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.

Human carboxylesterase 2A (hCES2A) plays pivotal roles in prodrug activation and hydrolytic metabolism of ester-bearing chemicals. Targeted inhibition of intestinal hCES2A represents a feasible strategy to mitigate irinotecan-triggered gut toxicity (ITGT), but the orally active, selective, and efficacious hCES2A inhibitors are rarely reported. Here, a novel drug-like hCES2A inhibitor was developed via three rounds of structure-based drug design (SBDD) and structural optimization. Initially, donepezil was identified as a moderate hCES2A inhibitor from 2000 US Food and Drug Administration (FDA)-approved drugs. Following two rounds of SBDD and structural optimization, a donepezil derivative (B7) was identified as a strong reversible hCES2A inhibitor. Subsequently, nine B7 carbamates were rationally designed, synthesized and biologically assayed. Among all synthesized carbamates, C3 showed the most potent time-dependent inhibition on hCES2A (IC₅₀ = 0.56 nmol/L), excellent specificity and favorable drug-like properties. C3 could covalently modify the catalytic serine of hCES2A with high selectivity, while this agent also showed favorable safety profiles, high intestinal exposure, and impressive effects for ameliorating ITGT in both human intestinal organoids and tumor-bearing mice. Collectively, this study showcases a rational strategy for developing drug-like and serine-targeting covalent inhibitors against target serine hydrolase(s), while C3 emerges as a promising orally active drug candidate for ameliorating ITGT.

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.

A method was developed for determination of dilauryl thiodipropionate(DLTDP)in fried foods by coupling solid-phase extraction(SPE)pretreatment with reverse-phase liquid chromatography-tandem mass spectrometry(RPLC-MS/MS)detection.Samples were extracted with n-hexane as the solvent,purified using a neutral alumina SPE cartridge,and finally analyzed by RPLC-MS/MS.Quantitative analysis was performed using matrix-matched calibration curves combined with an external standard method under optimal experimental conditions.The results showed that DLTDP exhibited good linearity in the range of 2.0-50.0 μg/L,with a correlation coefficient(R2)≥0.999.The limit of detection(LOD)and the limit of quantification(LOQ)of the method were 0.15 mg/kg and 0.5 mg/kg,respectively.The mean recoveries at three fortification levels(0.5,1.0,and 200 mg/kg)in different samples ranged from 84.8%to 96.8%,with the relative standard deviations(RSDs)all less than 8.0%.The developed method was highly sensitive,accurate and reliable,and easy to operate,making it well suited for the routine quantitative analysis of DLTDP in fried foods.

Objective To understand the current situation of healthcare-associated infection(HAI)in China,pro-vide data support and decision-making basis for formulating scientific and effective strategies for HAI prevention and control.Methods A nationwide cross-sectional survey on HAI was conducted among various types and levels of medical institutions in China according to a unified protocol of bedside surveys and case investigations.Results In 2024,a total of 5 736 medical institutions and 2 751 765 patients were surveyed.Among them,34 889 HAI cases were identified,with a prevalence rate of 1.27％.The number of HAI episodes was 38 032,and case prevalence rate was 1.38％.The prevalence rate of HAI in medical institutions in different regions of China ranged from 0.66％to 2.35％.Among medical institutions of different scales,those with a bed capacity of ≥900 had the high-est incidence of HAI,reaching 1.65％.The most common infection site was the lower respiratory tract(44.66％),followed by the urinary tract(12.94％),surgical site(9.32％),upper respiratory tract(7.02％),and bloodstream infection(5.78％).The top 3 departments with the highest HAI rates were the general intensive care unit(10.02％),department of neurosurgery(5.51％),and department(group)of hematology(5.34％).A total of 23 238 strains of HAI pathogens were detected,with 10 714 strains(46.10％)from lower respiratory tract speci-mens.The top 5 detected strains were Klebsiella pneumoniae(14.76％),Pseudomonas aeruginosa(13.33％),Escherichia coli(12.79％),Acinetobacter baumannii(9.23％),and Staphylococcus aureus(7.88％).231 944 pa-tients underwent class Ⅰ incision surgery were monitored,with 1 647 cases experienced surgical site infection,and the prevalence rate of surgical site infection was 0.71％.The number of patients who should undergo pathogen de-tection(patients receiving therapeutic and therapeutic combined prophylactic antimicrobial agents)was 715 179,while the actual number was 480 492,with a pathogen detection rate of 67.18％.425 225 patients received patho-genic detection before treatment,with a detection rate of 59.46％.Conclusion The overall HAI prevalence in Chi-na is lower,showing disparities among medical institutions of different regions and scales.Therefore,precise imple-mentation of measures is necessary for HAI prevention and control,with a focus on high-risk institutions and high-risk departments,key areas,and critical procedures.All levels of medical institutions should continuously reduce the incidence of HAI by strengthening monitoring,standardizing the use of antimicrobial agents,and reinforcing basic HAI prevention and control measures.

Objective To understand the current situation of healthcare-associated infection(HAI)in China,pro-vide data support and decision-making basis for formulating scientific and effective strategies for HAI prevention and control.Methods A nationwide cross-sectional survey on HAI was conducted among various types and levels of medical institutions in China according to a unified protocol of bedside surveys and case investigations.Results In 2024,a total of 5 736 medical institutions and 2 751 765 patients were surveyed.Among them,34 889 HAI cases were identified,with a prevalence rate of 1.27％.The number of HAI episodes was 38 032,and case prevalence rate was 1.38％.The prevalence rate of HAI in medical institutions in different regions of China ranged from 0.66％to 2.35％.Among medical institutions of different scales,those with a bed capacity of ≥900 had the high-est incidence of HAI,reaching 1.65％.The most common infection site was the lower respiratory tract(44.66％),followed by the urinary tract(12.94％),surgical site(9.32％),upper respiratory tract(7.02％),and bloodstream infection(5.78％).The top 3 departments with the highest HAI rates were the general intensive care unit(10.02％),department of neurosurgery(5.51％),and department(group)of hematology(5.34％).A total of 23 238 strains of HAI pathogens were detected,with 10 714 strains(46.10％)from lower respiratory tract speci-mens.The top 5 detected strains were Klebsiella pneumoniae(14.76％),Pseudomonas aeruginosa(13.33％),Escherichia coli(12.79％),Acinetobacter baumannii(9.23％),and Staphylococcus aureus(7.88％).231 944 pa-tients underwent class Ⅰ incision surgery were monitored,with 1 647 cases experienced surgical site infection,and the prevalence rate of surgical site infection was 0.71％.The number of patients who should undergo pathogen de-tection(patients receiving therapeutic and therapeutic combined prophylactic antimicrobial agents)was 715 179,while the actual number was 480 492,with a pathogen detection rate of 67.18％.425 225 patients received patho-genic detection before treatment,with a detection rate of 59.46％.Conclusion The overall HAI prevalence in Chi-na is lower,showing disparities among medical institutions of different regions and scales.Therefore,precise imple-mentation of measures is necessary for HAI prevention and control,with a focus on high-risk institutions and high-risk departments,key areas,and critical procedures.All levels of medical institutions should continuously reduce the incidence of HAI by strengthening monitoring,standardizing the use of antimicrobial agents,and reinforcing basic HAI prevention and control measures.

Objective:To explore the effect of psychological nursing on the prognosis of male patients with urethral riding inju-ry treated by ureteroscopic urethral catheter implantation(UCI).Methods:This study included 63 male patients with urethral strad-dle injury treated in the General Hospital of Southern Theater Command from February 2020 to March 2023.We divided the patients into a control(n=29)and an experimental group(n=34)according to the odd-or even-numbered days of admission and treated them by ureteroscopic UCI.Meanwhile those of the former group received routine nursing care and the latter underwent psychological nursing intervention in addition.We obtained the Self-Rating Anxiety Scale(SAS)and Self-Rating Depression Scale(SDS)scores of the patients,recorded their postoperative pain scores,catheter-removal time,hospitalization days,postoperative complications and o-verall recovery status,and compared the data collected between the two groups.Results:At 3 days after surgery,both the SAS and SDS scores were significantly lower in the experimental group than in the control(SAS:45.2±2.9 vs 50.4±3.6,P<0.05;SDS:41.9±2.5 vs 48.3±4.0,P<0.05),and so were the pain scores at24 hours(6.2±0.6 vs 6.8±0.9,P<0.05),48 hours(4.9±0.7 vs 6.1±0.8,P<0.05)and 72 hours after surgery(2.5±0.6 vs 3.9±0.9,P<0.05).The hospitalization time was remarkably shorter in the experimental than in the control group([14.1±2.9]vs[16.1±3.4]d,P<0.05),but there was no statistically significant difference in the time of postoperative catheterization between the two groups of patients([19.3±3.7]vs[19.6±4.4]d,P>0.05).A30-day postoperative follow-up found2 cases of difficult urination in the control group but no com-plications in the experimental group.Conclusion:Ureteroscopic UCI is a safe,effective and minimally invasive treatment method for male urethral riding injury,and psychological nursing helps not only shorten the time of catheterization and hospitalization but also a-void postoperative complications.

As artificial intelligence technology rapidly advances, its deployment within the medical sector presents substantial ethical challenges. Consequently, it becomes crucial to create a standardized, transparent, and secure framework for processing medical data. This includes setting the ethical boundaries for medical artificial intelligence and safeguarding both patient rights and data integrity. This consensus governs every facet of medical data handling through artificial intelligence, encompassing data gathering, processing, storage, transmission, utilization, and sharing. Its purpose is to ensure the management of medical data adheres to ethical standards and legal requirements, while safeguarding patient privacy and data security. Concurrently, the principles of compliance with the law, patient privacy respect, patient interest protection, and safety and reliability are underscored. Key issues such as informed consent, data usage, intellectual property protection, conflict of interest, and benefit sharing are examined in depth. The enactment of this expert consensus is intended to foster the profound integration and sustainable advancement of artificial intelligence within the medical domain, while simultaneously ensuring that artificial intelligence adheres strictly to the relevant ethical norms and legal frameworks during the processing of medical data.
Artificial Intelligence/legislation & jurisprudence* ; Humans ; Consensus ; Computer Security/standards* ; Confidentiality/ethics* ; Informed Consent/ethics*