>0.05），组间具有可比性。术后3个月，4个组UDVA分别为0.11±0.11、0.04±0.08、0.09±0.07及0.10±0.11，差异无统计学意义（P=0.054）。双焦点组CDVA最优（-0.02±0.07），显著优于单焦点组（0.07±0.09）和衍射型EDOF组（0.05±0.05）（P=0.033），且MTF截止频率及SR显著优于单焦点组和非衍射型EDOF组（均P<0.05）。两组EDOF组DCIVA（衍射型EDOF组0.12±0.07、非衍射型EDOF组0.16±0.10）均优于单焦点组（0.32±0.18）和双焦点组（0.38±0.16）（P<0.001），衍射型EDOF组在2.00 D以内的近视范围内、非衍射型EDOF组在1.50 D以内的近视范围内维持≤0.2的视力。DCNVA方面，单焦点组最差（0.60±0.23），其余3个组间差异无统计学意义。术后等效球镜度数单焦点组（-0.15±0.44）D和双焦点组（0.12±0.37）D呈正视状态，非衍射型EDOF组（-0.63±0.51）D和衍射型EDOF组（-0.73±0.41）D呈轻度近视状态（P<0.001）。非衍射型EDOF组球差显著高于其余3个组（P<0.05），总高阶像差显著高于双焦点组和衍射型EDOF组（均P<0.05），但VF-14评分最高，夜间驾驶及精细手工方面表现优异。相关性及回归分析进一步证实，MTF截止频率和SR与眼内高阶像差独立相关。 结论： EDOF型IOL可提供更优的中距离视力及连续功能性视程，非衍射型EDOF 组患者主观视觉功能评分呈较高趋势，但四组间VF-14评分差异无统计学意义；双焦点IOL客观光学性能更佳。IOL的选择应综合考虑功能性视程、光学质量及患者的个体视觉需求。.
Weak spectral responses, overlapping bands, and background interference remain key challenges in spectroscopic quantitative analysis. Therefore, this study developed a cross-shaped metal-insulator-metal metamaterial (MIM-MM) platform to enhance Raman, near-infrared (NIR), and visible (VIS) spectroscopy. The optimized MIM-MM consisted of an Au/SiO2/Au multilayer structure with periodically arranged cross-shaped Au nanoarrays. With a simulated juice matrix as the validation model system, MIM-MM-enhanced spectra combined with the partial least squares regression (PLSR) model generally exhibit better predictive performance in predicting sugar concentration than unenhanced spectra. In mixed sugar systems, Raman fingerprint (FP) spectra showed the best quantitative analysis ability, with the PLSR model achieving an average RMSE of 0.6330 and an average R2 of 0.9502. In addition, Raman high wavenumber (HW), NIR, and VIS spectra can provide supplementary information spectral information. Furthermore, the variable importance in projection (VIP) analysis method confirmed that the MIM-MM enhanced key spectral variables related to sugar molecular vibrations and optical responses. Overall, these results indicate that MIM-MM shows potential as a front-end signal enhancement platform for improving weak spectral responses and feature distinguishability in spectroscopic sugar prediction. This study provides a controlled feasibility evaluation of material-assisted spectral analysis in an aqueous multicomponent model system.
Chemokine and chemokine receptors (CKRs) regulate cell growth, tumor immunity, migration, angiogenesis, and metastasis. We evaluated associations of inherited chemokine/CKR variants with subsequent primary melanomas among patients with a previous melanoma. Our population-based cohort included 2458 incident single primary and 1205 multiple primary melanoma (MPM) cases. We examined 215 a priori candidate polymorphisms selected based on prior evidence and predicted functional relevance, along with nine haplotype blocks. Logistic regression estimated odds ratios and 95% confidence intervals, adjusting for age at diagnosis, sex, age-by-sex interaction, and study center. Thirty variants across 14 genes were nominally associated with MPM (P < 0.05), with effect sizes ranging from 16 to 80%. Five haplotypes in CXCL12, CCL18, CXCR2, and CCR9 were significantly associated with MPM (Pglobal < 0.03) with stronger effects than individual-effect alleles. Functional annotation identified 160 associated index variants and linkage disequilibrium proxies acting as skin expression quantitative trait loci for CXCR2, XCR1, BCL9L, FYCO1, PNKD, RUFY4, and RP11-378A13.1 (P < 1.06E-05 to 4.3E-34). These variants were classified functionally as candidate causal variants based on convergent regulatory annotation and skin‑specific expression quantitative trait locus evidence, linking chemokine/CKR variants and haplotypes to downstream genes involved in epithelial-mesenchymal-like transition, matrix remodeling, and other tumorigenic processes, thereby implicating immune regulatory mechanisms in melanoma development. To the best of our knowledge, this is the first study to demonstrate an association of a priori chemokine/CKR variants with subsequent melanoma risk. Validation and evaluation of associations with tumor characteristics and survival are warranted.
Retained intraocular foreign bodies (IOFBs) are important causes of vision-threatening ocular trauma and preventable visual impairment. This systematic review and meta-analysis evaluated the clinical manifestations, diagnostic approaches, surgical management and visual outcomes associated with IOFBs. A systematic search of PubMed, Embase, Scopus and Web of Science identified studies published between 1 January 2000 and 30 November 2025. Eligible studies reporting clinical presentation, imaging findings, management strategies or outcomes of retained IOFBs were included. Data were synthesised using a random-effects meta-analysis model. A total of 59 studies involving 4289 patients were included of which 42 were eligible for quantitative meta-analysis. Reduced visual acuity (78%), ocular pain (42%) and intraocular inflammation (36%) were the most common clinical manifestations. Endophthalmitis occurred in 12.4% and retinal detachment in 9.8% of cases. Favourable visual outcomes (Best Corrected Visual Acuity-BCVA≥20/40) were achieved in 65.7% of patients. CT demonstrated the highest diagnostic accuracy (91.2%), while ultrasonography showed 83.6% sensitivity particularly in media opacity. Pars plana vitrectomy was the predominant surgical approach, with an overall surgical success rate of 96.5%. Early IOFB removal (≤24 hours) was associated with improved visual recovery and reduced complications. Early diagnosis and timely surgical intervention are critical for optimising outcomes in retained IOFBs. CT and pars plana vitrectomy remain central to management, while delayed intervention increases the risk of severe complications. Standardised protocols and prospective multicentre studies are needed to strengthen evidence and improve patient care.
Aneurysmal subarachnoid hemorrhage (aSAH) is a life-threatening condition with high morbidity and mortality. Long-term outcomes remain difficult to predict. Computed tomography (CT) is central to diagnosis and early risk stratification, yet the prognostic value of non-contrast CT at the end of the acute care phase is not fully established. We conducted a retrospective single-center study of adult patients with confirmed aSAH admitted to the Neurocritical Care Unit (NCCU) of the University Hospital Zurich (January 2016-June 2024). Survivors with CT at discharge from the NCCU and 12-month follow-up were included. Functional outcome was assessed using the Glasgow Outcome Scale Extended (GOSE) and dichotomized as favorable (5-8) or unfavorable (1-4). CT scans were evaluated for infarcts, hemorrhage distribution, hydrocephalus requiring shunt, midline shift, and decompressive surgery. Multivariable logistic regression with multiple imputation by chained equations (MICE) addressed missing data, with days from admission to discharge CT included as a covariate in the clinical-severity model to control for differential timing. Discriminative performance was assessed using receiver operating characteristic (ROC) curve analysis. Of 500 admitted patients, 325 met inclusion criteria. At 12 months, 234 had favorable and 91 unfavorable outcomes. In a radiological model, increasing age and infarcts involving the thalamus, basal ganglia, corpus callosum, and cerebellum, together with cortical infarction and intraventricular hemorrhage, were independently associated with unfavorable outcome; sulcal subarachnoid hemorrhage showed only a non-significant trend toward a protective association. In a separate clinical-severity model, increasing age, ventriculoperitoneal shunt placement, decompressive craniectomy, and days-to-CT scan were independently associated with unfavorable outcome. ROC analysis showed low-to-moderate discriminative ability (AUC range 0.529-0.688), with intraventricular hemorrhage performing best. Discharge non-contrast CT provides incremental prognostic information in aSAH survivors. Deep or midline infarctions are particularly decisive for early prognostication and rehabilitation using routinely available imaging.
Per- and polyfluoroalkyl substances (PFAS) pose long-term risks to ecosystems and human health because of their persistence, bioaccumulation, long-range transport, and toxicity. Focusing on the Northeast Black Soil Region of China, a major grain production base with intensive agricultural activities and complex industrial inputs, this study developed a comprehensive analytical framework integrating multi-source data (2005-2025), including literature, environmental monitoring, soil properties, and toxicity databases. The framework combined meta-analysis, principal component analysis (PCA), machine learning, and entropy-weight-based risk assessment to systematically characterize PFAS pollution patterns, spatial heterogeneity, and priority risks. Random-effects meta-analysis revealed moderate-to-strong positive correlations among PFAS congeners (pooled effect size = 0.336) with significant heterogeneity (I² = 96.80%). PCA showed that the first two principal components explained over 48% of the total variance, whereas six components accounted for more than 85%. Among eleven machine-learning models, Logistic Regression achieved the best performance and stable generalization when the sample size exceeded 60 samples. SHapley Additive exPlanations (SHAP) identified latitude, longitude, and soil organic carbon as the dominant predictors. Risk assessment indicated that PFOS and PFOA exhibited the highest priority levels (ToxPi: 0.872 and 0.499; EHPi: 0.839 and 0.679), whereas approximately 60% of PFAS were classified as low risk and 20% as medium-to-high risk. This framework provides quantitative support for targeted PFAS management and offers a transferable strategy for regional-scale assessment of emerging contaminants.
Intraorbital space-occupying lesions impair visual function, yet whether this stems primarily from optic nerve compression, elongation or orbital apex crowding remains unclear. This study investigated the differential impact of these mechanical factors on visual acuity (VA) and visual field (VF) in patients with intraorbital cavernous haemangiomas. This retrospective cohort study analysed 130 patients with unilateral intraorbital cavernous haemangiomas. Key imaging parameters were measured from CT/MRI and categorised into optic nerve elongation, compression (eg, optic-nerve aspect ratio (ONAR)) and orbital crowding (eg, tumour-to-orbit (TO) ratio, tumour mid-position (TMP)). Univariate and multivariate linear regression models were used to correlate these metrics with visual outcomes. Regression analysis revealed two distinct injury mechanisms. Direct compression, best quantified by ONAR, was the strongest predictor for VA decline (R2=0.15, p<0.0001) but correlated weakly with VF loss (R2=0.09, p=0.0015). Conversely, orbital apex crowding (TO-ratio and TMP) correlated more strongly with VF defects (R2=0.19 and R2=0.20, p<0.0001) than with VA. Optic nerve elongation showed no significant correlation with visual dysfunction. The final multivariate model explained 17.68% of the variance in VA and 36.48% in VF. Visual loss from orbital tumours involves two mechanical pathways: direct compression preferentially impairs central vision (VA), while orbital apex crowding predominantly affects peripheral vision (VF). Imaging metrics like ONAR and TO-ratio can effectively quantify these mechanisms. They may serve as supplementary reference indicators for evaluating visual risk and informing clinical follow-up and surgical consultation, with ONAR indicating risk to central vision and TO-ratio signalling impending field loss.
From 2020 to 2022, an outbreak of listeriosis occurred in Austria, resulting in ten cases, three of which proved fatal. The outbreak was traced to a traditional southern European dairy product known as Kajmak, a lightly salted cream product. The subsequent sequencing of clinical and food-related isolates with clonal complex (CC)1 clonal type (CT) 6568 led to the identification of the outbreak's origin. CT6568 is classified under CC1, which exhibits notable diversification among human isolates, with 14 distinct CTs identified between 2020 and 2022. The clientele of the cheese factory primarily comprised local Balkan restaurants, delicatessen counters, and small bakeries. The existing self-monitoring system and the Hazard Analysis and Critical Control Points (HACCP) program did not meet the requisite standard of adequacy. Listeria monocytogenes was identified in the production and ripening room, indicating potential recontamination of Kajmak during the manual filling and ripening process. The highly perishable product in question was also incorrectly labeled with a best-before date (>21 days) instead of a use-by date. Furthermore, crucial labeling information for customers, such as the necessity of strict refrigeration at 4 °C, was absent. The outbreak was likely precipitated by the discontinuous production of Kajmak, which resulted in inadequate environmental hygiene. The retrospective report on the outbreak highlights the lack of risk analysis in research conducted on food safety risks associated with small-scale production of traditional foods. It is essential that manufacturers align their understanding with the principles of the HACCP concept and effective self-monitoring systems.
Colorectal cancer is a leading cause of cancer-related deaths in the United States, and colonoscopy remains the gold standard for early detection and prevention. However, many procedures are postponed due to inadequate bowel preparation, a preventable failure often caused by patients' difficulty in understanding and following written prep instructions. Prior interventions such as reminder apps and instructional videos have improved adherence only modestly, largely because they cannot answer patient-specific questions. Recent advances in large language models (LLMs) raise the possibility of developing conversational assistants that can provide interactive support to patients in procedure preparation. This study evaluated the correctness, harmfulness, and diversity of synthetic dialogues generated by leading LLMs acting as both simulated AI Coaches and patients for colonoscopy preparation. Five leading LLMs-OpenAI's o3, GPT-4.1, and GPT-5.1; Meta's Llama 3.3 70B; and Mistral's Large-2411-were used to generate 250 patient-AI Coach dialogues per model. Dialogues consisted of 3 to 7 question-answer pairs concerning diet, medications, and other prep-related topics. A multiprompt, multiquestion approach was designed to elicit diverse patient questions, and an error taxonomy was established to assess model capabilities in responding to questions. Human raters, including 3 medical experts, evaluated the generated questions for difficulty and the responses for correctness, error type, and potential harmfulness. Automatic evaluation using an LLM-as-a-judge approach complemented human evaluation. Question diversity was assessed using lexical diversity metrics (Distinct-1 and Distinct-2) and entropy. In addition, we evaluated a safety filtering mechanism in which responses judged incorrect by an automated evaluator were replaced with a deferral message instructing patients to contact their health care provider. Differences in response correctness across models were evaluated using permutation tests conducted at the dialogue level. Interrater agreement among human evaluators was assessed using the Gwet AC1 statistic. The study was conducted between May and September 2025. Automatic evaluation results closely aligned with human judgments: leading models approached but did not achieve adequate performance. Closed-weight models (GPT-5.1, GPT-4.1, and o3) outperformed open-weight models (Llama and Mistral) on correctness, with the reasoning models (GPT-5.1 and o3) performing best. This turn-level ranking was preserved under the supplementary single-prompt baseline, although dialogue-level rankings differed. All models produced harmful errors, primarily due to omissions or misinterpretations of prep instructions. The multiprompt generation strategy substantially increased the diversity of patient questions compared with a single-prompt baseline. Applying an automated safety filter reduced overall error rates but failed to eliminate harmful responses. Although LLMs demonstrate strong potential to support colonoscopy preparation, none are yet reliable enough for unsupervised deployment in patient-facing contexts. Persistent harmful errors and the limited effectiveness of simple filtering mechanisms highlight the need for improved instruction adherence, stronger safety mechanisms, and validation using real patient queries.
Objective: To evaluate the changes of binocular visual function and patient satisfaction after small-incision lenticule extraction (SMILE) with micro-monovision in myopic patients with presbyopia. Methods: This was a single-arm clinical trial. Myopic patients with presbyopia who underwent micro-monovision SMILE at Beijing Tongren Hospital, Capital Medical University, from January 2019 to March 2021 were enrolled. The surgical design principle was full correction of the dominant eye for distance vision, with the non-dominant eye undercorrected by -1.50 D to -0.50 D for near vision. Best corrected distance visual acuity and best corrected near visual acuity (CNVA) for the dominant eye, non-dominant eye, and both eyes were measured before surgery. Uncorrected distance and near visual acuity (UDVA, UNVA), CDVA, and CNVA for the dominant eye, non-dominant eye, and both eyes were measured at 1, 3, 6 months, 3 years, and 5 years after surgery. Accommodative function parameters were measured preoperatively and at each postoperative follow-up visit, including binocular accommodative amplitude (AMP) measured by the push-up method, accommodative response (BCC) measured by the binocular cross-cylinder method, negative relative accommodation (NRA), and positive relative accommodation (PRA), as well as distance and near stereopsis and Worth 4-dot tests. A visual satisfaction questionnaire was administered after surgery. Results: A total of 51 patients (102 eyes) with myopia and presbyopia were included, comprising 19 males and 32 females, aged (41.68±2.47) years. Sixteen patients (32 eyes) completed the 5-year follow-up. The binocular UDVA and UNVA were (-0.02±0.06) logMAR and (0.01±0.09) logMAR at 5 years after surgery, respectively. The spherical equivalent of the dominant and non-dominant eyes was (-0.24±0.22) D and (-0.80±0.37) D, respectively. The binocular AMP decreased from (5.66±0.28) D before surgery to (4.36±0.19) D at 1 month after surgery, and returned to the preoperative level at 3 months after surgery; The PRA decreased from (-1.93±0.17) D before surgery to (-1.33±0.13) D at 1 month after surgery, and returned to the preoperative level at 6 months after surgery. The BCC increased at 1 month after surgery and returned to the preoperative level at 6 months after surgery (differences across time points were all statistically significant, P<0.05). These indicators slightly decreased at 3 and 5 years postoperatively compared with their respective recovered levels. There were no statistically significant differences in NRA and near stereopsis at any postoperative follow-up time point compared with preoperative values. All patients achieved 60″ distance and near stereopsis at each postoperative follow-up visit point, and the Worth 4-dot tests showed both central and peripheral fusion function in both eyes. The subjective satisfaction score was (9.06±1.18) (out of 10) at 5 years after surgery, with a satisfaction rate of 15/16. Conclusions: SMILE with micro-monovision had an impact on accommodative function in the early postoperative stage. The AMP, PRA, and BCC returned to preoperative levels at 3 and 6 months postoperatively, with a slight decrease at 3 to 5 years. The procedure did not affect binocular visual functions such as fusion and stereopsis. At 5 years after surgery, patients exhibited good distance and near vision, maintained binocular visual functions, and reported high subjective satisfaction. 目的： 评估近视伴老视患者行微单眼视设计的飞秒激光小切口角膜基质透镜取出术（SMILE）后双眼视功能变化及患者满意度。 方法： 单臂临床试验。纳入2019年1月至2021年3月在首都医科大学附属北京同仁医院接受微单眼视设计的SMILE矫正近视伴老视的患者。手术设计原则为优势眼足矫看远，非优势眼预留-1.50~-0.50 D看近。术前测量优势眼、非优势眼、双眼最佳矫正远视力（CDVA）、最佳矫正近视力（CNVA）；分别于术后1个月、3个月、6个月、3年、5年时测量优势眼、非优势眼、双眼的裸眼远视力（UDVA）、裸眼近视力（UNVA）及CDVA、CNVA；分别于术前及术后各时间点测量调节功能参数，包括移近法测量双眼调节幅度（AMP）、交叉柱镜法测量双眼调节反应（BCC）、负相对调节（NRA）及正相对调节（PRA），以及远、近立体视和Worth四点灯检查；术后进行视觉满意度问卷调查。 结果： 共纳入近视伴老视患者51例（102只眼），其中男性19例、女性32例，年龄（41.68±2.47）岁，术后5年完成随访16例（32只眼）。术后5年双眼UDVA和UNVA分别为-0.02±0.06和0.01±0.09，优势眼和非优势眼等效球镜度数分别为（-0.24±0.22）D和（-0.80±0.37）D。双眼调节幅度由术前（5.66±0.28）D降至术后1个月（4.36±0.19）D，术后3个月恢复至术前水平；PRA由术前（-1.93±0.17）D降至术后1个月（-1.33±0.13）D，术后6个月恢复至术前水平；调节反应术后1个月升高，术后6个月恢复至术前水平（不同时间点间差异均有统计学意义，均P<0.05）；术后3年和5年上述指标较恢复期略有下降。NRA和近立体视在术后各时间点与术前比较差异均无统计学意义；所有患者术后各时间点远、近立体视均可达到60″，Worth四点灯检查均显示双眼中心及周边融合功能。术后5年主观满意度评分为（9.06±1.18）分，满意率为15/16。 结论： 微单眼视设计的SMILE术后早期对调节功能有影响，调节幅度、PRA和调节反应分别于术后3个月和6个月恢复到术前水平，术后3~5年略有下降。微单眼视设计的SMILE不影响融合和立体视等双眼视功能。术后5年表现出较好的远近视力、双眼视功能维持和较高的主观满意度。.
Phylogenomic discordance is widespread across plants, but its evolutionary significance is often obscured when conflict is treated primarily as analytical noise rather than as evidence of underlying processes. In woody lineages in particular, incomplete lineage sorting, introgression, and genome duplication can interact over long timescales to produce complex genomic histories that are not adequately summarized by a strictly bifurcating tree. Here, we use Malus as a model woody genus to investigate how these processes structure conflict across a genus-scale, accession-based phylogenomic framework. Using broad taxon sampling, hundreds of nuclear loci, plastid genomes, and genome-wide SNP summaries, we reconstruct a robust nuclear backbone for sampled Malus lineages and evaluate where discordance is concentrated and which processes best explain it. Nuclear analyses resolve eight major clades, whereas conflict is non-random and localized to recurrent hotspots rather than evenly distributed across the tree. Cytonuclear discordance is similarly concentrated, especially around Clade H, represented by sampled accessions of M. tschonoskii, where localized plastid-nuclear disagreement is consistent with candidate plastid capture or organellar introgression. Multiple complementary analyses further indicate that the strongest conflict is not explained by ILS alone, but instead reflects lineage-structured introgression, while polyploid complexes represent additional localized sources of evolutionary complexity. Together, these results provide evidence for a reticulate genomic backbone in Malus and show how integrating nuclear, plastid, and genome-wide conflict analyses can help distinguish background discordance from process-specific signals in woody plant radiations. Several lineage-level reticulation hypotheses identified here should now be tested with broader population-level sampling and curated reference accessions.
Subjective cognitive decline (SCD), which may represent the preclinical stage of Alzheimer's disease (AD) in a subset of individuals, lacks objective biomarkers, thus hindering the implementation of early intervention for AD. This study tentatively explored the potential value of neurite orientation dispersion and density imaging (NODDI) in assessing the microstructural integrity of gray matter (GM) and white matter (WM) in SCD. We enrolled 47 SCD subjects and 30 healthy controls (HCs). Through the TBSS and GBSS analysis based on NODDI, the neurite density index (NDI) and orientation dispersion index (ODI) were used to evaluate the microstructural integrity of GM and WM. NODDI parameters were extracted from multiple brain regions to assess their diagnostic utility in distinguishing SCD from HCs. In WM microstructure, NDI of the genu of corpus callosum was decreased, and the ODI of the left posterior thalamic radiation and posterior corona radiata were decreased. In GM microstructure, the NDI of the left hippocampus, amygdala and parahippocampal gyrus were decreased, whereas the ODI of the rolandic operculum, opercular part of inferior frontal gyrus, and precentral gyrus were increased. Logistic regression analysis results showed that the NDI value of GM, the ODI value of GM and the ODI value of WM were independent predictors and the combination of the three parameters had the best classification effect. This exploratory study suggests that NODDI may detect microstructural differences in GM and WM associated with SCD. These preliminary findings provide hypothesis-generating clues for future research into potential imaging markers for identifying SCD individuals. Additional studies are necessary to confirm these preliminary results.
This single-center, open-label, Phase 1 trial evaluated the safety and tolerability of chimeric antigen receptor (CAR)-T cells targeting interleukin-13 receptor α2 (IL13Rα2) in patients with malignant gliomas. Adults with World Health Organization Grade 3-4 gliomas, recurrent after standard treatment and expressing IL13Rα2 (confirmed by immunohistochemistry), received a single intravenous infusion of autologous CAR-T cells. Primary objectives were to measure the maximum-tolerated dose (MTD) and recommended Phase 2 dose. Secondary objectives included assessment of pharmacokinetics, objective response rate, progression-free survival (PFS), and overall survival (OS). Ten patients were sequentially assigned to three dosing cohorts (1.0×107, 3.0×107, 1.0×108 cells/kg). The MTD was not reached, as no dose-limiting toxicities were observed, and all adverse events (AEs) were Grade ≤3 without irreversible sequelae. The most common treatment-related AEs were pyrexia (70%) and elevated transaminase (70%). CAR-T cell levels in blood reached peak concentrations at a median of 7 days (range 3-14), and the transgene was detected in cerebrospinal fluid from five of six patients treated at the higher two doses. Nine patients were evaluable for survival analysis, among whom six were available for response assessment at 3 months; best response was stable disease in three cases. Median PFS was 2.6 months (95% confidence interval [CI]: 0.95-5.75), and median OS was 10.9 months (95% CI: 8.2-26.6). Intravenous IL13Rα2-targeted CAR-T therapy was well tolerated and demonstrated preliminary safety. Further studies are warranted to optimize dosing and preconditioning regimens and identify exploratory biomarkers for efficacy evaluation.
To evaluate the efficacy and safety of en face optical coherence tomography (OCT)-guided navigated focal laser photocoagulation for central serous chorioretinopathy (CSCR). Prospective, non-randomized non-inferiority study. Patients with non-resolving CSCR and focal leakage diagnosed by fluorescein angiography (FA) received navigated focal laser photocoagulation planned in a masked fashion based on either en face OCT or FA. The main outcome measure was complete resolution of subretinal fluid (SRF) within three months. Seventy-eight eyes of 78 patients and 48 eyes of 48 patients were included in the OCT-guided and FA-guided groups, respectively. The mean LogMAR baseline best-corrected visual acuity (BCVA) and central retinal thickness (CRT) in the OCT-guided group were 0.16 ± 0.12 and 379.1 ± 59.2 µm and improved to -0.03 ± 0.06 and 227.1 ± 46.2 µm, respectively (p < 0.001). The mean LogMAR BCVA and CRT in FA-guided group were 0.11 ± 0.08 and 373.8 ± 56.3 µm and improved to -0.03 ± 0.07 and 221.9 ± 43.1 µm, respectively (p < 0.001). Complete resolution of SRF at 3 months was achieved in 66 out of 78 (84.6%) eyes in the OCT-guided group and in 39 out of 48 (81.3%) eyes in the FA-guided group (p = 0.62). No adverse events were registered in either treatment group. Focal navigated laser photocoagulation planned solely based on en face OCT demonstrated non-inferior safety and efficacy compared with conventional FA-guided treatment for non-resolving CSCR with focal leakage.