OBJECTIVE To construct a predictive model for evaluating the efficacy of a triple antiemetic regimen （neurokinin- 1 receptor antagonist+5-hydroxytryptamine 3 receptor antagonist+dexamethasone） for preventing nausea and vomiting induced by highly emetogenic chemotherapy （HEC） based on interpretable deep learning algorithms. METHODS Clinical data of cancer patients who received HEC and were treated with the standard triple antiemetic regimen in the oncology department of Tianjin Medical University General Hospital from January 2018 to December 2022 were collected retrospectively. Demographic， clinical and metabolism-related variables were integrated. After data pre-processing， two deep learning algorithms （deep random forest and dense neural network） and four machine learning algorithms （support vector machine， categorical boosting， random forest and decision tree） were used to build predictive models. Subsequently， model performance evaluation and model interpretability analysis were conducted. RESULTS Among the six candidate models， the deep random forest model demonstrated the best predictive performance on the test set， with an area under the receiver operating characteristic curve of 0.850， an accuracy of 0.911， a precision of 0.805， a recall of 0.783， an F1 score of 0.793， and a Brier score of 0.075. Interpretability analysis revealed that creatinine clearance rate （Ccr） was the key predictive factor， and low Ccr levels， female gender， younger age， highly emetogenic drugs （particularly cisplatin-containing chemotherapy regimens）， and anticipatory nausea and vomiting were positively correlated with the risk of HEC-related nausea and vomiting. CONCLUSIONS The deep random forest model exhibits the best performance in predicting the efficacy of triple antiemetic regimen for preventing HEC-related nausea and vomiting. The key predictors in this model primarily include Ccr，anticipatory nausea and vomiting， gender， age， and highly emetogenic drugs.

OBJECTIVE To study the improvement effect and mechanism of Tripterygium wilfordii multiglucoside （TWM） on nephrotic syndrome in rats. METHODS The nephrotic syndrome model was established by intravenous injection of adriamycin via the tail vein. The modeling rats were randomly divided into the model group （distilled water）， prednisone group （10 mg/kg）， and TWM high- and low-dose groups （10 and 5 mg/kg， respectively）. Additionally， blank group （distilled water） without model induction was established. Each group consisted of 9 rats. Rats in each group were administered the corresponding drugs or distilled water by gavage， once a day， for 6 consecutive weeks. The histopathological morphology of kidney tissues in rats was observed； the levels of 24-hour urinary protein （24 h-UTP） and serum biochemical indicators [albumin （ALB）， blood urea nitrogen （BUN）， serum creatinine （SCr）， cholesterol （CHOL）， and triglyceride （TG）] in rats were determined； the levels of oxidative stress indicators [superoxide dismutase （SOD）， malondialdehyde （MDA）] in kidney tissue of rats were determined； expressions of hypoxia-inducible factor-1α （HIF-1α）/microRNA-155-5p （miR-155-5p）/nuclear factor erythriod 2- related factor 2 （Nrf2） signaling pathway-related mRNA and protein in the renal tissues of rats were detected. RESULTS Compared with the blank group， the rats in the model group exhibited disordered renal tissue structure， with a small amount of glomerular necrosis and edema of the renal tubular epithelial cells. 24 h-UTP， serum levels of SCr， BUN， CHOL and TG， MDA content， mRNA and protein expressions of HIF-1α and Keap1 as well as the expression of miR-155-5p in renal tissues were increased significantly （ P ＜0.05）. Serum level of ALB， SOD level in renal tissue as well as mRNA and protein expressions of Nrf2 were decreased significantly （ P ＜0.05）. Compared with the model group， TWM high-dose and low-dose groups exhibited significant improvements in renal injury， with notable reversals in the levels of the above quantitative indicators （ P ＜0.05）. CONCLUSIONS TWM can alleviate oxidative stress-induced damage and thereby improve nephrotic syndrome in rats by regulating the HIF-1α/miR-155-5p/Nrf2 signaling pathway.

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.

Plant-derived extracellular vesicles (PDEVs) are nanoscale extracellular vesicles secreted by plant cells, characterized by a lipid bilayer structure. These vesicles carry a variety of bioactive molecules, including proteins, nucleic acids, and lipids, and play essential roles in intercellular communication and physiological regulation in plants. Compared to animal-derived extracellular vesicles, PDEVs offer several advantages, such as a broad range of sources, high biocompatibility, low immunogenicity, and low production costs. Furthermore, PDEVs have demonstrated remarkable potential as natural nanocarriers for drug delivery, due to their ability to efficiently traverse biological barriers, such as the blood-brain barrier, making them promising candidates for drug delivery systems. This review systematically elaborates on the complex composition of PDEVs, which consists of lipids, proteins, and nucleic acids, the typical structural characteristics of their lipid bilayers ranging from 30 to 150 nm, and their versatile loading capabilities as drug carriers, efficiently encapsulating various types of therapeutic agents such as hydrophilic small molecules, hydrophobic drugs, nucleic acids, and proteins. We systematically summarize the recent advancements in strategies for enhancing the loading efficiency of PDEVs, which include methods such as co-incubation, ultrasound-assisted loading, electroporation, freeze-thaw cycles, and microfluidic technology. These techniques are evaluated based on their underlying principles, suitable drug types, and their respective advantages. In addition to loading strategies, we focus on the engineered approaches to achieve targeted delivery using PDEVs, such as genetic engineering modifications, chemical ligand conjugation, membrane fusion technology, and polyethylene glycol (PEG) modification. We discuss the mechanisms of these strategies in enhancing targeting efficiency, prolonging in vivo circulation time, and improving therapeutic efficacy. Further, this review highlights the application of PDEVs in various disease models, including tumor, skin inflammation, metabolic disorders, and neurodegenerative diseases, showcasing their therapeutic potential as multifunctional delivery platforms. The ability of PDEVs to encapsulate diverse therapeutic agents and target specific tissues or cells opens up new avenues for the treatment of complex diseases, offering advantages over conventional drug delivery systems. However, despite the promising applications of PDEVs, several challenges remain in their development and clinical translation. These challenges include variability in source materials, standardization of preparation processes, quality control, scalability of production, and the need for clinical validation. To overcome these obstacles, the integration of advanced technologies such as artificial intelligence-assisted design and multi-omics analysis is proposed as a way to facilitate the precise development of PDEVs. These emerging technologies hold the potential to further enhance the precision and effectiveness of plant-based drug delivery systems, ultimately advancing the field of precision medicine. In conclusion, the use of PDEVs as a platform for drug delivery represents a promising area of research with the potential to revolutionize therapeutic strategies. Their ability to encapsulate and deliver a wide variety of bioactive molecules, along with their inherent advantages in biocompatibility and versatility, makes them a valuable tool in the development of more efficient and targeted therapeutic interventions. Continued research and innovation in this field will pave the way for the clinical implementation of PDEVs in the treatment of various diseases, offering new hope for more effective and sustainable therapeutic options.

Purpose: The aim of this study was to investigate the feasibility of an intelligent handheld ultrasound (US) device for assisting non-expert general practitioners (GPs) in detecting carotid plaques (CPs) in community populations. Methods: This prospective parallel controlled trial recruited 111 consecutive community residents. All of them underwent examinations by non-expert GPs and specialist doctors using handheld US devices (setting A, setting B, and setting C). The results of setting C with specialist doctors were considered the gold standard. Carotid intima-media thickness (CIMT) and the features of CPs were measured and recorded. The diagnostic performance of GPs in distinguishing CPs was evaluated using a receiver operating characteristic curve. Inter-observer agreement was compared using the intragroup correlation coefficient (ICC). Questionnaires were completed to evaluate clinical benefits. Results: Among the 111 community residents, 80, 96, and 112 CPs were detected in settings A, B, and C, respectively. Setting B exhibited better diagnostic performance than setting A for detecting CPs (area under the curve, 0.856 vs. 0.749; P<0.01). Setting B had better consistency with setting C than setting A in CIMT measurement and the assessment of CPs (ICC, 0.731 to 0.923). Moreover, measurements in setting B required less time than the other two settings (44.59 seconds vs. 108.87 seconds vs. 126.13 seconds, both P<0.01). Conclusion Using an intelligent handheld US device, GPs can perform CP screening and achieve a diagnostic capability comparable to that of specialist doctors.

Successful cloning through somatic cell nuclear transfer (SCNT) faces significant challenges due to epigenetic obstacles. Recent studies have highlighted the roles of H3K4me3 and H3K27me3 as potential contributors to these obstacles. However, the underlying mechanisms remain largely unclear. In this study, we generated genome-wide maps of H3K4me3 and H3K27me3 in mouse pre-implantation NT embryos. Our analysis revealed that aberrantly over-represented broad H3K4me3 domain and H3K27me3 signal lead to increased bivalent marks at gene promoters in NT embryos compared with naturally fertilized (NF) embryos at the 2-cell stage, which may link to relatively low levels of H3K36me3 in NT 2-cell embryos. Notably, the overexpression of Setd2, a H3K36me3 methyltransferase, successfully restored multiple epigenetic marks, including H3K36me3, H3K4me3, and H3K27me3. In addition, it reinstated the expression levels of ZGA-related genes by reestablishing H3K36me3 at gene body regions, which excluded H3K27me3 from bivalent promoters, ultimately improving cloning efficiency. These findings highlight the excessive bivalent state at gene promoters as a potent barrier and emphasize the removal of these barriers as a promising approach for achieving higher cloning efficiency.
Animals ; Mice ; Histone-Lysine N-Methyltransferase/biosynthesis* ; Histones/genetics* ; Nuclear Transfer Techniques ; Female ; Gene Expression Regulation, Developmental ; Promoter Regions, Genetic ; Epigenesis, Genetic ; Embryo, Mammalian/metabolism*