ObjectiveTo establish an ultra-performance liquid fingerprint and multi-components determination method for Naomaili granules. To evaluate the quality of different batches by chemometrics， and the anti-neuroinflammatory effects of water extract and main components of Naomaili granules were tested in vitro. MethodsThe similarity and common peaks of 27 batches of Naomaili granules were evaluated by using Ultra performance liquid chromatography （UPLC） fingerprint detection. Ultra-performance liquid chromatography-tandem mass spectrometry （UPLC-MS/MS） technology was used to determine the content of the index components in Naomaili granules and to evaluate the quality of different batches of Naomaili granules by chemometrics. LPS-induced BV-2 cell inflammation model was used to investigate the anti-neuroinflammatory effects of the water extract and main components of Naomaili granules. ResultsThe similarity of fingerprints of 27 batches of samples was > 0.90. A total of 32 common peaks were calibrated， and 23 of them were identified and assigned. In 27 batches of Naomaili granules， the mass fractions of 14 components that were stachydrine hydrochloride， leonurine hydrochloride， calycosin-7-O-glucoside， calycosin，tanshinoneⅠ， cryptotanshinone， tanshinoneⅡ_A， ginsenoside Rb₁， notoginsenoside R₁， ginsenoside Rg₁， paeoniflorin， albiflorin， lactiflorin， and salvianolic acid B were found to be 2.902-3.498， 0.233-0.343， 0.111-0.301， 0.07-0.152， 0.136-0.228， 0.195-0.390， 0.324-0.482， 1.056-1.435， 0.271-0.397， 1.318-1.649， 3.038-4.059， 2.263-3.455， 0.152-0.232， 2.931-3.991 mg∙g^-1， respectively. Multivariate statistical analysis showed that paeoniflorin， ginsenoside Rg₁， ginsenoside Rb₁ and staphylline hydrochloride were quality difference markers to control the stability of the preparation. The results of bioactive experiment showed that the water extract of Naomaili granules and the eight main components with high content in the prescription had a dose-dependent inhibitory effect on the release of NO in the cell supernatant. Among them， salvianolic acid B and ginsenoside Rb₁ had strong anti-inflammatory activity， with IC₅₀ values of （36.11±0.15） mg∙L^-1 and （27.24±0.54） mg∙L^-1， respectively. ConclusionThe quality evaluation method of Naomaili granules established in this study was accurate and reproducible. Four quality difference markers were screened out， and eight key pharmacodynamic substances of Naomaili granules against neuroinflammation were screened out by in vitro cell experiments.

Malignant tumors represent a major threat to global health. Conventional anti-tumor pharmacotherapy often encounters challenges such as drug resistance, highlighting an urgent need for the development of novel therapeutic strategies. Fatty acid synthase (FASN), the key enzyme catalyzing de novo fatty acid synthesis, is subject to precise regulation at multiple levels, including transcriptional control, various post-translational modifications such as ubiquitination and phosphorylation, as well as modulation by diverse signaling pathways. Recent studies have revealed that FASN is aberrantly overexpressed in various malignant tumors and is closely associated with tumor progression and poor patient prognosis. FASN is a homodimer composed of seven functional domains that catalyzes the NADPH-dependent condensation of acetyl-CoA and malonyl-CoA to generate saturated fatty acids, primarily palmitic acid. Its stability is regulated by multiple ubiquitin ligases and deubiquitinating enzymes. Additionally, FASN is subject to upstream regulation via neural precursor cell-expressed developmentally downregulated 8 (Nedd8) modification and the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) pathway, thereby establishing a metabolic-signaling positive feedback loop. As a core executor of metabolic reprogramming, FASN promotes tumorigenesis through dual mechanisms. First, its fatty acid synthesis product, palmitate, participates in membrane phospholipid synthesis, lipid raft formation, and protein palmitoylation, thereby activating several key oncogenic signaling pathways, including PI3K/AKT/mTOR, wingless-type MMTV integration site family member (Wnt)/β‑catenin, and signal transducer and activator of transcription 3 (STAT3)/matrix metalloproteinase (MMP), leading to tumor development and progression. Second, FASN plays a pivotal role in modulating the anti-tumor functions of immune cells and remodeling the tumor immune microenvironment. Specifically, FASN enhances immune checkpoint inhibition by inducing programmed death-ligand 1 (PD-L1) palmitoylation, suppresses the activation of cytotoxic T lymphocytes and natural killer cells, and promotes the polarization of M2-type macrophages, consequently facilitating tumor immune evasion and malignant progression. Precisely due to its significant overexpression in tumor cells, its critical functional role, and its differential expression compared to normal cells, FASN has emerged as a highly promising target for anti-tumor drug development. Highly selective small-molecule inhibitors, notably represented by TVB-2640, have advanced to clinical trial stages and demonstrated favorable anti-tumor activity. Furthermore, the combination of FASN inhibitors with other chemotherapeutic agents or targeted drugs can overcome the limitations of monotherapy through synergistic effects or by resensitizing tumor cells to conventional drugs, achieving a “1+1>2” therapeutic outcome. With the advancement of modern traditional Chinese medicine (TCM), numerous active ingredients derived from TCM have been confirmed to exert anti-tumor effects by modulating FASN-related pathways. This integrated approach leverages the precision of Western medicine while simultaneously harnessing the holistic regulatory benefits of TCM to alleviate the side effects of radiotherapy and chemotherapy. Despite the promising prospects of FASN-targeted therapies, challenges remain, including tumor cell metabolic plasticity, tumor context-dependent responses, and heterogeneity. This review systematically summarizes the molecular structure, physiological functions, and mechanisms of FASN in tumorigenesis, as well as recent advances in targeted therapies. Future directions—including the precise identification of responsive patient populations using spatial transcriptomics, the development of novel combination regimens, and the active exploration of integrative strategies combining traditional Chinese and Western medicine—will facilitate the clinical translation of FASN-targeted therapies and open new avenues for improving the quality of life and prognosis of cancer patients.

Objective To explore the risk factors of painful diabetic neuropathy (PDN) in patients with type 2 diabetes mellitus (T2DM). Methods A total of 269 patients with type 2 diabetes mellitus (T2DM) who were treated in the Department of Endocrinology at Zhangjiagang Hospital Affiliated to Soochow University from January 2020 to December 2024 were selected. The patients were divided into two groups: T2DM without PDN (n=190) and T2DM with PDN (n=79). The general characteristics and biochemical indicators of the two groups of patients were compared. Multivariate logistic regression was used to analyze the associated factors with PDN in T2DM. The receiver operating characteristic (ROC) curve was used to evaluate fasting C-peptide (FC-P), body mass index (BMI), and disease duration to predict the risk of PDN. Results Compared with the T2DM group without concurrent PDN, the T2DM group with concurrent PDN had a longer disease course, lower BMI, higher HDL-C, lower FC-P, and a higher proportion of diabetic retinopathy. The differences between the two groups were statistically significant (P<0.05). The results of multivariate logistic regression analysis showed that BMI, duration, and FC-P were associated factors of PDN. Conclusion BMI, duration and FC-P are associated factors of painful neuropathy complicated with type 2 diabetes.

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.

The prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD) is higher among individuals with inflammatory bowel disease (IBD) than in the general population. Emerging evidence indicates that the development of MASLD in patients with IBD may occur independently of traditional metabolic risk factors, suggesting a unique pathophysiological mechanism distinct from conventional MASLD pathways. Bile acids (BAs), which act as critical signaling molecules in enterohepatic circulation, play an essential role in maintaining gastrointestinal homeostasis through bidirectional gut–liver axis communication. These molecules are increasingly recognized as pivotal regulators in the progression of both MASLD and IBD. While previous studies have characterized the dynamics of BA in blood or fecal samples under both conditions, a comprehensive understanding of their metabolic profiles and the associated interactions within the gut–liver axis is still lacking. This review synthesizes current evidence from studies employing BA metabolomics to investigate IBD and MASLD. By focusing on BA signatures, we summarize conserved alterations between these two conditions and their potential mechanisms of disease progression. Our review advances understanding of BA-mediated pathways in IBD and MASLD, providing a foundation for assessing their roles in contributing to the pathogenesis of MASLD in patients with IBD.

AIM: To compare the efficacy of highly aspherical lenslets(HAL), defocus incorporated multiple segments(DIMS), and single-vision spectacle lenses(SVL)in slowing myopia progression in children and adolescents.

METHODS: In this retrospective cohort study, myopic children and adolescents aged 6-16 y who were prescribed spectacle lenses at the hospital between January 2023 and June 2023 were included. Participants were divided into the HAL, DIMS, and SVL groups according to the correction modality voluntarily selected by participants and their guardians. The follow-up duration was 12 mo. After 12 mo of wear, changes from baseline in sphere, cylinder, spherical equivalent(SE), axial length(AL), flat corneal curvature(K1), steep corneal curvature(K2), mean corneal curvature, and binocular visual function \〖binocular cross-cylinder(BCC), negative/positive relative accommodation(NRA/PRA)\〗 were compared among the three groups after 12-month spectacle wear. Three groups were further stratified by age into age≤10 y and age>10 y subgroups to explore the effect of age on myopia control efficacy of defocus lenses.

RESULTS: A total of 159 myopic participants(159 eyes; only the right eye was included for each subject)were enrolled. In the HAL group(55 participants), there were 24 males and 31 females, with a mean age of 10.07±2.14 y. The DIMS group consisted of 60 participants(33 males, 27 females), with a mean age of 10.02±1.88 y. The SVL group included 44 participants(21 males, 23 females), with a mean age of 10.59±1.93 y. There were no significant differences in K1, K2, mean corneal curvature, NRA, BCC, and PRA among the three groups after 12-month spectacle wear(all P>0.05). There were statistically significant differences in the changes in sphere, SE, and AL(all P<0.05). The changes in sphere, cylinder, SE, and AL in the HAL and DIMS groups were significantly lower than those in the SVL group(all P<0.0167). The DIMS group showed a significantly smaller cylinder progression than the SVL group(P<0.0167). In participants aged≤10 y, the change in spherical power in the HAL group was significantly smaller than those in the DIMS group and SVL group after 12-month spectacle wear(all P<0.0167). The change in cylindrical power in the DIMS group was smaller than that in the SVL group(all P<0.0167). Among participants aged>10 y, the changes in spherical power, SE, and AL in both the HAL group and DIMS group were significantly smaller than those in the SVL group after 12-month spectacle wear(all P<0.0167).

CONCLUSION: Both HAL and DIMS lenses can effectively slow the progression of refractive error and AL elongation in children and adolescents, without significant adverse effects on binocular visual function. Among children≤10 y, HAL lenses yield superior spherical power control, whereas DIMS lenses provide better astigmatism control than SVL.

Metabolic dysfunction-associated fatty liver disease （MAFLD） has become the most common chronic liver disease worldwide. Its pathological process is centered on the “two-hit” theory， with gut-liver axis dysregulation running through the entire disease course from steatosis to liver cirrhosis and even hepatocellular carcinoma， and it forms an extensive cross-system regulatory mechanism with the metabolic， cardiovascular， renal， and psychological and nervous systems through multi-dimensional pathways such as the “gut-liver-brain axis”， the “liver-kidney axis”， and the “liver-heart axis”. Disease progression is not limited to the liver itself， and it also involves functional imbalance of multiple organ systems throughout the body. This article systematically elaborates on the association mechanism of “liver pathological progression-multi-axis regulation-comorbidity occurrence” in MAFLD， proposes synergistic management strategies integrating early screening based on risk stratification， targeted intervention， and multidisciplinary diagnosis and treatment， and analyzes the limitations of current research and future development directions， in order to provide a theoretical basis and practical guidance for precise diagnosis and treatment and individualized management of MAFLD.

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.

Sleep deprivation (SD) has emerged as a significant modifiable risk factor for Alzheimer’s disease (AD), with mounting evidence demonstrating its multifaceted role in accelerating AD pathogenesis through diverse molecular, cellular, and systemic mechanisms. SD is refined within the broader spectrum of sleep-wake and circadian disruption, emphasizing that both acute total sleep loss and chronic sleep restriction destabilize the homeostatic and circadian processes governing glymphatic clearance of neurotoxic proteins. During normal sleep, concentrations of interstitial Aβ and tau fall as cerebrospinal fluid oscillations flush extracellular waste; SD abolishes this rhythm, causing overnight rises in soluble Aβ and tau species in rodent hippocampus and human CSF. Orexinergic neurons sustain arousal, and become hyperactive under SD, further delaying sleep onset and amplifying Aβ production. At the molecular level, SD disrupts Aβ homeostasis through multiple converging pathways, including enhanced production via beta-site APP cleaving enzyme 1 (BACE1) upregulation, coupled with impaired clearance mechanisms involving the glymphatic system dysfunction and reduced Aβ-degrading enzymes (neprilysin and insulin-degrading enzyme). Cellular and histological analyses revealed that these proteinopathies are significantly exacerbated by SD-induced neuroinflammatory cascades characterized by microglial overactivation, astrocyte reactivity, and sustained elevation of pro-inflammatory cytokines (IL-1β, TNF-α, IL-6) through NF‑κB signaling and NLRP3 inflammasome activation, creating a self-perpetuating cycle of neurotoxicity. The synaptic and neuronal consequences of chronic SD are particularly profound and potentially irreversible, featuring reduced expression of critical synaptic markers (PSD95, synaptophysin), impaired long-term potentiation (LTP), dendritic spine loss, and diminished neurotrophic support, especially brain-derived neurotrophic factor (BDNF) depletion, which collectively contribute to progressive cognitive decline and memory deficits. Mechanistic investigations identify three core pathways through which SD exerts its neurodegenerative effects: circadian rhythm disruption via BMAL1 suppression, orexin system hyperactivity leading to sustained wakefulness and metabolic stress, and oxidative stress accumulation through mitochondrial dysfunction and reactive oxygen species overproduction. The review critically evaluates promising therapeutic interventions including pharmacological approaches (melatonin, dual orexin receptor antagonists), metabolic strategies (ketogenic diets, and Mediterranean diets rich in omega-3 fatty acids), lifestyle modifications (targeted exercise regimens, cognitive behavioral therapy for insomnia), and emerging technologies (non-invasive photobiomodulation, transcranial magnetic stimulation). Current research limitations include insufficient understanding of dose-response relationships between SD duration/intensity and AD pathology progression, lack of long-term longitudinal clinical data in genetically vulnerable populations (particularly APOE ε4 carriers and those with familial AD mutations), the absence of standardized SD protocols across experimental models that accurately mimic human chronic sleep restriction patterns, and limited investigation of sex differences in SD-induced AD risk. The accumulated evidence underscores the importance of addressing sleep disturbances as part of multimodal AD prevention strategies and highlights the urgent need for clinical trials evaluating sleep-focused interventions in at-risk populations. The review proposes future directions focused on translating mechanistic insights into precision medicine approaches, emphasizing the need for biomarkers to identify SD-vulnerable individuals, chronotherapeutic strategies aligned with circadian biology, and multi-omics integration across sleep, proteostasis and immune profiles may delineate precision-medicine strategies for at-risk populations. By systematically examining these critical connections, this analysis positions sleep quality optimization as a viable strategy for AD prevention and early intervention while providing a comprehensive roadmap for future mechanistic and interventional research in this rapidly evolving field.