ObjectiveThe widespread adoption of portable fundus cameras for primary care and community screening is hindered by limitations in current autofocus(AF) technologies. Image-based methods relying on sharpness evaluation require iterative searches, resulting in slow convergence, while projection-based techniques are susceptible to optical artifacts and calibration errors. To address these challenges, this study introduces a novel AF system based on direct wavefront sensing, designed to deliver simultaneous high speed, high precision, and operational robustness within the compact form factor essential for portable ophthalmic devices. MethodsOur approach fundamentally reimagines the AF process by directly measuring the ocular wavefront aberration. We developed a custom portable fundus camera integrating a miniaturized Shack-Hartmann wavefront sensor (SHWS) into the optical path. An 850 nm laser diode projects a point source onto the retina via oblique illumination to minimize corneal reflections. Light scattered from this spot carries the eye’s refractive error through the imaging optics and is directed to the SHWS, positioned at a plane optically conjugate to the primary color CMOS imaging sensor. A microlens array within the SHWS samples the incident wavefront, generating a pattern of focal spots on a CCD. Real-time centroid analysis of these spots provides a map of local wavefront slopes. These measurements are processed through a singular value decomposition (SVD) algorithm to fit a Zernike polynomial basis set, enabling real-time reconstruction of the wavefront phase. The defocus component (S) is extracted from the second-order Zernike coefficients, providing a direct, quantitative measure of the refractive error in diopters. This value serves as a precise error signal in a closed-loop control system, which commands a voice-coil actuated focusing lens to its null position in a single, deterministic step, eliminating the need for iterative search algorithms. ResultsComprehensive evaluation demonstrated the system’s high performance. Testing on a calibrated model eye (OEMI-7) established a highly linear relationship between the computed defocus S and the focusing lens position across a ±20 Diopter (D) compensation range, achievable within a 5 mm mechanical travel. The system achieved a focusing precision of 0.08 D, corresponding to an 18-fold improvement over a conventional projection spot-size method tested under identical conditions. The total focus acquisition time, encompassing wavefront measurement, computation, and lens actuation, averaged under 0.5 s. Clinical validation with 25 human volunteers (50 eyes, refractive range -15 D to +10 D) confirmed practical efficacy. The wavefront-sensing AF succeeded in 92% of attempts with a mean time of 0.5 s, substantially outperforming a projection-based benchmark which achieved only a 32% success rate with an average time of 4.25 s. The system provided instantaneous directional guidance and maintained stability during minor ocular movements. Objective assessment of image quality, via amplitude contrast of retinal vasculature, showed consistent and significant enhancement following AF correction across the entire tested diopter range. ConclusionThis work successfully implements and validates a direct wavefront-sensing autofocus paradigm for portable fundus cameras. By directly quantifying and compensating for the optical defocus aberration, this method bypasses the fundamental limitations of image-processing and projection-based techniques, enabling rapid, precise, and deterministic diopter compensation. The developed system delivers an exceptional combination of a wide operational range (±20 D), high accuracy (0.08 D), fast convergence (0.5 s), and a compact physical footprint. This technology provides a practical and high-performance focusing solution capable of enhancing the reliability, throughput, and diagnostic utility of portable retinal imaging in large-scale screening applications. Future efforts will be directed towards system cost optimization and performance adaptation for diverse ocular conditions.

ObjectiveDiscriminating atypical hepatocellular carcinoma (HCC) from other malignancies in liver nodules classified as Liver Imaging Reporting and Data System category M (LR-M) remains a significant diagnostic challenge on conventional ultrasound examination. The LR-M category, originally intended to capture non-HCC malignancies, paradoxically contains up to 63% of atypical HCCs that deviate from classic enhancement patterns, leading to potential misdiagnosis and suboptimal treatment planning. While deep learning has shown promise in HCC diagnosis, most existing models rely exclusively on single-modality ultrasound, overlooking the diagnostic benefits of integrating complementary information from multiple imaging sources. To address this gap, we propose a novel attention-weighted tri-modal ultrasound network (TUS-Net) that integrates contrast-enhanced ultrasound (CEUS), B-mode ultrasound (BUS), and time-intensity curves (TICs) to improve diagnostic accuracy for these clinically challenging lesions. MethodsOur framework incorporates a three-dimensional convolutional neural network (C3D) backbone to extract spatiotemporal features from CEUS videos, capturing dynamic vascular patterns critical for lesion characterization. To effectively fuse complementary modalities, we introduce a dual-channel feature fusion module (DCFFM) that adaptively combines features from CEUS and BUS through channel-wise attention mechanisms, allowing the model to dynamically weigh the contribution of each modality based on diagnostic relevance. Additionally, we propose a temporal intensity feature fusion module (TIFFM) that leverages quantitative hemodynamic information from TICs to guide the model’s attention toward diagnostically critical temporal phases, such as arterial wash-in and portal venous washout. The model is further enhanced by automated lesion localization using YOLOX and class activation mapping for interpretability, ensuring that predictions align with clinically meaningful imaging features. ResultsEvaluated on a tri-modal ultrasound dataset comprising 161 patients with pathologically confirmed LR-M nodules (131 atypical HCC and 30 non-HCC malignancies), our model achieved an accuracy of 86.83%, a sensitivity of 92.50%, a specificity of 75.50%, and an AUC of 89.32% in screening atypical HCC. Compared to single-modality baselines, TUS-Net demonstrated superior specificity, a clinically critical metric given the higher risk associated with misclassifying non-HCC malignancies. Ablation studies confirmed the contribution of each module, with the full model outperforming both standard C3D and 3D ResNet backbones integrated with attention mechanisms. A reader study involving junior and senior radiologists further validated the clinical utility of AI assistance, showing consistent improvements in specificity and inter-reader consistency, particularly for less experienced clinicians. ConclusionThese results surpass existing benchmark models and demonstrate the potential of our approach to enhance diagnostic precision in clinically specific cases. By intelligently fusing multi-modal ultrasound data with attention-guided mechanisms, TUS-Net offers a reliable and interpretable tool that holds promise for improving the non-invasive diagnosis of atypical HCC in challenging LR-M liver nodules.

ObjectiveDiscriminating atypical hepatocellular carcinoma (HCC) from other malignancies in liver nodules classified as Liver Imaging Reporting and Data System category M (LR-M) remains a significant diagnostic challenge on conventional ultrasound examination. The LR-M category, originally intended to capture non-HCC malignancies, paradoxically contains up to 63% of atypical HCCs that deviate from classic enhancement patterns, leading to potential misdiagnosis and suboptimal treatment planning. While deep learning has shown promise in HCC diagnosis, most existing models rely exclusively on single-modality ultrasound, overlooking the diagnostic benefits of integrating complementary information from multiple imaging sources. To address this gap, we propose a novel attention-weighted tri-modal ultrasound network (TUS-Net) that integrates contrast-enhanced ultrasound (CEUS), B-mode ultrasound (BUS), and time-intensity curves (TICs) to improve diagnostic accuracy for these clinically challenging lesions. MethodsOur framework incorporates a three-dimensional convolutional neural network (C3D) backbone to extract spatiotemporal features from CEUS videos, capturing dynamic vascular patterns critical for lesion characterization. To effectively fuse complementary modalities, we introduce a dual-channel feature fusion module (DCFFM) that adaptively combines features from CEUS and BUS through channel-wise attention mechanisms, allowing the model to dynamically weigh the contribution of each modality based on diagnostic relevance. Additionally, we propose a temporal intensity feature fusion module (TIFFM) that leverages quantitative hemodynamic information from TICs to guide the model’s attention toward diagnostically critical temporal phases, such as arterial wash-in and portal venous washout. The model is further enhanced by automated lesion localization using YOLOX and class activation mapping for interpretability, ensuring that predictions align with clinically meaningful imaging features. ResultsEvaluated on a tri-modal ultrasound dataset comprising 161 patients with pathologically confirmed LR-M nodules (131 atypical HCC and 30 non-HCC malignancies), our model achieved an accuracy of 86.83%, a sensitivity of 92.50%, a specificity of 75.50%, and an AUC of 89.32% in screening atypical HCC. Compared to single-modality baselines, TUS-Net demonstrated superior specificity, a clinically critical metric given the higher risk associated with misclassifying non-HCC malignancies. Ablation studies confirmed the contribution of each module, with the full model outperforming both standard C3D and 3D ResNet backbones integrated with attention mechanisms. A reader study involving junior and senior radiologists further validated the clinical utility of AI assistance, showing consistent improvements in specificity and inter-reader consistency, particularly for less experienced clinicians. ConclusionThese results surpass existing benchmark models and demonstrate the potential of our approach to enhance diagnostic precision in clinically specific cases. By intelligently fusing multi-modal ultrasound data with attention-guided mechanisms, TUS-Net offers a reliable and interpretable tool that holds promise for improving the non-invasive diagnosis of atypical HCC in challenging LR-M liver nodules.

ObjectivePrimary liver cancer, predominantly hepatocellular carcinoma (HCC), is a significant global health issue, ranking as the sixth most diagnosed cancer and the third leading cause of cancer-related mortality. Accurate and early diagnosis of HCC is crucial for effective treatment, as HCC and non-HCC malignancies like intrahepatic cholangiocarcinoma (ICC) exhibit different prognoses and treatment responses. Traditional diagnostic methods, including liver biopsy and contrast-enhanced ultrasound (CEUS), face limitations in applicability and objectivity. The primary objective of this study was to develop an advanced, light-weighted classification network capable of distinguishing HCC from other non-HCC malignancies by leveraging the automatic analysis of brightness changes in CEUS images. The ultimate goal was to create a user-friendly and cost-efficient computer-aided diagnostic tool that could assist radiologists in making more accurate and efficient clinical decisions. MethodsThis retrospective study encompassed a total of 161 patients, comprising 131 diagnosed with HCC and 30 with non-HCC malignancies. To achieve accurate tumor detection, the YOLOX network was employed to identify the region of interest (ROI) on both B-mode ultrasound and CEUS images. A custom-developed algorithm was then utilized to extract brightness change curves from the tumor and adjacent liver parenchyma regions within the CEUS images. These curves provided critical data for the subsequent analysis and classification process. To analyze the extracted brightness change curves and classify the malignancies, we developed and compared several models. These included one-dimensional convolutional neural networks (1D-ResNet, 1D-ConvNeXt, and 1D-CNN), as well as traditional machine-learning methods such as support vector machine (SVM), ensemble learning (EL), k-nearest neighbor (KNN), and decision tree (DT). The diagnostic performance of each method in distinguishing HCC from non-HCC malignancies was rigorously evaluated using four key metrics: area under the receiver operating characteristic (AUC), accuracy (ACC), sensitivity (SE), and specificity (SP). ResultsThe evaluation of the machine-learning methods revealed AUC values of 0.70 for SVM, 0.56 for ensemble learning, 0.63 for KNN, and 0.72 for the decision tree. These results indicated moderate to fair performance in classifying the malignancies based on the brightness change curves. In contrast, the deep learning models demonstrated significantly higher AUCs, with 1D-ResNet achieving an AUC of 0.72, 1D-ConvNeXt reaching 0.82, and 1D-CNN obtaining the highest AUC of 0.84. Moreover, under the five-fold cross-validation scheme, the 1D-CNN model outperformed other models in both accuracy and specificity. Specifically, it achieved accuracy improvements of 3.8% to 10.0% and specificity enhancements of 6.6% to 43.3% over competing approaches. The superior performance of the 1D-CNN model highlighted its potential as a powerful tool for accurate classification. ConclusionThe 1D-CNN model proved to be the most effective in differentiating HCC from non-HCC malignancies, surpassing both traditional machine-learning methods and other deep learning models. This study successfully developed a user-friendly and cost-efficient computer-aided diagnostic solution that would significantly enhances radiologists’ diagnostic capabilities. By improving the accuracy and efficiency of clinical decision-making, this tool has the potential to positively impact patient care and outcomes. Future work may focus on further refining the model and exploring its integration with multimodal ultrasound data to maximize its accuracy and applicability.

Objective To investigate the clinical efficacy and safety of facilitated percutaneous coronary intervention(PCI)with half-dose recombinant staphylokinase(r-SAK)in patients with ST-segment elevation myocardial infarction(STEMI)who are expected to undergo PCI within 120 minutes.Methods From October 2021 to August 2022,a total of 200 STEMI patients in eight centers were included and randomly assigned in a 1﹕1 ratio to either r-SAK group or control group.Patients received loading doses of aspirin and ticagrelor and intravenous heparin and were randomized to receive an intravenous bolus of either 5 mg r-SAK or normal saline prior to PCI.The outcomes were set as ST-segment resolution(STR)at 60-90 minutes after PCI,the proportion and transition of pathological Q waves on the 5th day after PCI,and the proportion of high-sensitivity cardiac troponin T(hs-cTnT)peaking within 12 hours of onset.The safety outcome was major bleeding events defined as Bleeding Academic Research Consortium(BARC)≥type 3 bleeding during hospitalization.Results Compared with the control group,the r-SAK group had a higher proportion of STR≥70%within 60-90 minutes after PCI(58.3%vs.40.3%,P=0.009);a lower proportion of pathological Q waves(59.1%vs.74.1%,P=0.040);a lower rate of Q wave progression(14.8%vs.43.2%,P＜0.001);a higher rate of Q wave disappearance(12.5%vs.3.7%,P=0.027);and a higher proportion of hs-cTnT peaking within 12 hours of symptom onset[31/40(77.5%)vs.17/33(51.5%),P=0.027].Regarding the safety outcome,no significant difference in BARC≥type 3 bleeding was found between the two groups during hospitalization(P＞0.05).Conclusions For STEMI patients who were expected to undergo primary PCI within 120 minutes of symptom onset,the facilitated PCI with half-dose r-SAK significantly increased the proportion of STR≥70%at 60-90 minutes after PCI,reduced the formation of pathological Q waves,and shortened the time to peak hs-cTnT,without increasing the risk of bleeding,which should be an alternative reperfusion strategy worthy of further study.

Objective:To explore the application of the flipped classroom teaching method in pediatric internship, evaluate the feedback from both students and faculty, and provide evidence for optimizing clinical medical education strategies.Methods:Fourth-year students ( n=174) enrolled in 2019 in the eight-year clinical medicine program at Peking University Health Science Center and instructors ( n=42) participated in this study. Questionnaire survey and exit examination scores were used to assess the effectiveness of the flipped classroom method. A statistical analysis was performed using SPSS 26.0. Kruskal-Wallis test and Dunn's multiple comparisons were used to evaluate the differences in satisfaction across teaching components. Chi-square test and Fisher's exact test were applied to compare satisfaction between high-score (top 30%) and low-score (bottom 30%) student groups. Results:①Overall student satisfaction with the flipped classroom reached 91.33% ("very satisfied" + "satisfied"). High satisfaction was reported for pre-class self-learning videos (80.35%), recommended clinical guidelines (82.80%), and English literature (71.10%), while satisfaction with the flipped classroom lectures was comparatively low (52.60%), with significant differences compared to other components ( P<0.05). ②High-score students exhibited significantly lower satisfaction than low-score students across all components (overall evaluation, 84.61% vs. 98.08%, Fisher's exact test, P=0.031; pre-class videos, 53.85% vs. 76.92%, χ2=6.12, P=0.013; preparatory assignments, 61.54% vs. 80.77%, χ2=4.68, P=0.030; English literature, 53.85% vs. 75.00%, χ2=5.80, P=0.016; and flipped class lectures, 36.54% vs. 59.62%, χ2=5.55, P=0.019). ③Enhanced competency in ≥2 core areas was reported in 71.68% of students, including theoretical knowledge acquisition (49.71%), self-directed learning (35.84%), clinical thinking (31.21%), research capabilities (25.43%), and communication skills (22.54%). ④Faculty feedback indicated that 83.33% of instructors perceived the flipped classroom as equivalent or superior to traditional teaching, particularly in cultivating clinical thinking (90.48%), self-directed learning (85.71%), theoretical knowledge acquisition (76.19%), and communication skills (76.19%). ⑤Student participation willingness was primarily influenced by pre-class time investment (46.51% reported excessive effort in preparing discussion materials), while faculty engagement depended on implementation effectiveness (42.86%) and curriculum design (35.71%). Conclusions:The flipped classroom method demonstrated promising initial outcomes in pediatric internship, with high satisfaction among both students and faculty. However, further exploration and practice are required in optimizing teaching components, implementing differentiated instructional strategies, and managing time investment.

With the rapid development of generative artificial intelligence(GAI)technologies,their widespread application in academic research and writing is continuously expanding the boundaries of sci-entific inquiry.However,this trend has also raised a series of ethical and regulatory challenges,inclu-ding issues related to authorship,content authenticity,citation accuracy,and accountability.In light of the growing involvement of AI in generating academic content,establishing an open,controllable,and trustworthy ethical governance framework has become a key task for safeguarding research integrity and maintaining trust within the academic community.This expert consensus outlines ethical requirements across key stages of AI-assisted academic writing-including topic selection,data management,citation practices,and authorship attribution.It aims to clarify the boundaries and ethical obligations surrounding AI use in academic writing,ensuring that technological tools enhance efficiency without compromising in-tegrity.The goal is to provide guidance and institutional support for building a responsible and sustainable research ecosystem.

Objective:To understand the coverage and information consistency rate of National Immunization Program (NIP) vaccines among children born during 2020-2021 in Zhejiang Province, Chongqing City, and Shanxi Province (3 provinces) of China .Methods:A simple random sampling method was used to randomly select 3 counties (districts) from each of the 3 provinces, 5 townships from each county (district), and 5 villages from each township. Vaccination information for seven NIP vaccines was collected for children born between 2020 and 2021 in each village. The vaccination coverage, timely coverage, and consistency rates between the survey data and the Immunization Planning Information System data were analyzed.Results:A total of 1 117 children were investigated. The vaccination coverage for each dose of NIP vaccine ranged from 99.10% to 100.00%, with those in Zhejiang Province, Chongqing City, and Shanxi Province ranging from 99.19% to 100.00%, 98.92% to 100.00%, and 99.20% to 100.00%, respectively. The timely coverage of each dose of NIP vaccine ranged from 89.79% to 99.82%, with those in Zhejiang Province, Chongqing City, and Shanxi Province ranging from 94.09% to 99.73%, 89.52% to 99.73%, and 78.55% to 100.00%, respectively. The consistency rate of information on each dose of NIP vaccine ranged from 94.36% to 99.91%, with those in Zhejiang Province, Chongqing City, and Shanxi Province ranging from 97.85% to 99.73%, 98.92% to 100.00%, and 86.06% to 100.00%, respectively.Conclusions:Coverage of NIP vaccines was generally high among children born during 2020-2021 in the 3 provinces of China, but there were regional differences in the timely coverage of some vaccine doses and the vaccination information consistency rate. It is necessary to strengthen the timely vaccination of children's vaccine booster doses and optimize the management of vaccination services.

To promote the effective implementation of China's Essential Public Health Services Program(EPHSP)and ensure the secure and efficient utilization of project funds,China officially initiated the performance evaluation of EPHSP in 2011.This performance evaluation has evolved through three distinct phases:initial exploration(2011-2014),steady advancement(2015-2018),and reform and enhancement(2019—present).The evaluation objectives have progressively expanded from an initial focus on fund security and service coverage to a broader emphasis on enhancing service quality,improving residents'health status and sense of benefit,and facilitating the refinement of policy frameworks and the implementation of primary responsibilities.Performance evaluation has emerged as a critical instrument for strengthening project governance and optimizing resource allocation,gradually establishing a performance-driven incentive mechanism that aligns rewards with both the quantity and quality of work.This approach has effectively contributed to the continuous improvement of service quality.To further advance the high-quality development of EPHSP,future efforts should optimize the performance evaluation system,prioritize the adoption of a full-cycle performance management approach and the integration of health outcome-based indicators.Additionally,it is essential to deepen the application of information technologies to enhance the precision and efficiency of evaluations,and to innovate mechanisms for utilizing evaluation results to reinforce accountability at the local level.These measures will collectively strengthen project performance management capabilities.

This study aimed to comprehensively examine the association of gallstones, cholecystectomy, and cancer risk. Multivariable logistic regressions were performed to estimate the observational associations of gallstones and cholecystectomy with cancer risk, using data from a nationwide cohort involving 239 799 participants. General and gender-specific two-sample Mendelian randomization (MR) analysis was further conducted to assess the causalities of the observed associations. Observationally, a history of gallstones without cholecystectomy was associated with a high risk of stomach cancer (adjusted odds ratio (aOR)=2.54, 95% confidence interval (CI) 1.50-4.28), liver and bile duct cancer (aOR=2.46, 95% CI 1.17-5.16), kidney cancer (aOR=2.04, 95% CI 1.05-3.94), and bladder cancer (aOR=2.23, 95% CI 1.01-5.13) in the general population, as well as cervical cancer (aOR=1.69, 95% CI 1.12-2.56) in women. Moreover, cholecystectomy was associated with high odds of stomach cancer (aOR=2.41, 95% CI 1.29-4.49), colorectal cancer (aOR=1.83, 95% CI 1.18-2.85), and cancer of liver and bile duct (aOR=2.58, 95% CI 1.11-6.02). MR analysis only supported the causal effect of gallstones on stomach, liver and bile duct, kidney, and bladder cancer. This study added evidence to the causal effect of gallstones on stomach, liver and bile duct, kidney, and bladder cancer, highlighting the importance of cancer screening in individuals with gallstones.
Humans ; Mendelian Randomization Analysis ; Gallstones/complications* ; Female ; Male ; Cholecystectomy/statistics & numerical data* ; Middle Aged ; Risk Factors ; Aged ; Adult ; Neoplasms/etiology* ; Stomach Neoplasms/epidemiology*