Metabolic associated fatty liver disease (MAFLD) is fundamentally driven by an imbalance in hepatic fatty-acid flux: the influx of fatty acids exceeds the liver’s capacity for disposal, resulting in excessive hepatic lipid accumulation, predominantly in the form of triglycerides (TGs). The occurrence and progression of MAFLD depend on disordered regulation across multiple metabolic steps, including fatty-acid uptake, de novo lipogenesis (DNL), fatty-acid oxidation (FAO), and very low-density lipoprotein (VLDL) export. Forkhead box protein O1 (FOXO1) is a key transcriptional regulator within the hepatic network coordinating glucose and lipid metabolism. Under metabolic stress and insulin resistance (IR), FOXO1 expression is frequently increased, whereas its inhibitory phosphorylation is reduced. These changes enhance FOXO1 nuclear localization and transcriptional activity, thereby reprogramming the expression of genes related to metabolism in the liver. Because hepatic lipid deposition is the central pathological feature of MAFLD, the functional status of FOXO1 directly influences hepatic lipid homeostasis. Growing evidence suggests that FOXO1 can exert bidirectional, environment-dependent effects on hepatic lipid accumulation; however, the molecular basis for this functional switch remains incompletely understood. This review systematically summarizes the biological functions and regulatory mechanisms of FOXO1 and its roles in hepatic lipid metabolism, with a particular focus on its crosstalk with insulin signaling. FOXO1 expression is shaped by RNA modifications and epigenetic regulation mediated by non-coding RNAs. Its transcriptional output is precisely governed by post-translational modifications—such as phosphorylation and acetylation—as well as by coordinated nucleocytoplasmic shuttling. Notably, these regulatory patterns vary markedly across nutritional states, degrees of insulin resistance, and stages of disease. In the fed state, insulin/IGF-1 signaling activates the PI3K-AKT pathway, promoting the inhibitory phosphorylation of FOXO1 and facilitating additional modifications, including acetylation, methylation, and ubiquitination. Together, these events drive FOXO1 export from the nucleus and dampen its transcriptional activity, suppressing gluconeogenesis and constraining lipogenic programs. Conversely, during fasting or when insulin signaling is weakened, FOXO1 inhibition is relieved. FOXO1 accumulates in the nucleus, binds to DNA, and regulates the transcription of downstream target genes. Mechanistically, FOXO1 can aggravate hepatic lipid accumulation by activating genes involved in TG synthesis while repressing FAO-related pathways, thereby favoring storage over oxidation. However, under specific conditions, FOXO1 may also alleviate the hepatic lipid burden by promoting TG hydrolysis and enhancing VLDL secretion, thereby reducing the net hepatic lipid load. In addition, lipotoxic signals mediated by ceramides and diacylglycerols (Cer/DAG) activate atypical protein kinase C (aPKC), further exacerbating the disruption of the AKT-FOXO1 axis. This vicious cycle ultimately produces a metabolic paradox in which increased hepatic glucose output coexists with persistent, insulin-independent lipogenesis, accelerating MAFLD progression. Importantly, FOXO1 regulation is not uniform: during early metabolic overload, insulin-mediated suppression may remain effective, whereas in advanced insulin resistance, the loss of AKT control permits sustained FOXO1 activity. Such stage-dependent dynamics may help explain why FOXO1 can either promote steatosis or, in certain contexts, support programs that facilitate lipid turnover. Accordingly, interventions should be liver-specific and tuned to the disease stage, aiming to curb maladaptive FOXO1 signaling while preserving its capacity to promote triglyceride hydrolysis and VLDL secretion when advantageous. Overall, this review offers an important perspective on MAFLD pathogenesis, emphasizing FOXO1 as a potential therapeutic target and providing a theoretical basis for developing liver-specific, disease-course-dependent precision interventions.

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.

Flap transplantation is a critical surgical strategy for the reconstruction of tissue defects caused by trauma, tumor resection, and congenital malformations, and its survival rate directly determines surgical efficacy and patient prognosis. Following transplantation, flaps inevitably undergo ischemia-reperfusion (I/R) injury, during which oxidative stress, inflammatory responses, and metabolic disturbances are intricately intertwined, ultimately leading to cellular injury and tissue necrosis. Recent studies have demonstrated that multiple forms of programmed cell death—including apoptosis, pyroptosis, ferroptosis, necroptosis, and PANoptosis—play central roles in flap I/R injury. The extensive crosstalk and molecular interactions among these pathways form a highly complex cell death network. Specifically, apoptosis is mediated by the imbalance of Bcl-2 family proteins and the activation of cysteine-dependent aspartate-specific protease (caspase) cascades; pyroptosis is driven by the NLRP3-caspase-1-GSDMD axis, resulting in membrane pore formation and the release of pro-inflammatory cytokines; ferroptosis is characterized by iron-dependent lipid peroxidation and dysfunction of glutathione peroxidase 4 (GPX4); necroptosis is triggered by the receptor-interacting serine/threonine-protein kinase 1 (RIPK1)-RIPK3-MLKL signaling complex, leading to membrane rupture; and PANoptosis represents an integrated form of inflammatory cell death that coordinates multiple death pathways. Importantly, these forms of programmed cell death are not independent but are interconnected through extensive signaling crosstalk. Key regulatory molecules, including caspase-8, reactive oxygen species (ROS), nuclear factor-κB (NF-κB), and nuclear factor erythroid 2-related factor 2 (Nrf2), collectively modulate the dynamic balance among these pathways. Therefore, the multidimensional interplay and spatiotemporal dynamics of programmed cell death constitute a fundamental pathological basis of flap I/R injury. This review systematically summarizes the latest advances in the mechanisms and interactions of various programmed cell death pathways in flap I/R injury, aiming to elucidate the underlying regulatory network. These insights may provide novel theoretical foundations for optimizing flap protection strategies, improving flap survival, and promoting tissue repair.

Objective To investigate the incidence and adverse prognosis of late onset sepsis(LOS)in neonates in neonatal intensive care unit(NICU).Methods A retrospective study was conducted to collect and analyze the peri-natal condition,underlying diseases,invasive procedures,and adverse prognosis of neonates in NICU of a regional maternal and child healthcare hospital from 2019 to 2023.According to whether LOS occurred during hospitaliza-tion,neonates were divided into LOS group and non-LOS group.The LOS group was divided into 5 subgroups based on whether invasive procedures were performed:LOS plus umbilical vein catheter(UVC)group,LOS plus peripherally inserted central catheter(PICC)group,LOS plus sequential catheter group,LOS plus tracheal intuba-tion group,and LOS plus lumbar puncture group,the relationship between LOS and adverse prognosis was ana-lyzed.Results Among 2 945 neonates in NICU,354(12.02％)developed LOS.Comparison between LOS groups and non-LOS group were as follows:in term of perinatal condition of neonates,there were statistically significant difference in weight,gestational age,and whether they were twins between the two groups(all P＜0.001);in term of underlying diseases,there were statistically significant differences in the number of cases of maternal gestational hypertension,neonatal asphyxia,neonatal congenital heart disease,neonatal ventricular dilation,neonatal pneumo-nia,neonatal hyperthyrotropinemia,and neonatal anemia,as well as five invasive procedures between the two groups(all P＜0.05).Compared with the non-LOS group,the incidences of retinopathy of prematurity(ROP),neonatal necrotizing enterocolitis(NNEC),bronchopulmonary dysplasia(BPD),and neonatal respiratory distress syndrome(NRDS)in LOS group were all higher(all P＜0.001).Regression analysis showed that compared with the non-LOS groups,the risk of ROP increased in the LOS group and its subgroups,with the LOS plus sequential catheter group having a 2.27-fold higher risk of ROP than non-LOS group;the risk of NNEC increased in the LOS group and its subgroups,with the LOS plus UVC group having an 8.29-fold higher risk of NNEC than the non-LOS group.Except for the LOS plus UVC group,the risk of BPD increased in the LOS group and other subgroups,with the LOS plus PICC group and LOS plus sequential catheter group having 4.68-and 4.64-fold higher risk of BPD than the non-LOS group,respectively;the risk of NRDS in the LOS plus PICC group was 6.84-fold higher than the non-LOS group(all P＜0.05).The top three pathogens causing LOS were coagulase negative Staphylococcus,Klebsiella pneumoniae,and Escherichia coli.Conclusion LOS can significantly increase the risks of ROP,NNEC,BPD,and NRDS.LOS plus invasive procedures can further increase the risk of adverse prognosis.

Objective:To study the relationship between the levels of multiple elements in urine and the risk of arsenic poisoning in populations exposed to drinking water arsenic in Inner Mongolia Autonomous Region (Inner Mongolia).Methods:From April 2023 to January 2024, a case-control study method was used to select 128 individuals with a residence time of ≥10 years in drinking water arsenic exposed areas in Inner Mongolia as study subjects. Eighty-one individuals diagnosed with arsenic poisoning were selected as the case group, and 47 healthy individuals were selected as the control group for urine sample collection and questionnaire survey. Inductively coupled plasma mass spectrometry was employed to determine the levels of 10 elements (chromium, manganese, cobalt, nickel, copper, zinc, arsenic, molybdenum, cadmium and lead) in urine. The levels of each element in urine were divided into four groups ( Q1, Q2, Q3, and Q4 groups) based on quartiles. The associations between the levels of various elements in urine and the risk of arsenic poisoning were studied using binary logistic regression model and restricted cubic spline (RCS). Results:The age of the control group and the case group [ M ( Q1, Q3)] were 61 (53, 69) and 61 (56, 67) years old, respectively. There were 19 and 43 males, and 28 and 38 females, respectively. There was no statistically significant differences in age and and gender composition between the two groups ( Z = - 0.39, P = 0.700; χ 2 = 1.91, P = 0.167). The levels of urinary copper and cadmium of the case group were higher than those of the control group, and the differences were statistically significant ( Z = - 2.66, - 2.16, P < 0.05). The results of univariate logistic regression analysis showed that urinary copper was an influencing factor for arsenic poisoning ( P = 0.017). The results of multivariate logistic regression analysis revealed that after adjusting for covariates, urinary copper and arsenic were independent influencing factors of arsenic poisoning ( P < 0.05). Taking Q1 group as a reference, urinary copper in Q3 group [ OR (95% CI) = 8.23 (1.81, 37.39), P = 0.006] increased the risk of arsenic poisoning, while urinary arsenic in Q2, Q3, and Q4 groups [ OR (95% CI) = 0.24 (0.06, 0.92), 0.12 (0.03, 0.53), 0.15 (0.04, 0.63), P < 0.05] decreased the risk of arsenic poisoning. After adjusting for covariates, RCS did not show a dose-response relationship between urinary copper, urinary arsenic, and arsenic poisoning ( P > 0.05). Conclusion:Urinary arsenic and copper are associated with the risk of arsenic poisoning in the drinking water arsenic exposed areas of Inner Mongolia, copper exposure may contribute significantly to arsenic poisoning.

Objective To investigate the methodology,technical considerations,and precautions of transcatheter arterial emboliza-tion(TAE)using medical adhesive as the primary embolic agent for visceral artery aneurysm(VAA).Methods A total of 14 patients with VAA treated with medical adhesive-based TAE were retrospectively analyzed.Among the 14 patients,6 cases were hepatic artery pseudoaneurysms,4 cases were true gastroduodenal aneurysms,3 cases were gastroduodenal pseudoaneurysms,and 1 case was renal artery pseudoaneurysm.Among the 14 patients,medical adhesive alone was used in 12 cases,while the other 2 cases were employed medical adhesive combined with supplemental coil embolization.Follow-up assessments including abdominal computed tomography angiography(CTA)and laboratory tests were conducted at 3,6,and 12 months post-procedure.Results Technical success was achieved in all cases with mean procedure duration of(60.4±8.2)min and average hospital stay of(8.9±2.1)d.No recanalization,aneurysm enlargement,or non-target embolization were observed during follow-up.Conclusion Medical adhesive-based TAE demonstrates high technical success when performed by experienced interventionists,with low complication rate,low recurrence rate,and low pro-cedure-related mortality,representing a cost-effective preferred treatment for VAA.

Aim To study the effect of evodiamine(EVO)regulating forkhead box protein Ml(FOXM1)on the proliferation and apoptosis of colorectal cancer-resistant cells HCT8/5-FU.Methods CCK-8 assay and EdU assay were used to detect the effect of EVO on cell proliferation ability.Clone formation assay was employed to detect the effect of EVO on the clone for-mation ability of cells.Flow cytometric counting was applied to detect apoptosis.Western blot was utilized to detect the expression of cellular Bcl-2,Bax,FOXM1,β-catenin,c-MYC,and CyclinD1;Molecular docking was used to explore the EVO-FOXM1 interac-tion.Nude mouse transplant tumor model was estab-lished to validate the effect of EVO on HCT8/5-FU cells in vivo.Results CCK-8 assay showed that EVO inhibited the proliferation of HCT8/5-FU cells in a time-and concentration-dependent manner.EdU assay found that the newly proliferated cells in the EVO-trea-ted group were significantly reduced.The results of the clone formation assay showed that EVO inhibited the clone-forming ability of HCT8/5-FU cells.Flow cyto-metric counting found that apoptosis rate of the cells in the EVO group significantly increased.Western blot showed that FOXM1 and β-catenin were significantly highly expressed in HCT8/5-FU cells,and EVO down-regulated the expression of FOXM1,β-cateniin,c-MYC,CyclinD1,and Bcl-2,and up-regulated the ex-pression of Bax.Molecular docking revealed strong in-teractions between EVO and FOXM1.The in vivo ex-perimental results demonstrated that EVO exerted a substantial inhibitory effect on the growth of subcutane-ously implanted HCT8/5-FU xenograft tumors and regulated the expression of related proteins.HE stai-ning revealed significant nuclear consolidation and fragmentation of tumor cells in the EVO group.Con-clusions The findings suggest that EVO could sup-press the activation of the Wnt signaling pathway through a mechanism involving the downregulation of FOXM1 protein expression,thus inhibiting the prolifer-ation of HCT8/5-FU cells and induce their apoptosis.

Objective To assess the clinical feasibility of a self-developed augmented reality(AR)system in surgeries for cerebral arteriovenous malformation(AVM),dural arteriovenous fistula(DAVF),moyamoya,and internal carotid artery stenosis.This system integrates preoperative vascular imaging(CTA,MRA,DSA)with intraoperative real-time visualization through high-precision patient-image registration and virtual-real integration technology.Methods A retrospective analysis was conducted on 6 patients(1 cerebral AVM,1 DAVF,3 moyamoya and 1 internal carotid artery stenosis)collected between March 2023 and April 2024.AR with three-dimensional reconstruction was used for preoperative precise localization and intraoperative navigation guidance.Clinical feasibility was evaluated and analyzed using an intraoperative self-controlled method.Results All 6 patients with diverse etiologies successfully underwent preoperative precise localization and intraoperative navigation guidance under AR three-dimensional reconstruction modeling.This technology achieved visualization of intracranial arteriovenous structures and precise lesion locations,providing surgeons with a visual reference for accurate planning of the surgical approach and operative field.Conclusion The application of AR with three-dimensional reconstruction is safe and feasible in neurosurgical procedures for cerebrovascular diseases.It demonstrates satisfactory effectiveness in preoperative localization and intraoperative navigation guidance.

This article summarizes the nursing experience of a patient who underwent percutaneous transluminal septal branch microsphere embolization combined with MitraClip surgery.Nursing key points:before the operation,a multidisciplinary case nursing team was established to jointly optimize the surgical risk plan;checklist-based nursing management was adopted to fully implement preoperative preparations;dynamic psychological assessment was conducted and early psychological intervention was carried out.Intraoperative monitoring and prevention of complications were well implemented.Postoperative care:close monitoring of cardiac function;good volume management to maintain electrolyte balance;implementation of risk management for bleeding or thrombosis;enhanced care for urethral bleeding.The patient was discharged 11 days after the operation and followed up for 3 months.The 6-minute walk distance increased from 240 m before the operation to 560 m,and the patient returned to normal life.