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.

Background: and Purpose Live cell-based assays (LCBA) are increasingly used for serological antibody detection due to their ability to preserve antigen conformation, offering moderately higher sensitivity than fixed cell-based assays. However, the clinical necessity of prioritizing LCBA for the detection of aquaporin-4 immunoglobulin G (AQP4-IgG) in neuromyelitis optica spectrum disorder (NMOSD) remains unclear, especially when compared to its established role in diagnosing myelin oligodendrocyte glycoprotein antibody-associated disease. Methods: We compared the performance of live cell-based assays using immunofluorescence (LCBA-IF) and fixed cell-based assays using immunofluorescence (FCBA-IF) in detecting AQP4-IgG in 90 cases of NMOSD meeting 2015 International Panel for Neuromyelitis Optica Diagnosis criteria, alongside 40 controls. Additionally, we further investigated the relationship between AQP4-IgG titers as measured by LCBA-IF and FCBA-IF and clinical parameters in NMOSD patients. Results: Results showed 96.9% agreement between LCBA-IF and FCBA-IF (Cohen’s κ=0.935), with a strong Spearman correlation (0.977, p<0.001). Both methods demonstrated 100% specificity, with LCBA-IF showing slightly higher sensitivity compared to FCBA-IF. Within LCBAIF-tested groups, statistically significant differences in annualized relapse rates were observed across all pairwise comparisons (low-titer vs. moderate-titer, low-titer vs. high-titer, and moderate-titer vs. high-titer; all p<0.050). However, this association reached statistical significance in some FCBA-IF-tested groups. Conclusions Overall, there is a strong concordance between LCBA-IF and FCBA-IF in detecting AQP4-IgG, where LCBA-IF shows slightly higher sensitivity. Furthermore, there is a potential link between elevated AQP4-IgG titers and an increased risk of relapse, and this correlation may appear more pronounced when using LCBA-IF.

OBJECTIVE: To analyze the distribution characteristics of glucose-6-phosphate-dehydrogenase (G6PD) mutations and their enzyme activity in newborns patients with G6PD deficiency in Guilin region. METHODS: From July 2022 to July 2024, umbilical cord blood samples from 4 554 newborns in Guilin were analyzed for G6PD mutations using fluorescence PCR melting curve analysis. Enzyme activity was detected in 4 467 cases using the rate assay. RESULTS: Among 4 467 newborns who underwent G6PD activity testing, 162 newborns (3.63%) were identified as G6PD-deficient, including 142 males (6.04%) and 20 females (0.94%), the prevalence of G6PD deficiency was significantly higher in males than in females (P < 0.001). Genetic analysis of 4 554 newborns detected G6PD mutations in 410 cases (9%), including 171 males (7.13%) and 239 females (11.09%), with a significantly higher mutation detection rate in females than in males (P < 0.001). A total of nine single mutations and four compound heterozygous mutations were identified. The most common mutations were c.1388G>A (33.66%), c.1376G>T (23.66%) and c.95A>G (16.34%). Among newborns who underwent both enzyme activity and genetic mutation testing, males with G6PD mutations had significantly lower enzyme activity than that of females with G6PD mutations(P < 0.001). Specifically, among newborns carrying the mutations c.1388G>A, c.1376G>T, c.95A>G, c.1024C>T or c.871G>A, males consistently exhibited lower enzymatic activity than females with the same mutations (P < 0.001). Furthermore, in male G6PD-deficient newborns, the enzyme activity levels in those carrying c.1388G>A, c.1376G>T, c.95A>G, c.1024C>T, or c.871G>A were lower than those in both the control group and the c.519C>T group (P < 0.05). CONCLUSION This study provides a comprehensive profile of G6PD deficiency incidence and mutation spectrum in the Guilin region. By analyzing enzyme activity and genetic mutation results, this study provides insights into potential intervention strategies and personalized management approaches for the prevention and treatment of neonatal G6PD deficiency in the region.
Humans ; Infant, Newborn ; Glucosephosphate Dehydrogenase Deficiency/epidemiology* ; Glucosephosphate Dehydrogenase/genetics* ; Female ; Male ; Mutation ; China/epidemiology*

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 relationship between sleep disorders and coronary heart disease through big data combined with Mendelian randomization analysis.Methods Data from 2005 to 2018 National Health and Nutrition Examination Survey in the United States were utilized.Logistic regression analysis was employed to evaluate the association between sleep disorders and coronary heart disease,while analyzing relevant influencing factors.A two-sample Mendelian randomization approach was implemented using Genome-Wide Association Studies to establish causal relationships.Results Logistic regression analysis demonstrated a significant association between sleep disorders and coronary heart disease(P＜0.001),with the neutrophil-to-lymphocyte ratio serving as a mediating factor in this relationship(P＜0.001).Mendelian randomization analysis revealed a positive correlation between sleep disorders and coronary heart disease(OR=1.030,95％CI:1.01-1.04).Conclusion Sleep disorders can increase the risk of coronary heart disease by activating inflammatory factors.

Cervical spinal cord injury without fracture-dislocation (CSCIWFD) is referred to as a special type of cervical spinal cord injury characterized by traumatic spinal cord dysfunction and no significant bony structural abnormalities on imagines. Duo to the high risk of missed diagnosis during the initial consultation, CSCIWFD may lead to progressive neurological deterioration or even complete paralysis, severely impacting patients′ prognosis. Currently, there are no established consensuses over the diagnosis and treatment of CSCIWFD, such as the lack of evidence-based standards for indications of non-surgical treatment and risk of secondary neurological injury, as well as debates over the optimal timing for surgical intervention and indications for different surgical approaches. To address these issues, the Spine Trauma Group of the Orthopedic Branch of the Chinese Medical Doctor Association organized experts in the relevant fields to formulate Diagnosis and treatment guideline for acute cervical spinal cord injury without fracture- dislocation in adults ( version 2025) . Based on evidence-based medicine and the principles of scientific rigor and clinical applicability, the guidelines proposed 11 recommendations covering terminology, diagnosis, evaluation treatment, and rehabilitation, etc., aiming to standardize the management of CSCIWFD.

Vertebral refracture following percutaneous vertebral augmentation (PVA) is commonly seen in elderly patients with osteoporotic thoracolumbar compression fractures (OTLCF). It can lead to recurrent pain, loss of vertebral height, progression of kyphosis, and even neurological dysfunction, significantly impairing patients′ quality of life. Current diagnosis and treatment face multiple challenges, including high misdiagnosis rate, difficulty in choosing between surgical and non-surgical treatment options, lack of standardized surgical protocols, interference from intralesional bone cement during procedures, inadequate stability of internal fixation in osteoporotic bone, and suboptimal compliance of anti-osteoporotic therapy. Establishing a standardized diagnostic and therapeutic framework is urgently needed. To standardize the management process and improve outcomes for vertebral refractures after PVA in elderly OTLCF patients, Spinal Trauma Group of the Orthopedic Branch of Chinese Medical Doctor Association organized experts in the field to develop Guideline for the diagnosis and treatment of vertebral refracture after percutaneous vertebral augmentation in elderly patients with osteoporotic thoracolumbar compression fractures ( version 2025), based on current literature and clinical experience, and adhering to principles of scientific rigor and clinical applicability. A total of 11 recommendations were proposed, encompassing diagnosis, treatment, and rehabilitation of vertebral refracture after PVA in elderly patients with OTLCF, aiming to provide a foundation for a standardized management.

Acute symptomatic osteoporotic thoracolumbar compression fracture (ASOTLF) can lead to chronic low back pain, kyphosis deformity, pulmonary dysfunction, loss of mobility, and even life-threatening complications. Vertebral augmentation is currently the mainstream treatment method for this condition. In 2019, the Editorial Board of Chinese Journal of Trauma and the Spinal Trauma Group of Orthopedic Surgeons Branch of Chinese Medical Doctor Association collaboratively led the development of Clinical guideline for vertebral augmentation for acute symptomatic osteoporotic thoracolumbar compression fractures. Six years later, with advances in clinical diagnosis and treatment techniques as well as accumulating evidence in related fields, the 2019 guideline requires updating. To this end, the Spinal Trauma Group of Orthopedic Surgeons Branch of Chinese Medical Doctor Association, the Spinal Health Professional Committee of China Human Health Science and Technology Promotion Association, and the Minimally Invasive Orthopedics Professional Committee of Shaanxi Medical Doctor Association have organized experts in the field to develop the Clinical guideline for vertebral augmentation of acute symptomatic osteoporotic thoracolumbar compression fractures ( version 2025) , based on the latest evidence-based medical researches. This guideline incorporates 3 recommendations retained from the 2019 version with updated strength of evidence, along with 12 new recommendations. It provides recommendations from six aspects of diagnosis, pain management, treatment option selection, prevention of postoperative complications, anti-osteoporosis therapy, and postoperative rehabilitation, aiming to provide a reference for standard treatment of vertebral augmentation for ASOTLF in hospitals at all levels.

Cardiac fibrosis is characterized by an elevated amount of extracellular matrix (ECM) within the heart. However, the persistence of cardiac fibrosis ultimately diminishes contractility and precipitates cardiac dysfunction. Circular RNAs (circRNAs) are emerging as important regulators of cardiac fibrosis. Here, we elucidate the functional role of a specific circular RNA CELF1 in cardiac fibrosis and delineate a novel feedback loop mechanism. Functionally, circ-CELF1 was involved in enhancing fibrosis-related markers' expression and promoting the proliferation of cardiac fibroblasts (CFs), thereby exacerbating cardiac fibrosis. Mechanistically, circ-CELF1 reduced the ubiquitination-degradation rate of BRPF3, leading to an elevation of BRPF3 protein levels. Additionally, BRPF3 acted as a modular scaffold for the recruitment of histone acetyltransferase KAT7 to facilitate the induction of H3K14 acetylation within the promoters of the Celf1 gene. Thus, the transcription of Celf1 was dramatically activated, thereby inhibiting the subsequent response of their downstream target gene Smad7 expression to promote cardiac fibrosis. Moreover, Celf1 further promoted Celf1 pre-mRNA transcription and back-splicing, thereby establishing a feedback loop for circ-CELF1 production. Consequently, a novel feedback loop involving CELF1/circ-CELF1/BRPF3/KAT7 was established, suggesting that circ-CELF1 may serve as a potential novel therapeutic target for cardiac fibrosis.