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.

Background: Aspiration pneumonia (AP), which can be caused by gastric content inhalation into the lower airways, causes acute lung injury (ALI) through complex mechanisms, including inflammation, oxidative stress, and apoptosis. Here, we evaluated the efficacy of exosomes derived from human placental mesenchymal stem cells (hpMSCs) in mitigating ALI in a murine model of AP. We also investigated the role of hsa-let-7i-5p, the most abundant miRNA in hpMSC-derived exosomes, in this respect. Methods: Adult male C57BL/6 mouse AP models were administered hpMSC-derived exosomes (APExo group) or phosphate-buffered saline (AP group) intra-tracheally. After 48 h, the mice were euthanized and evaluated. The effects of hsa-let-7i-5p were assessed by specific inhibition or overexpression. Results: Compared with the APExo group, the AP group exhibited significantly greater ALI, as evidenced by histological damage, increased lung injury scores, impaired lung function, increased leukocyte infiltration, and elevated tissue edema (all P < 0.05). The untreated AP group also showed more inflammation, characterized by nuclear factor-κB upregulation, macrophage M1 polarization, and cytokine level elevation (tumor necrosis factor-α, interleukin-1β, and interleukin-6), as well as increased oxidation and activation of the apoptosis pathway (all P < 0.05). Notably, the therapeutic effects of hpMSC-derived exosomes were compromised by specific inhibition of hsa-let-7i-5p. Furthermore, engineered exosomes derived from genetically modified RAW264.7 overexpressing hsa-let-7i-5p demonstrated therapeutic effects against AP similar to those obtained with hpMSC-derived exosomes. Conclusions In a murine AP model, intra-tracheal administration of hpMSC-derived exosomes has ALI-mitigating effects, involving inflammation, oxidation, and apoptosis modulation, with hsa-let-7i-5p playing a pivotal mediating role.

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.

Objectives: . Idiopathic sudden sensorineural hearing loss (ISSNHL) is a significant cause of hearing loss. Intratympanic steroid injection (ITSI) is commonly used as an initial or salvage treatment; however, the lack of a standardized treatment protocol has resulted in variability in clinical practice. In addition, no efficient prediction model currently exists to support personalized management. Therefore, this study aimed to develop tailored management strategies for ISSNHL using a machine-learning model. Methods: . This retrospective multicenter cohort study was conducted between January 2015 and December 2020, with data analysis performed between January 2021 and March 2024. Patients were selected based on the International Classification of Diseases, 10th Revision criteria for ISSNHL, along with relevant medication and procedure codes. Patients with pure-tone audiogram results not meeting ISSNHL criteria, better initial hearing in the affected ear, an identifiable etiology, no post-treatment audiogram, or delayed treatment (>6 weeks) were excluded. We included 770 patients diagnosed with ISSNHL who received ITSI. The primary outcome was the area under the receiver operating characteristic curve for prediction performance. Recovery status was determined using the last pure-tone audiogram. Modeling was conducted on the Quanta for Medical Care AI platform using five machine-learning algorithms and a nested cross-validation framework, in which feature selection and hyperparameter tuning were performed in the inner folds and model performance was evaluated in the outer folds. Results: . A random forest classifier outperformed the other models in predicting hearing outcomes, achieving an area under the receiver operating characteristic curve of 0.788. Time to ITSI was the most influential treatment-related factor, with ITSI administered within 10 days of hearing loss being associated with better outcomes. This model can be used to provide personalized prognostic estimates under different treatment protocols. Conclusion . The machine-learning-based prediction model facilitates personalized treatment strategies and timely treatment adjustments for ISSNHL, thereby optimizing the likelihood of complete recovery.

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.

The combination of Aidi Injection(ADI) and doxorubicin(DOX) is a common strategy in the treatment of cancer, which can achieve synergistic anti-tumor effects while attenuating the cardiotoxicity caused by DOX. This study aims to investigate the mechanism of ADI in attenuating DOX-induced cardiotoxicity by multi-omics. DOX was used to induce cardiotoxicity in mice, and the cardioprotective effects of ADI were evaluated based on biochemical indicators and pathological changes. Based on the results, transcriptomics, proteomics, and metabolomics were employed to analyze the changes of endogenous substances in different physiological states. Furthermore, data from multiple omics were integrated to screen key regulatory pathways by which ADI attenuated DOX-induced cardiotoxicity, and important target proteins were selected for measurement by ELISA kits and immunohistochemical analysis. The results showed that ADI significantly reduced the levels of cardiac troponin T(cTnT) and N-terminal pro-B-type natriuretic peptide(NT-proBNP) and effectively ameliorated myocardial fibrosis and intracellular vacuolization, indicating that ADI showed therapeutic effect on DOX-induced cardiotoxicity. The transcriptomics analysis screened out a total of 400 differentially expressed genes(DEGs), which were mainly enriched in inflammatory response, oxidative stress, and myocardial fibrosis. After proteomics analysis, 70 differentially expressed proteins were selected, which were mainly enriched in the inflammatory response, cardiac function, and energy metabolism. A total of 51 differentially expressed metabolites were screened by the metabolomics analysis, and they were mainly enriched in multiple signaling pathways, including the inflammatory response, lipid metabolism, and energy metabolism. The integrated data of multiple omics showed that linoleic acid metabolism, arachidonic acid metabolism, and glycerophosphate metabolism pathways played an important role in DOX-induced cardiotoxicity, and ADI may exert therapeutic effects by modulating these pathways. Target validation experiments suggested that ADI significantly regulated abnormal protein levels of cyclooxygenase-1(COX-1), cyclooxygenase-2(COX-2), prostaglandin H2(PGH2), and prostaglandin D2(PGD2) in the model group. In conclusion, ADI may attenuate DOX-induced cardiotoxicity by regulating linoleic acid metabolism, arachidonic acid metabolism, and glycerophosphate metabolism, thus alleviating inflammation of the body.
Doxorubicin/toxicity* ; Animals ; Mice ; Cardiotoxicity/genetics* ; Drugs, Chinese Herbal/administration & dosage* ; Male ; Proteomics ; Metabolomics ; Injections ; Humans ; Multiomics

OBJECTIVE: Pseudomonas aeruginosa( P. aeruginosa) is a prevalent pathogenic bacterium involved in meningitis; however, the virulence factors contributing to this disease remain poorly understood. METHODS: The virulence of the P. aeruginosa A584, isolated from meningitis samples, was evaluated by constructing in vitro blood-brain barrier and in vivo systemic infection models. qPCR, whole-genome sequencing, and drug efflux assays of A584 were performed to analyze the virulence factors. RESULTS: Genomic sequencing showed that A584 formed a phylogenetic cluster with the reference strains NY7610, DDRC3, Pa58, and Pa124. Its genome includes abundant virulence factors, such as hemolysin, the Type IV secretion system, and pyoverdine. A584 is a multidrug-resistant strain, and its wide-spectrum resistance is associated with enhanced drug efflux. Moreover, this strain caused significantly more severe damage to the blood-brain barrier than the standard strain, PAO1. qPCR assays further revealed the downregulation of the blood-brain barrier-associated proteins Claudin-5 and Occludin by A584. During systemic infection, A584 exhibited a higher capacity of brain colonization than PAO1 (37.1 × 10 ⁶ CFU/g brain versus 2.5 × 10 ⁶ CFU/g brain), leading to higher levels of the pro-inflammatory factors IL-1β and TNF-α. CONCLUSION This study sheds light on the virulence factors of P. aeruginosa involved in meningitis.
Pseudomonas aeruginosa/genetics* ; Virulence Factors/metabolism* ; Animals ; Virulence ; Mice ; Pseudomonas Infections/microbiology* ; Blood-Brain Barrier/microbiology* ; Humans ; Female

A microchannel-based electrochemiluminescence(ECL)sensor was developed for detection of tetracycline(TC)utilizing luminol/H2O2 as ECL reporting system.The low excitation potential of luminol/H2O2 effectively mitigated the impact of clamping voltage,thereby enhancing the detection performance of the microchannel-based ECL sensor.The microchannel modified with TC aptamer selectively recognized and captured target TC.The positively charged TC reduced the surface charge density within the microchannel,thereby increasing the ionic current in the microchannel,leading to change of ECL signal of system.The experimental conditions such as electrolyte concentration,TC-aptamer concentration,and reaction time between TC and TC-aptamer were optimized.Under optimal conditions,the difference of ECL signal in the absence and presence of TC(?ECL)exhibited a good linear relationship with TC concentration in the range from 1.00 ng/mL to 200 ng/mL,with a detection limit as low as 0.69 ng/mL.The sensor had good selectivity and was successfully used in detection of TC in milk samples.

Objective To identify risk factors for severe hand-foot-mouth disease(HFMD)cases in Fuzhou,providing scientific evidence for disease prevention and control.Methods Severe pediatric HFMD cases in Fuzhou between 2017 and 2024 were collected from the"China Disease Prevention and Control Information System".Non-severe cases were frequency-matched based on the same onset period,same affected region,same parents'educational level,and age(±1 year)to form a non-severe group.Demographic characteristics,clinical symptoms,medical history,and pathogen types of both groups were collected.Logistic regression was used to identify the risk factors for severe cases.Results From 2017 to 2024,Fuzhou reported 503 severe HFMD cases and 1053 matched non-severe cases.Demographically,severe group had a higher proportion of home-based childcare,rural residence,care-seeking delay>2 d,while EV71 vaccination rates,handwashing habits before meals and after defecation,and caregivers with HFMD prevention education were lower than those in non-severe group,with statistically significant difference(P<0.001).Clinically,severe group showed more frequent fever>3 d,altered consciousness/seizure,limb tremors/convulsions,the proportion of infectious disease and the proportion of EV71-positive cases than those in non-severe group(P<0.001).Multivariate logistic regression analysis showed that home-based childcare(OR=3.213,95%CI 1.913-5.398),rural residence(OR=2.121,95%CI 1.513-2.973),lack of EV7 vaccination(OR=3.141,95%CI 1.996-4.945),care-seeking delay>2 d(OR=2.004,95%CI 1.410-2.849),no handwashing habits before meals and after defecation(OR=3.927,95%CI 1.718-5.356),caregiver without HFMD education(OR=2.465,95%CI 1.807-3.362),fever>3 d(OR=2.585,95%CI 1.801-3.709),altered consciousness/seizures(OR=4.059,95%CI 2.731-6.031),limb tremors/convulsions(OR=2.087,95%CI 1.398-3.117),history of infectious disease(OR=3.369,95%CI 1.725-6.335)and EV71 positivity(OR=3.854,95%CI 2.678-5.545)were risk factors for severe HFMD cases.The recovery time in severe group was significantly longer than that in non-severe group[18(5,32)d vs.11(4,23)d,P<0.001].Conclusions Prevention and control efforts should target high-risk groups by strengthening health education,promoting EV71 vaccination and improving the treatment of severe HFMD cases to reduce the incidence of severe HFMD.