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.

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.

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.

ObjectiveAngle-closure glaucoma (ACG) is one of the major eye-blinding diseases. To diagnose ACG, it is crucial to examine the anterior chamber angle. Current diagnostic tools include slit lamp gonioscopy, water gonioscopy, ultrasound biomicroscopy (UBM), and anterior segment optical coherence tomography (AS-OCT). Slit lamp and water gonioscopy allow convenient observation of the anterior chamber angle, but pose risks of invasive operation and eye infections. UBM can accurately measure the structure of the anterior chamber angle. However, it is complex to operate and unsuitable for patients, who have undergone trauma or ocular surgery. Although AS-OCT provides detailed images, it is costly. The aim of this study is to explore a non-invasive, non-destructive optical reflection tomography (ORT) technique. This technique can achieve low-cost three-dimensional imaging and full-field anterior chamber angle measurement of the porcine eye. MethodsThe experiment involved assembling an optical reflection tomography system, which included a complementary metal oxide semiconductor (CMOS) camera, a telecentric system, a stepper motor, and a white light source, achieving a spatial resolution of approximately 8.5 μm. The process required positioning the porcine eye at the center of the field of the imaging system and rotating it around its central axis using a stepper motor. Reflection projection images were captured at each angle with an exposure time of 1.0 ms and an interval of 2°. The collected reflection-projection data were processed using a filtered reflection tomography algorithm, generating a series of two-dimensional slice data. These slices essentially represented cross-sectional views of the three-dimensional structural image, and were reconstructed into a complete three-dimensional structural image. Based on the reconstructed three-dimensional structural image of the porcine eye, the anterior chamber angles at different positions were measured, and a distribution map of these angles was drawn. Simultaneously, the ORT measurements were compared with the standard results obtained from optical coherence tomography (OCT) to assess the accuracy of ORT measurements. ResultsIn this study, we successfully obtained the reflection projection data of a porcine eye using ORT technology, reconstructed its three-dimensional structural image, and measured the anterior chamber angle, generating the corresponding distribution map. To better distinguish the different structural parts of porcine eye, the three-dimensional structural image was marked with blue, green, and yellow dashed lines from the outer to the inner layers. The area between the blue and green dashed lines corresponded to the sclera. The area between the green and yellow dashed lines corresponded to the iris. The area inside the yellow dashed line corresponded to the pupil. The three-dimensional structural image clearly revealed the key anatomical features of the porcine eye. It was able to measure the anterior chamber angle at different positions. Additionally, the anterior chamber angle measurements of the porcine eye using ORT were compared with the measurements obtained using a TEL320C1 type OCT system, showing an average deviation of 0.51° and a mean square error MSE of 0.317. ConclusionORT is a non-invasive, non-destructive, low-cost, and high-resolution imaging technique capable of achieving three-dimensional structural imaging and full-field anterior chamber angle measurement of a porcine eye. This technology offers a new perspective for the diagnosis of angle-closure glaucoma and is significant for the screening, diagnosis, and monitoring of eye diseases, potentially benefiting clinics and small hospitals in remote areas in the future.

Objective: To analyze the spatial and temporal distribution characteristics of seasonal A(H3N2) influenza [influenza A(H3N2)] in China and to provide a reference for scientific prevention and control. Methods: The influenza A(H3N2) surveillance data in 2014-2019 was derived from China Influenza Surveillance Information System. A line chart described the epidemic trend analyzed and plotted. Spatial autocorrelation analysis was conducted using ArcGIS 10.7, and spatiotemporal scanning analysis was conducted using SaTScan 10.1. Results: A total of 2 603 209 influenza-like case sample specimens were detected from March 31, 2014, to March 31, 2019, and the influenza A(H3N2) positive rate was 5.96%(155 259/2 603 209). The positive rate of influenza A(H3N2) was statistically significant in the north and southern provinces in each surveillance year (all P<0.05). The high incidence seasons of influenza A (H3N2) were in winter in northern provinces and summer or winter in southern provinces. Influenza A (H3N2) clustered in 31 provinces in 2014-2015 and 2016-2017. High-high clusters were distributed in eight provinces, including Beijing, Tianjin, Hebei, Shandong, Shanxi, Henan, Shaanxi, and Ningxia Hui Autonomous Region in 2014-2015, and high-high clusters were distributed in five provinces including Shanxi, Shandong, Henan, Anhui, and Shanghai in 2016-2017. Spatiotemporal scanning analysis from 2014 to 2019 showed that Shandong and its surrounding twelve provinces clustered from November 2016 to February 2017 (RR=3.59, LLR=9 875.74, P<0.001). Conclusion: Influenza A (H3N2) has high incidence seasons with northern provinces in winter and southern provinces in summer or winter and obvious spatial and temporal clustering characteristics in China from 2014-2019.
Humans ; Influenza, Human/epidemiology* ; China/epidemiology* ; Influenza A Virus, H3N2 Subtype ; Seasons ; Cluster Analysis

COVID-19 ; Female ; Humans ; Infectious Disease Transmission, Vertical ; Pregnancy ; Pregnancy Complications, Infectious/drug therapy* ; Pregnancy Outcome ; Premature Birth ; Retrospective Studies ; SARS-CoV-2

OBJECTIVE: To study the long-term clinical effect of multicenter multidisciplinary treatment (MDT) in children with renal malignant tumors. METHODS: A retrospective analysis was performed on the medical data of 55 children with renal malignant tumors who were diagnosed and treated with MDT in 3 hospitals in Hunan Province from January 2015 to January 2020, with GD-WT-2010 and CCCG-WT-2016 for treatment regimens. A Kaplan-Meier survival analysis was used to analyze the survival of the children. RESULTS: Of the 55 children, 10 had stage I tumor, 14 had stage Ⅱ tumor, 22 had stage Ⅲ tumor, 7 had stage IV tumor, and 2 had stage V tumor. As for pathological type, 47 had FH type and 8 had UFH type. All children underwent complete tumor resection. Of the 55 children, 14 (25%) received preoperative chemotherapy. All children, except 1 child with renal cell carcinoma, received postoperative chemotherapy. Among the 31 children with indication for radiotherapy, 21 (68%) received postoperative radiotherapy. One child died of postoperative metastasis. The incidence rate of FH-type myelosuppression was 94.4%, and the incidence rate of UFH-type myelosuppression was 100%. The median follow-up time was 21 months and the median survival time was 26 months for all children, with an overall survival rate of 98% and an event-free survival rate of 95%. CONCLUSIONS Multicenter MDT has the advantages of high success rate of operation and good therapeutic effect of chemotherapy in the treatment of children with renal malignant tumors, with myelosuppression as the most common side effects, and radiotherapy is safe and effective with few adverse events. Therefore, MDT has good feasibility, safety, and economy.
Child ; Family ; Humans ; Kidney Neoplasms/therapy* ; Progression-Free Survival ; Retrospective Studies