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.

OBJECTIVE To compare the risk of pneumonitis and interstitial lung disease （ILD） associated with different antibody-drug conjugates （ADC） in the treatment of breast cancer. METHODS CNKI， VIP， PubMed， Embase and other Chinese and English databases， and ClinicalTrials.gov were searched from the inception to June 15， 2025. Randomized controlled trials （RCT） about pneumonitis and ILD associated with ADC （T-DM1， T-DXd， SG， Dato-DXd， SHR-A1811，ARX788， and T-Duo） in the treatment of breast cancer were included. After literature screening， data extraction， and quality assessment， a network meta-analysis was conducted using Stata 17.0 software， and the surface under the cumulative ranking curve（SUCRA） of all interventions were ranked. RESULTS A total of 19 RCTs involving 10 556 patients were included. The overall incidence of pneumonitis with ARX788 and T-DXd was significantly higher than that with T-DM1， T-DM1 plus TPC（T-DM1combined with pertuzumab or atezolizumab）， TPC（treatment of primary care）， and SG （ P ＜0.05）， for grade 1-2 pneumonitis， ARX788 and T-DXd showed significantly higher incidence than T-DM1， T-DM1 plus TPC， and TPC （ P ＜0.05）. For both indicators， ARX788 and T-Duo were ranked as the top two by SUCRA. For the incidence of grade ≥3 pneumonitis， T-DXd and T-DM1 were significantly higher than SG （ P ＜0.05）， T-Duo and Dato-DXd were ranked as the top two by SUCRA. For overall incidence of ILD， ARX788 was significantly higher than T-DM1， SHR-A1811， TPC， and SG （ P ＜0.05）， for the incidence of grade 1-2 ILD， ARX788 was significantly higher than T-DM1， SHR-A1811， and TPC （ P ＜0.05）， for the incidence of grade ≥3 ILD， ARX788 was significantly higher than TPC （ P ＜0.05）. For three indicators above， ARX788 and T-DXd combined with pertuzumab were ranked as the top two by SUCRA. CONCLUSIONS Compared with other ADCs， ARX788 and T-Duo are associated with a higher risk of pneumonitis and ILD in patients with breast cancer.

With the improvement of China’s scientific research capabilities， the number of innovative drug enterprises and clinical trials has increased， and disputes over damage compensation have also risen accordingly. It has become a consensus to resolve such disputes through institutionalization. Given the similarities between China and India in terms of population and development level， this study analyzed India’s compensation mechanism for clinical trial damage， including the definition of damage， compensation procedures， and calculation methods. Through a comparison of the mechanisms between China and India， this paper put forward suggestions for improving China’s compensation mechanism for clinical trial damage， that is， clarifying the compensation process for clinical trial damage， the scope and calculation methods of damage compensation， and the compensation ratio corresponding to the correlation between damage and clinical trials. It also proposed to use disability weights to measure the participants’ health status before their enrollment in clinical trials， and to calculate the amount of compensation for disability and death based on their health status.

The diagnosis of occupational radiation diseases, with the diagnostic criteria as main technical basis, has always been one of the emphases of China's occupational health work. This study compared the diagnostic standards for occupational radiation diseases between China and the International Labour Organization (ILO) across three aspects—types, classifications/grades, and diagnostic dose requirements—categorizing the diseases into tissue reactions and stochastic effects. ① For tissue reactions, 8 types are included both in China and ILO standards, 13 unique to China's standards, and 4 unique to the ILO standards; significant differences exist in classifaton/grading; China's standards are either consistent with or stricter (i.e., requiring higher diagnostic dose values) than the ILO standards. ② For stochastic effects (radiation-induced cancer), 12 types were included in both standards; China's standards provide more comprehensive coverage for certain malignancies (e.g., bone cancer); while 5 types were unique to the ILO standards have, some (e.g. salivary gland cancer) are primarily observed in non-occupational populations. Overall, China’s standards include more disease types and maintain more stringent diagnostic dose requirements. However, specific discrepancies require further deliberation during future revisions. Finally, based on these results, recommendations were provided for basic research and the evaluation of Chinese diagnostic criteria.

The diagnosis of occupational radiation diseases, with the diagnostic criteria as main technical basis, has always been one of the emphases of China's occupational health work. This study compared the diagnostic standards for occupational radiation diseases between China and the International Labour Organization (ILO) across three aspects—types, classifications/grades, and diagnostic dose requirements—categorizing the diseases into tissue reactions and stochastic effects. ① For tissue reactions, 8 types are included both in China and ILO standards, 13 unique to China's standards, and 4 unique to the ILO standards; significant differences exist in classifaton/grading; China's standards are either consistent with or stricter (i.e., requiring higher diagnostic dose values) than the ILO standards. ② For stochastic effects (radiation-induced cancer), 12 types were included in both standards; China's standards provide more comprehensive coverage for certain malignancies (e.g., bone cancer); while 5 types were unique to the ILO standards have, some (e.g. salivary gland cancer) are primarily observed in non-occupational populations. Overall, China’s standards include more disease types and maintain more stringent diagnostic dose requirements. However, specific discrepancies require further deliberation during future revisions. Finally, based on these results, recommendations were provided for basic research and the evaluation of Chinese diagnostic criteria.

Objective To establish a fluorescent recombinase-aided amplification (RAA) assay for detection of Strongyloides stercoralis nucleic acid and to preliminarily evaluate its performance. Methods Six sets of specific primers targeting S. stercoralis 18S ribosomal RNA (18S rRNA) gene and one fluorescent probe were designed and synthesized. The optimal primer-probe set was determined through systematic screening and optimization to establish the fluorescent RAA assay. The assay was evaluated using S. stercoralis genomic DNA at concentrations of 100, 10, and 1 pg/μL, and 100, 10, and 1 fg/μL, as well as recombinant pUC57 plasmids containing the target gene fragments at 1 × 10⁵, 1 × 10⁴, 1 × 10³, 1 × 10², 1 × 10¹, 1 × 10⁰ copies/reaction, to determine the analytical sensitivity. Genomic DNA from Ascaris lumbricoides, Ancylostoma duodenale, Enterobius vermicularis, Angiostrongylus cantonensis, Trichinella spiralis, Clonorchis sinensis, Schistosoma japonicum, and Taenia saginata was used to assess assay specificity. A total of 25 stool samples from patients suspected of S. stercoralis infection were tested by the modified Baermann funnel technique, PCR, and the established fluorescent RAA assay. The sensitivity, specificity, concordance rate and their 95% confidence intervals (CI) of these three techniques were estimated, and agreement between methods was evaluated using the Kappa coefficient. Results Exo-4 was identified as the optimal primer set screened from the six primer sets, and the best amplification performance was achieved when the final concentrations of the forward and reverse primers were 0.44 μmol/L and a probe concentration was 0.20 μmol/L. The limit of detection of the fluorescent RAA assay was 100 fg/μL for genomic DNA of S. stercoralis and 1 × 10⁰ copies/reaction for recombinant plasmids. Specific fluorescence signals were detected within 5 min, with no cross-reactivity observed with A. lumbricoides, A. duodenale, E. vermicularis, A. cantonensis, T. spiralis, C. sinensis, S. japonicum, or T. saginata. Among the 25 clinical stool samples from patients suspected of S. stercoralis infections, the modified Baermann funnel technique and fluorescent RAA assay detected 19 positives and 6 negatives, whereas PCR detected 18 positives and 7 negatives. The fluorescent RAA assay showed a sensitivity of 100.00% [95% CI: (82.35%, 100.00%)], specificity of 100.00% [95% CI: (54.07%, 100.00%)], concordance rate of 100.00% [95% CI: (86.28%, 100.00%)], and a Kappa coefficient of 1.00 [95% CI: (1.00, 1.00)] (P < 0.001) relative to the modified Baermann funnel technique, and a sensitivity of 100.00% [95% CI: (81.47%, 100.00%)], specificity of 85.71% [95% CI: (42.13%, 99.64%)], concordance rate of 96.00% [95% CI: (79.65%, 99.90%)], and a Kappa coefficient of 0.90 [95% CI: (0.70, 1.00)] (P < 0.001). Positive amplification products emitted green fluorescence under a portable blue-light device, enabling visual interpretation of results. Conclusions The fluorescent RAA assay established in this study is rapid, highly sensitive, and highly specific. It enables detection of S. stercoralis nucleic acid under isothermal conditions and allows visual interpretation of results, providing a novel tool for rapid clinical diagnosis and field screening of S. stercoralis infections.

To investigate the association between sleep-related phenotypes and the risk of malignancy in pulmonary nodules, and to provide complementary evidence from a general population cohort and genetic analyses.

Background: High-mobility group box protein 1 (HMGB1) is a chromatin-binding protein involved in arthritis, ischemia, sepsis, atherosclerosis, neurodegenerative disorders, meningitis, and cancer. HMGB1 exhibits dual roles in cancer, acting as either a tumor suppressor or oncoprotein depending on context. Objective: This research aimed to elucidate HMGB1’s functional significance in cutaneous squamous cell carcinoma (cSCC). Methods: We overexpressed HMGB1 in cSCC cell lines using recombinant adenovirus and examined its effects on cell proliferation, colony formation, and cell migration. Results: Immunohistochemical analysis revealed elevated HMGB1 expression levels in cSCC tissue relative to normal epidermis. To assess the influence of HMGB1, we employed recombinant adenoviruses expressing HMGB1 to transduce SCC cell lines (SCC12 and SCC13). Enhanced HMGB1 expression significantly promoted cellular proliferation and colony formation capacity.Notably, HMGB1 overexpression elevated the levels of proliferation regulators, including P63, SOX2, CDK4 and CDK6. Furthermore, HMGB1 overexpression substantially enhanced tumor invasiveness, accompanied by upregulation of epithelial-mesenchymal transition (EMT) biomarkers. Mechanistically, overexpression of HMGB1 enhanced transforming growth factor-β signaling by increasing phosphorylation of SMAD2/3, the key mediators of EMT. Conclusion These data imply that HMGB1 acts as a tumor-promoting factor in cSCC.

Objective: To develop and validate a clinically actionable prediction model for postoperative pulmonary complications (PPCs) in cardiac surgery patients, focusing on modifiable preoperative risk factors amenable to targeted optimization. Methods: In this prospective observational cohort study, 492 adults undergoing open-chest cardiac surgery between August 15, 2023 and December 31, 2023 were analyzed. Prespecified predictors included gas exchange variables, pulmonary function, inspiratory muscle strength, and physical performance. Univariable and multivariable logistic regression analyses were used to develop the prediction model. Discrimination was assessed by the area under the receiver operating characteristic curve (AUC). Results: A total of 90 patients (14.1%) developed PPCs after surgery. Five independent predictors were identified: elevated arterial PaCO2 (odds ratio [OR] 1.12, 95% confidence interval [CI] 1.00–1.26), oxygen desaturation (SpO2<93%) (OR 12.47, 95% CI 3.51–48.13), reduced gait speed (OR 0.17, 95% CI 0.04–0.71), lower FEV1/FVC ratio (OR 0.96, 95% CI 0.92–1.00), and diminished inspiratory muscle strength (MIP % predicted) (OR 0.96, 95% CI 0.92–0.99). The model demonstrated good discriminative ability with an AUC of 0.86 (95% CI 0.80–0.93) in the training cohort and 0.87 (95% CI 0.74–0.93) in the validation cohort. Conclusion This parsimonious model achieved high predictive accuracy using five modifiable physiological variables. By targeting abnormalities in gas exchange, pulmonary mechanics, muscle strength, and functional reserve, the model offers a practical tool to guide individualized prehabilitation strategies for reducing PPC risk in cardiac surgery patients.