Intravitreal anti-vascular endothelial growth factor(anti-VEGF)therapy has been widely used, but the variability in its therapeutic efficacy limits individualized treatment. In recent years, the application of artificial intelligence(AI)has opened up new avenues for personalized treatment response prediction, and its core branches include machine learning(ML)and deep learning(DL). This review systematically retrieved and analyzed 41 relevant studies published up to April 2025. Comprehensive analysis reveals that AI predictive models are evolving from forecasting single endpoints(such as visual acuity or central retinal thickness)to integrating multi-dimensional endpoints(encompassing anatomical, functional, and treatment demand parameters)and generating predictive imaging outputs. In terms of technical approaches, DL models(28 studies, accounting for 68.3%)dominate this field due to their robust image interpretation capabilities, while ML models(10 studies, 24.4%)retain significant value in the analysis of structured clinical data. Cross-disease comparisons indicate that research efforts are most concentrated on age-related macular degeneration(ARMD)and diabetic macular edema(DME), with shared conceptual frameworks for model construction, yet distinct anatomical and functional indicators are prioritized for each disease. Currently, the field confronts several key challenges, including insufficient prospective clinical validation, limited model interpretability(the “black box problem”), and a scarcity of high-quality multi-center datasets. Moving forward, it is imperative to advance real-world validation and develop explainable AI techniques to expedite the clinical translation of these predictive models.

Background Gestational weight gain is closely related to maternal and infant health outcomes. Pregnant women are simultaneously exposed to four metals—lithium (Li), vanadium (V), uranium (U), and bismuth (Bi)—through inhalation of fine particulate matter and consumption of contaminated food and water. Existing studies suggest that exposure to these metals may be associated with gestational weight gain. However, no study has yet explored the complex relationships between exposure to mixtures of these four metals and weight gain at different stages of pregnancy. Objective To investigate the associations between mixed exposure to Li, V, U, and Bi in early pregnancy and the average weekly gestational weight gain during both early pregnancy and mid-to-late pregnancy. Methods This prospective study recruited eligible women in early pregnancy from an obstetrics clinic of a tertiary hospital in Jinan, China, between September 2021 and July 2023. Pre-pregnancy weight, current weight (at 11⁺⁰ to 13⁺⁶ weeks of gestation), and spot urine samples (≥5.0 mL) were collected at enrollment. Urinary concentrations of Li, V, Bi, and U were determined using inductively coupled plasma mass spectrometry. Participants were followed up in late pregnancy (≥28 weeks of gestation) to collect information on physical activity via questionnaire; weight measurements at the last antenatal visit (35⁺⁰ to 37⁺⁶ weeks of gestation) were obtained from the hospital information system. After adjusting for covariates, multiple linear regression and generalized additive models were used to assess the associations of individual metals with weekly weight gain in early pregnancy and in mid-to-late pregnancy. Bayesian kernel machine regression (BKMR) and quantile-based g-computation (Qgcomp) were applied to evaluate the joint effects of the metal mixture exposure on weekly weight gain at the two gestational stages. Results A total of 313 pregnant women were included. The geometric means of urinary Li, V, U, and Bi concentrations were 37.07, 0.20, 0.06, and 0.04 μg·L⁻¹, respectively; after creatinine adjustment, the corresponding values were 46.82, 0.25, 0.07, and 0.05 μg·g⁻¹ (Cr). The mean weekly gestational weight gain was (0.19±0.25) kg in early pregnancy and (0.53 ± 0.18) kg in mid-to-late pregnancy. Both multiple linear regression and generalized additive models showed that urinary V concentration was positively associated with average weekly gestational weight gain in early pregnancy, while no significant associations were found for other metals or for gestational weight gain in mid-to-late pregnancy. In the BKMR model with early-pregnancy weight gain as the outcome, V had the strongest association [posterior inclusion probability (PIP)=0.773]. When other metals were fixed at their medians, V showed a positive non-linear association with the outcome. A significant single-metal effect of V and its interaction with Li were observed. Compared with the 50th percentile of the metal mixture, the average weekly weight gain in early pregnancy increased by 0.016 (95%CI: 0.003, 0.029) and 0.018 (95%CI: 0.001, 0.036) at the 60th and 65th percentiles, respectively; conversely, at the 25th percentile, it decreased by 0.026 (95%CI: 0.002, 0.050). Overall, the joint effect of the metal mixture on early- pregnancy weight gain showed an upward trend. In the BKMR model for mid-to-late pregnancy gestational weight gain, all PIPs were<0.5, and no significant single-metal effects, interactions, or joint effects were identified. Qgcomp results confirmed a positive association between the metal mixture and early-pregnancy weight gain (b=0.031, 95%CI: 0.010, 0.051; P<0.01), with V contributing the highest positive weight (0.71). No significant association was found for weight gain in mid-to-late pregnancy (b=0.007, P=0.339). Conclusion Higher levels of co-exposure to the Li, V, Bi, and U metal mixture during early pregnancy may be associated with increased average weekly weight gain in early pregnancy. Among these metals, V exhibits a predominant role and appears to interact with Li. No association is observed between early-pregnancy metal mixture exposure and average weekly gestational weight gain in mid-to-late pregnancy. These findings suggest that monitoring and managing metal exposure during early pregnancy may be crucial for the rational regulation of gestational weight gain.

Objective:To analyze the correlation between sex hormone levels and micropenis after mild hypospadias surgery in children.Methods:A case control study was carried out.The clinical data of 71 children aged 1 to 13 years who underwent mild hypospadias surgery at the First Affiliated Hospital of Xinxiang Medical University from April 2022 to April 2024 were analyzed.Preoperatively, the children were divided into a mild hypospadias group (Group A) and a mild hypospadias with micropenis group (Group B) based on the stretched penile length (SPL).Prolactin, follicle-stimulating hormone (FSH), luteinizing hormone (LH), progesterone, and testosterone (TES) levels of the 2 groups were measured.Multivariate Logistic regression analysis was used to construct a risk prediction model.The discrimination capability of the model was evaluated using the receiver operating characteristic (ROC) curve.SPL and sex hormone levels were measured again 6 months after surgery.The children were divided into a normal penile group (Group AA) and a micropenis group (Group BB) after mild hypospadias according to SPL.Multivariate Logistic regression analysis was used to construct a risk prediction model, which was evaluated using the ROC curve.Results:The levels of FSH, LH and TES in group A before the operation were 3.28(2.02, 4.46) IU/L, 0.53(0.25, 0.79) IU/L and 25.24(17.94, 36.67) ng/dL, respectively, and those in group B were 1.42(1.10, 1.84) IU/L, 0.14(0.09, 0.23) IU/L and 15.73 (12.92, 17.00) ng/dL, respectively.The difference was all statistically significant (all P<0.05).Multivariate Logistic regression analysis showed statistical significance ( OR=0.515, 95% CI: 0.271-0.977; OR=0.035, 95% CI: 0.002-0.542; OR=0.883, 95% CI: 0.796-0.980).The area under the ROC curve (AUC) of the prediction model was 0.906, with a sensitivity of 75.00% and a specificity of 95.74%.The levels of FSH, LH and TES in the postoperative AA group were 2.97 (1.88, 4.28) IU/L, 0.46 (0.23, 0.78) IU/L and 20.92 (17.34, 33.27) ng/dL, respectively.The median levels of FSH, LH and TES in the BB group were 1.52 (1.27, 1.82) IU/L, 0.17 (0.12, 0.26) IU/L and 15.08(11.68, 16.68) ng/dL, respectively.The difference was all statistically significant (all P<0.05).Multivariate Logistic regression analysis showed statistical significance ( OR=0.484, 95% CI: 0.236-0.992; OR=0.061, 95% CI: 0.004-0.939; OR=0.891, 95% CI: 0.795-0.999).The AUC of the prediction model constructed was 0.877, with a sensitivity of 94.12% and a specificity of 68.52%. Conclusions:Lower FSH, LH and TES levels are risk factors for the micropenis after mild hypospadias surgery, and preoperative hormone levels have higher predictive value.

Objective:To investigate the value of MULTI-parametric MR imaging with flexible design (MTP) combined with microview diffusion weighted imaging (DWI) in predicting pituitary tumor invasion of cavernous sinus.Methods:This was a cross-sectional study. The clinicopathological and imaging data of 59 patients with suspected pituitary tumor who were admitted to the First Affiliated Hospital of Dalian Medical University from December 2021 to August 2024 were prospectively analyzed. According to Knosp classification, patients with pituitary tumor were divided into non-invasive group (28 cases) and invasive group (31 cases). All patients underwent MTP and microview DWI sequence scanning, and the corresponding parameter values [T 1, R 2*, T 2* and apparent diffusion coefficient (ADC)] were measured on the MTP quantitative map and ADC map at the maximum diameter level of the pituitary tumor. Independent sample t test and Mann-Whitney U test were used to compare the clinical data, quantitative parameter values and proliferation index (Ki-67) between the non-invasive group and the invasive group. Quantitative parameter values with statistically significant differences were jointly included in logistic regression to construct the model. Receiver characteristic curve and area under the curve (AUC) were used to analyze single variables and logistic model in predicting the value invasion of pituitary tumor cavernous sinus invasion. Spearman correlation was used to analyze the correlation between the quantitative values and Ki-67 level. Results:There was no significant difference in age and gender between the non-invasive group and the invasive group ( P>0.05). T 1 and T 2* values in invasion group were higher than those in non-invasion group ( Z=4.42, P<0.001; Z=3.43, P<0.001), R 2* values and ADC values in the invasion group were lower than those in the non-invasion group, and the difference was statistically significant (t=4.07, P<0.001; Z=-3.07, P=0.002). The AUC values of T 1, R 2*, T 2*, ADC, logistic model 1(T 1+R 2*+ADC) and logistic model 2(T 1+T 2*+ADC) for predicting pituitary tumor cavernous sinus invasion were 0.835, 0.767, 0.760, 0.733, 0.926 and 0.925, respectively. Correlation analysis showed that the value of pituitary tumor T 1 was positively correlated with Ki-67 ( r=0.431, P=0.002). Conclusion:Quantitative parameters of MTP sequence and microview DWI sequence can quantitatively evaluate pituitary tumor invasion of cavernous sinus, and the combination of the two sequences can significantly improve the prediction efficiency.

Natural products (NPs) have long held a significant position in various fields such as medicine, food, agriculture, and materials. The chemical space covered by NPs is extensive but often underexplored. Therefore, high-throughput and efficient methodologies for the annotation and discovery of NPs are desired to address the complexity and diversity of NP-based systems. Mass spectrometry (MS) has emerged as a powerful platform for the annotation and discovery of NPs. MS databases provide vital support for the structural characterization of NPs by integrating extensive mass spectral data and sample information. Additionally, the released annotation methodologies, based on a variety of informatics tools, continuously improve the ability to annotate the structure and properties of compounds. This review examines the current mainstream databases and annotation methodologies, focusing on their advantages and limitations. Prospects for future technological advancements are then discussed in terms of novel applications and research objectives. Through a systematic overview, this review aims to provide valuable insights and a reference for MS-based NPs annotation, thereby promoting the discovery of novel natural entities.
Biological Products/chemistry* ; Mass Spectrometry/methods* ; Databases, Factual ; Drug Discovery/methods* ; Humans

[Objective]To investigate the role of endothelial microparticles(EMPs)in diabetic macrovascular complications(DMCs),summarize the research progress and analyze the research progress on traditional Chinese medicine(TCM)regulation of EMPs for the prevention and treatment of DMCs.[Methods]Base on the research on the pathological mechanism of DMCs,the biological characteristic of EMPs and their mechanism of action in DMCs in recent years,combined with the experimental and clinical studies on the regulation of EMPs by TCM for disease prevention and treatment,their paper summarizes the action targets of EMPs in regulating the pathological process of DMCs and the resarch progress of TCM intervention,on this basis,the limitations of current research and future prospects are summarized.[Results]EMPs are important biomarkers of endothelial injury,promoting mechanisms such as inflammation,oxidative stress,thrombosis and atherosclerosis,contributes to the pathological progression of DMCs.TCM intervenes in the generation and release of EMPs through mechanisms such as regulating inflammatory responses,alleviating oxidative stress,relieving endothelial dysfunction and promoting vascular repair.This approach helps maintain endothelial homeostasis and plays a crucial role in the prevention and treatment of DMCs.[Conclusion]EMPs play a significant role in the onset and progression of DMCs,serving as potential intervention targets.TCM offers novel therapeutic approaches for the prevention and treatment of DMCs by inhibiting inflammation and oxidative stress,promoting vascular remodeling and repair,and suppressing the formation of thrombosis and atherosclerosis.

Objective:Investigating the association of the systemic immune-inflammation index and the pan-inflammation index with incident non-alcoholic fatty liver disease (NAFLD).Methods:This retrospective cohort study included 42 891 participants who underwent at least two health examinations at the Second Affiliated Hospital of Dalian Medical University between 2014 and 2023. Based on their levels of the systemic immune inflammation index (SII) and the pan immune inflammation value (PIV), participants were respectively divided into four quartile groups (Q1 to Q4). Cox proportional hazards models were employed to analyze the association of different SII and PIV levels, as well as their quartile groups, with new onset NAFLD in the total population and across various subgroups. Restricted cubic splines were used to examine the dose response relationship between these inflammatory indices and incident NAFLD. Additionally, sensitivity analyses were conducted to confirm the robustness of the findings.Results:After adjusting for confounding factors, the natural logarithm-transformed lnSII ( HR=1.247, 95% CI: 1.184-1.314, P<0.001) and lnPIV ( HR=1.192, 95% CI: 1.148-1.238, P<0.001) were significantly positively associated with the risk of NAFLD. When the subjects were grouped by SII quartiles (Q1-Q4), compared with those in Q1, participants in Q2, Q3, and Q4 exhibited progressively higher risks of incident NAFLD:11.9% ( HR=1.119, 95% CI: 1.051-1.192, P<0.001), 17.1% ( HR=1.171, 95% CI: 1.100-1.248, P<0.001), and 29.1% ( HR=1.291, 95% CI: 1.211-1.377, P<0.001), respectively. Quartile analysis of PIV yielded similar trends: the risk of incident NAFLD increased for 10.4% ( HR=1.104, 95% CI: 1.034-1.179, P=0.003), 18.7% ( HR=1.187, 95% CI: 1.112-1.266, P<0.001), and 30.5% ( HR=1.305, 95% CI: 1.223-1.393, P<0.001) in Q2, Q3, and Q4 group respectively when compared to that in Q1 group. Subgroup analysis confirmed consistent associations of SII and PIV with elevated NAFLD risk across all subgroups. Conclusion:Elevated levels of SII and PIV are significantly associated with increased risk of NAFLD.

Objective:To investigate transition pattern of health status among middle-aged and elderly population in China based on frailty index.Methods:In this retrospective cohort study, middle-aged and elderly people were selected from the China Health and Retirement Longitudinal Study (CHARLS) in 2011; and 1 434 subjects were followed up to 2015. The frailty index was calculated from the prevalence of chronic diseases, daily activity ability and blood biomarkers, and the frailty state was divided by quartiles of the frailty index. Markov models were constructed to determine the transition probabilities of different frailty states.Results:The mean age of the 1 434 subjects was (59.0±9.4) years and the mean frailty index was 0.11±0.05. In the healthy individuals, 63.0% remained healthy after a four-year follow-up; during the same follow-up period, 40.9% of the mildly frail individuals and 23.0% of the moderately frail individuals remained in their baseline frailty status. Increasing age leaded to a gradual increase in the probability of the population shifting to a severely frailty state. Women were more likely to shift to severe frailty status than men (0.029 vs 0.019, Z=3.03, P=0.002). Conclusion:Among middle-aged and elderly population in China, the transition of health states follows a pattern where higher frailty levels are associated with lower stability. Advanced age and female gender are identified as risk factors for progression to severe frailty.

Objective To analyze the risk factors of fluconazole resistance in Candida tropicalis(C.tropicalis)urinary tract infection(UTI),and evaluate the efficacy of different treatment regimens.Methods Patients with C.tropicalis UTI at Xiangya Hospital of Central South University from January 2021 to December 2023 were in-cluded for single center retrospective study.The minimum inhibitory concentration(MIC)of fluconazole was deter-mined by microbroth dilution.Patients were divided into a fluconazole-resistant group and a fluconazole-sensitive group based on fluconazole resistance.Risk factors for fluconazole resistance were analyzed based on clinical data,and therapeutic efficacy in patients in fluconazole-resistant group was analyzed.Results A total of 198 patients were included in the study.133(67.2％)C.tropicalis strains were detected to be sensitive to fluconazole,while 65(32.8％)strains were resistant,and 63.1％(n=41)had MIC values ≥128 μg/mL.Compared with fluconazole-sensitive group,fluconazole-resistant group had a higher proportion of pulmonary infection(P=0.019).Pulmonary infection(OR=3.282)was a risk factor for fluconazole resistance in C.tropicalis UTI,while urinary system ob-struction(OR=0.269)was a protective factor for fluconazole resistance in C.tropicalis UTI.There was no statis-tically significant difference in the usage rate of different antimicrobial agent types between the two groups(all P＞0.05).The therapeutic efficacy analysis showed that the effective rates of treatment with fluconazole dosage regi-mens of ≤200 mg/d,≥400 mg/d,and fluconazole monotherapy against fluconazole-resistant strains were 66.7％(6/9),83.3％(5/6),and 100％(6/6),respectively.For patients treated with monotherapy using other drugs or with multidrug sequential treatment regimens,the treatment effective rate was 60.0％(3/5).The proportion of pa-tients in the effective treatment group who removed their urinary catheters after detecting C.tropicalis was higher than that in the ineffective treatment group(P＜0.001).Conclusion The fluconazole resistance of C.tropicalis is related to urinary tract obstruction and concurrent pulmonary infection.When treating UTI caused by fluconazole-resistant strains,the catheter should be removed as early as possible.In addition to increasing the dosage of flucon-azole,other antifungal drugs such as flucytosine alone or sequential treatment with multiple drugs can also be con-sidered.

Objective:To develop a rapid assessment scale for healthy aging suitable for the Chinese population.Methods:Based on existing healthy aging assessment scales, national standards, and expert consensus, an initial Healthy Aging Rapid Assessment Scale was drafted through two rounds of expert consultation. A pre-survey was conducted with 3 220 subjects recruited from Guangzhou between July 2023 and July 2024. Items were screened through item analysis and exploratory factor analysis to form the final scale. Reliability and validity of the final scale were validated across five cities: Guangzhou, Dongguan, Shenzhen, Baoding, and Chuxiong.Results:The initial version comprised 36 items, while the finalized scale contained 18 items across three dimensions: metabolic health, mental health, and cognitive health. Test-retest reliability ranged from 0.71 to 0.81 across all study sites. The Spearman-Brown coefficient varied between 0.91-0.96, Cronbach′s α between 0.77-0.83, comparative fit index (CFI) between 0.90-0.98, goodness-of-fit index (GFI) between 0.90-0.99, and root-mean-square error of approximation (RMSEA) between 0.03-0.09. For the three dimensions, reliability and validity metrics demonstrated consistency: Spearman-Brown coefficients 0.87-0.99, Cronbach′s α 0.77-0.83, CFI 0.90-0.98, GFI 0.90-0.99, and RMSEA 0.03-0.09 across four regions.Conclusion:The developed Healthy Aging Rapid Assessment Scale for the Chinese population exhibits robust reliability and validity.