Explore the vast range of titles available.

ObjectivesTo evaluate the performance of reasoning large language models (LLMs) with human-like thinking in ophthalmic question answering.MethodsWe evaluated two state-of-the-art open-source reasoning LLMs (DeepSeek-R1 and QwQ-32B) and one conventional non-reasoning LLM (LLaMA-3.3-70B-Instruct) models on ophthalmology questions, assessing not only answer accuracy (ACC) but also the quality of their reasoning processes. First, we curated MedQA-Eye, a dataset of 967 ophthalmology questions across 10 subspecialties, 3 scenarios, 5 medical entities and 3 languages. Second, we proposed a novel framework considering human thinking patterns essential to medical practice to evaluate the thinking performance of reasoning LLMs on MedQA-Eye.ResultsDeepSeek-R1 demonstrated superior overall ACC (90.59%, 95% CI 88.59% to 92.27%) to LLaMA-3.3-70B-Instruct (87.90%, 95% CI 85.69% to 89.81%, p=0.015) and QwQ-32B (84.28%, 95% CI 81.85% to 86.44%, p<0.001) with performance varying across subspecialties. Analysis of reasoning LLMs revealed incorrect logical inference as the primary point of failure, accounting for 93.41%–94.74% of incorrectly answered questions. We further quantified semantic uncertainty in reasoning LLM thinking as a predictor of answer reliability. DeepSeek-R1 exhibited lower semantic uncertainty (1.04±3.63) compared with QwQ-32B (4.31±40.70), p<0.001.ConclusionReasoning LLMs demonstrated superior performance in ophthalmology question answering, with DeepSeek-R1 achieving the highest ACC. Our findings demonstrate that reasoning LLM can better simulate human-like thinking processes compared with conventional non-reasoning LLM, suggesting its potential for more trustworthy LLM systems in ophthalmology.

Food saving is a pressing global challenge, and understanding its divergent framings is vital for designing effective strategies. This study examines the alignment and divergence between media discourse and individual attitudes toward food saving in China using a mixed-methods approach that combines content analysis of 123 news articles with Q-methodology involving 60 participants. Drawing on Boltanski and Thévenot’s framework of seven value orders, the research identifies “domesticity” as a core value emphasized by both media and individuals, while media narratives underrepresent values such as “the environment” and “inspiration,” which are central to individual perspectives. This misalignment highlights why awareness campaigns often fail to effect substantial behavioral change. By analyzing the interplay between media narratives and public attitudes, the study not only provides actionable insights for policymakers and educators in China, but also offers a transferable model for interpreting frame (mis)alignment in other rapidly modernizing, state-led contexts. The proposed multi-level explanatory model contributes to international food policy scholarship by advancing the theoretical understanding of how value orders shape, and sometimes constrain, effective interventions in food saving governance.

Large-span spatial lattice structures generally have characteristics such as incomplete modal information, high modal density, and high degrees of freedom. To address the problem of misjudgment in the damage detection of large-span spatial structures caused by these characteristics, this paper proposed a damage identification method based on time series models. Firstly, the order of the autoregressive moving average (ARMA) model was selected based on the Akaike information criterion (AIC). Then, the long autoregressive method was used to estimate the parameters of the ARMA model and extract the residual sequence of the autocorrelation part of the model. Furthermore, principal component analysis (PCA) was introduced to reduce the dimensionality of the model while retaining the characteristic values. Finally, the Mahalanobis distance (MD) was used to construct the damage sensitive feature (DSF). The dome of Taiyuan Botanical Garden in China is one of the largest non-triangular timber lattice shells worldwide. Relying on the structural health monitoring (SHM) project of this structure, this paper verified the effectiveness of the damage identification model through numerical simulation and determined the damage degree of the dome structure through SHM measurement data. The results demonstrated that the proposed damage identification method can effectively identify the damage of large-span timber lattice structures, locate the damage position, and estimate the degree of damage. The constructed DSF had relatively strong robustness to small damage and environmental noise and has practical application value for SHM in engineering.

In patients with cirrhosis, differences between acute kidney injury (AKI) at the time of hospital admission (community-acquired) and AKI occurring during hospitalization (hospital-acquired) have not been explored. We aimed to compare patients with hospital-acquired AKI (H-AKI) and community-acquired AKI (C-AKI) in a large, prospective study. Hospitalized patients with cirrhosis were enrolled (N = 519) and were followed for 90 days after discharge for mortality. The primary outcome was mortality within 90 days; secondary outcomes were the development of de novo chronic kidney disease (CKD)/progression of CKD after 90 days. Cox proportional hazards and logistic regressions were used to determine the independent association of either AKI for primary and secondary outcomes, respectively. H-AKI occurred in 10%, and C-AKI occurred in 25%. In multivariable Cox models adjusting for significant confounders, only patients with C-AKI had a higher risk for mortality adjusting for model for end-stage liver disease-Na: (hazard ratio 1.64, 95% confidence interval [CI] 1.04-2.57, P = 0.033) and adjusting for acute on chronic liver failure: (hazard ratio 2.44, 95% CI 1.63-3.65, P < 0.001). In univariable analysis, community-acquired-AKI, but not hospital-acquired-AKI, was associated with de novo CKD/progression of CKD (odds ratio 2.13, 95% CI 1.09-4.14, P = 0.027), but in multivariable analysis, C-AKI was not independently associated with de novo CKD/progression of CKD. However, when AKI was dichotomized by stage, C-AKI stage 3 was independently associated with de novo CKD/progression of CKD (odds ratio 4.79, 95% CI 1.11-20.57, P = 0.035). Compared with H-AKI, C-AKI is associated with increased mortality and de novo CKD/progression of CKD in patients with cirrhosis. Patients with C-AKI may benefit from frequent monitoring after discharge to improve outcomes.

To explore alterations in tears and ocular surface throughout Descemet's Membrane Endothelial Keratoplasty (DMEK), thereby offering valuable insights for clinical intervention. This study was conducted on 84 patients (84 eyes) with CED and received DMEK who visited the Zhongshan Ophthalmic Center of Sun Yat-sen University from April 2022 to January 2024. All patients were evaluated preoperatively and 1, 3, 6, and 12 months postoperatively, with 31 patients (31 eyes) post-24 months above. This cohort was then compared against a normal control of 21 individuals (21 eyes). Main clinical outcomes included lipid layer thickness (LLT), blink counts and partial blink rate (PBR) measured by Lipiview interferometer, in vivo confocal microscopy (IVCM) was used to detect the density of epithelial wing cells, leukocytes, subbasal dendritic cells (DCs) and endothelial cells. Compared with normal control, LLT was decreased at 6 months (P = 0.053); total blink counts were significantly higher at 36 months (P < 0.05) after DMEK, while LLT was significantly lower (P < 0.05). Utilizing IVCM, an increased leukocytes density was identified in CED patients, which correlated positively with Log MAR visual acuity (r = 0.261, P = 0.04) and inversely with epithelial wing cells density (r = -0.319, P = 0.009). Leukocytes density decreased to comparable level at postoperative 3 months (P = 0.11). Subbasal DCs density was still higher at 6 months postoperatively (P < 0.05), while there was no difference at 12 months after DMEK (P > 0.05). Abnormalities of LLT can occur at different stages after DMEK. Besides, the immune reaction on cornea is still active within 6 months after DMEK.

IntroductionResveratrol is an immune modulator that can reduce M1 macrophage polarization in vitro . Reducing macrophage recruitment and M1 polarization can prevent corneal allograft rejection (CGR). In this study, rat corneal allograft rejection models were established to explore the effects of resveratrol on CGR and macrophages and the underlying mechanisms after corneal transplantation.MethodsCorneal allograft models were established, and 100 mg/kg resveratrol was injected intraperitoneally. The corneal allografts were assessed clinically using the Holland rejection scoring system, anterior segment photography, and anterior segment optical coherence tomography. Corneal macrophages, pro-inflammatory cytokines, and corneal lymphatic vessels were detected using hematoxylin and eosin staining, immunofluorescence staining, and real-time quantitative polymerase chain reaction (qRT-PCR). Dendritic cells (DCs) in cervical lymph nodes were explored using flow cytometry. RNA sequencing experiments were conducted to identify the mechanisms through which resveratrol affected CGR. The results were verified using Simple Western analysis. Pro-inflammatory cytokines by macrophages in vitro were measured using qRT-PCR and enzyme-linked immunosorbent assays.ResultsResveratrol significantly prolonged the survival of corneal grafts and reduced graft edema and central corneal thickness. Corneal macrophage recruitment and M1 macrophage polarization decreased significantly after corneal transplantation in the resveratrol group. Resveratrol also reduced pro-inflammatory cytokines in corneal grafts and suppressed the early generation of cornea lymphatic vessels and the recruitment of cornea inflammatory cells 14 days after surgery. Resveratrol decreased the proportion of DCs in ipsilateral cervical lymph nodes. The effect of resveratrol on CGR was related to the phosphatidylinositol 3-kinase/protein kinase-B (PI3K/Akt) pathway. Resveratrol reduced the secretion of pro-inflammatory cytokines by M1 macrophages in vitro .ConclusionOur findings suggest that resveratrol can reduce corneal macrophage recruitment and M1 macrophage polarization after corneal transplantation in rats and prevent CGR. The PI3K/Akt pathway may be an important mechanism that warrants further research.

Background To explore the molecular genetic cause of a four-generation autosomal dominant congenital cataract family in China. Methods Targeted region sequencing was performed to screen for the potential mutation, and Sanger sequencing was used to confirm the mutation. The homology model was constructed to identify the protein structural change, PolyPhen-2 and Provean were used to predict the mutation impact. Functional and cellular analysis of the wild and mutant GJA8 were performed in DF-1 cells by western blotting, dye uptake assay, immunofluorescence, Annexin V-FITC staining. Results A novel heterozygous mutation (c.205G > A; p.Ala69Thr) was identified within GJA8 , which cosegregated with congenital cataract phenotype in this family. Bioinformatics analysis showed the mutation was located in a highly conserved region, and the mutation was predicted to be pathogenic. Function analysis indicated that the mutation inhibited GJA8 hemichannel activity, reduced cell tolerance to oxidative stress, changed the protein distribution pattern and inhibited the cell growth. Conclusions We have identified a novel missense mutation in GJA8 (c.205G > A, p.Ala69Thr) in a four-generation Chinese family and our results will further broaden the gene mutation spectrum of GJA8 .

Background. The aim of this study was to compare the daily costs and cost effectiveness of fixed combination glaucoma drugs in China. Methods. This study included the following fixed combination drugs: brinzolamide 1% and timolol 0.5% (Azarga; Alcon, Inc., Fort Worth, TX, USA), travoprost 0.004% and timolol 0.5% (DuoTrav; Alcon, Inc.), bimatoprost 0.03% and timolol 0.5% (Ganfort; Allergan, Inc., Dublin, Ireland), and latanoprost 0.005% and timolol 0.5% (Xalacom; Pfizer, Inc., New York, NY, USA). Five bottles of each drug were measured. The mean actual volume, mean actual number of drops, volume per drop, daily cost, yearly cost, and per mmHg reduction cost for each drug were calculated. Results. The volumes per drop ranged from 32.61 ± 2.90 μl (DuoTrav) to 24.38 ± 0.23 μl (Ganfort). The number of usage days per bottle varied from 36 days (DuoTrav) to 61 days (Ganfort). Azarga had the lowest daily cost ($0.23) and yearly cost ($84.72), while DuoTrav had the highest daily cost ($0.79) and yearly cost ($287.02). Azarga costed $2.17–$3.30 per mmHg intraocular pressure reduction, which was lower than the other three drugs. For the prostaglandin and ß-adrenergic blocker FCs, Ganfort had the lowest daily cost ($0.35) and per mmHg reduction cost (from $3.40 to $4.04). Conclusions. The daily costs of these drugs were significantly different, with Azarga having the lowest daily cost and best cost effectiveness. For the prostaglandin and β-adrenergic blocker fixed combinations, Ganfort was the most economical choice with its lower daily cost and per mmHg reduction cost. The results of this study could provide drug selection guidance from an economic perspective, but various factors should be considered when making a decision.

Background To introduce a novel protocol to treat refractory acute primary angle closure (APAC): transscleral cyclophotocoagulation (TCP) followed by cataract surgery. Methods Thirteen APAC eyes (13 patients) were enrolled in this prospective case series as study group. All patients underwent emergency TCP (20 pulses of 2000 mW during 2000 ms applied to the inferior quadrant) followed by scheduled cataract surgery. They were compared to 13 age- and gender-matched patients treated with emergency phacotrabeculectomy. We recorded intraocular pressure (IOP), best corrected visual acuity (BCVA), and complications, and several ultrasound biomicroscopy (UBM) parameters before and after TCP. Results In the study group, IOP decreased from 51.5 ± 7.0 mmHg (mean ± standard deviation) before TCP to 16.4 ± 5.4 mmHg 1 day after TCP ( P  < 0.001). At 6 months, there was no significant difference in IOP between the study group (14.0 ± 3.4 mmHg) and control group (16.7 ± 4.3 mmHg; P  = 0.090); IOP lowering medications were used by 0/13 in the study group and 2/13 patients in the control group ( P  = 0.48). At 6 months, there was no significant difference in BCVA between the study group and the control group (20/25 (20/200 to 20/25) and 20/30 (20/50 to 20/25), respectively; P  = 1.0). The UBM parameters anterior chamber depth ( P  = 0.016), angle-opening distance at 500 μm ( P  = 0.011), and maximum ciliary body thickness ( P  < 0.001) increased significantly while the iris-ciliary process distance decreased significantly ( P  = 0.020) after TCP. Conclusions TCP effectively lowers IOP and modifies the anterior chamber morphology in APAC; TCP followed by cataract surgery can be considered an alternative to treat refractory APAC but needs further evaluation. Trial registration This project was registered in Chinese Clinical Trial Registry ( ChiCTR1800017475 ) at July, 31, 2018 ( http://www.chictr.org.cn/edit.aspx?pid=29629&htm=4 ).

Aiming at the damage misjudgment caused by incomplete modal information and unclear physical meaning of parameters in damage identification of long-span spatial structures, a damage identification method based on time-series analysis model established from measured data of structural health monitoring system is proposed. First, the selection method and modeling process of time-series analysis model based on acceleration response data are illustrated through the theoretical derivation. Subsequently, the damage sensitive feature (DSF) constructed by autoregressive parameters and Mahalanobis distance (MD) discriminant function is proposed for structural damage identification, which is dimension reduced by principal component analysis (PCA). Finally, based on the long-span steel structure in Taiyuan Botanical Garden, the effectiveness of damage identification method is studied by numerical simulation method, and the current damage status of the structure is identified through the measured data. The results show that the proposed method can effectively identify the structural damage, accurately locate the damage location and reflect the damage degree. The proposed DSF is sensitive to minor structural damage while reducing the amount of calculation. It is suitable for damage identification of complex structures and has application value in the analysis of health monitoring systems measured response data in practical engineering.