Intelligent prescription recommendation has become an important research direction in traditional Chinese medicine (TCM), offering new opportunities to support clinical decision-making and promote the modernization of TCM practice. With the rapid development of artificial intelligence (AI), a variety of computational approaches have been proposed to learn prescription patterns from clinical data and generate personalized treatment recommendations. However, despite increasing research activity, systematic and comprehensive reviews of AI-driven methods for TCM prescription recommendation remain limited. In this study, we present a comprehensive review of computational approaches for herbal prescription recommendation (HPR) in TCM. Existing methods are systematically categorized into several major paradigms, including traditional machine learning methods, topic model methods, sequential generative methods, deep learning and graph-based methods, and large language model–based frameworks. In addition to summarizing methodological developments, we also review commonly used public datasets and evaluation metrics in this field. Furthermore, representative models with publicly available implementations are experimentally evaluated on multiple benchmark datasets to provide a comparative analysis of their performance on the HPR task. Finally, we discuss the key challenges that hinder the practical deployment of intelligent prescription recommendation systems, including data heterogeneity, limited interpretability, and insufficient integration of TCM domain knowledge. Future research directions are outlined to facilitate the development of more reliable, interpretable, and clinically applicable AI-assisted HPR systems for TCM.

As China accelerates its efforts in deep space exploration and long-duration space missions, including the operationalization of the Tiangong Space Station and the development of manned lunar missions, safeguarding astronauts’ physiological and cognitive functions under extreme space conditions becomes a pressing scientific imperative. Among the multifactorial stressors of spaceflight, microgravity emerges as a particularly potent disruptor of neurobehavioral homeostasis. Dopamine (DA) plays a central role in regulating behavior under space microgravity by influencing reward processing, motivation, executive function and sensorimotor integration. Changes in gravity disrupt dopaminergic signaling at multiple levels, leading to impairments in motor coordination, cognitive flexibility, and emotional stability. Microgravity exposure induces a cascade of neurobiological changes that challenge dopaminergic stability at multiple levels: from the transcriptional regulation of DA synthesis enzymes and the excitability of DA neurons, to receptor distribution dynamics and the efficiency of downstream signaling pathways. These changes involve downregulation of tyrosine hydroxylase in the substantia nigra, reduced phosphorylation of DA receptors, and alterations in vesicular monoamine transporter expression, all of which compromise synaptic DA availability. Experimental findings from space analog studies and simulated microgravity models suggest that gravitational unloading alters striatal and mesocorticolimbic DA circuitry, resulting in diminished motor coordination, impaired vestibular compensation, and decreased cognitive flexibility. These alterations not only compromise astronauts’ operational performance but also elevate the risk of mood disturbances and motivational deficits during prolonged missions. The review systematically synthesizes current findings across multiple domains: molecular neurobiology, behavioral neuroscience, and gravitational physiology. It highlights that maintaining DA homeostasis is pivotal in preserving neuroplasticity, particularly within brain regions critical to adaptation, such as the basal ganglia, prefrontal cortex, and cerebellum. The paper also discusses the dual-edged nature of DA plasticity: while adaptive remodeling of synapses and receptor sensitivity can serve as compensatory mechanisms under stress, chronic dopaminergic imbalance may lead to maladaptive outcomes, such as cognitive rigidity and motor dysregulation. Furthermore, we propose a conceptual framework that integrates homeostatic neuroregulation with the demands of space environmental adaptation. By drawing from interdisciplinary research, the review underscores the potential of multiple intervention strategies including pharmacological treatment, nutritional support, neural stimulation techniques, and most importantly, structured physical exercise. Recent rodent studies demonstrate that treadmill exercise upregulates DA transporter expression in the dorsal striatum, enhances tyrosine hydroxylase activity, and increases DA release during cognitive tasks, indicating both protective and restorative effects on dopaminergic networks. Thus, exercise is highlighted as a key approach because of its sustained effects on DA production, receptor function, and brain plasticity, making it a strong candidate for developing effective measures to support astronauts in maintaining cognitive and emotional stability during space missions. In conclusion, the paper not only underscores the centrality of DA homeostasis in space neuroscience but also reflects the authors’ broader academic viewpoint: understanding the neurochemical substrates of behavior under microgravity is fundamental to both space health and terrestrial neuroscience. By bridging basic neurobiology with applied space medicine, this work contributes to the emerging field of gravitational neurobiology and provides a foundation for future research into individualized performance optimization in extreme environments.

OBJECTIVE: To explore the pain-location interaction between pressure-sensitive points on the body surface and traditional acupoints in patients with chronic non-specific low back pain (CNLBP) under different disease courses, using Euclidean distance and multivariate statistical analysis. METHODS: A pressure-sensitive point detection was performed on 30 CNLBP patients with varying disease courses. A constant pressure was applied using an FDK20 algometer within a designated lumbar area, a total of 50 points were tested, and the tested points were numbered; the visual analogue scale (VAS) pain score was recorded simultaneously. MatlabR2022a9.12. software was used to extract the positions of pressure-sensitive points, and preprocessing and normalization of point location and VAS scores data were conducted. Under constraint conditions (VAS≥8.0 ∩ Euclidean distance to acupoint≤0.5), the proportion of pressure-sensitive points within the Euclidean distance threshold to each acupoint (PVD_acupoint) was calculated, followed by multivariate statistical analysis. RESULTS: ①Constrained analysis of PVD_acupoint showed that PVD_{Qihaishu (BL24)} and PVD_{Dachangshu (BL25)} were positively correlated with disease course (r=0.55, P<0.01). ②Factor analysis and silhouette analysis revealed that PVD_{Shenshu (BL23)} and PVD_{Dachangshu (BL25)} exhibited trends consistent with disease course progression (P>0.05), with different degree (P<0.01). CONCLUSION The PVD_acupoint value based on Euclidean distance can characterize the pressure sensitivity features of traditional acupoints associated with disease. Multivariate statistical analysis of PVD_acupoint confirms that selecting the acupoint combination of Shenshu (BL23) and Dachangshu (BL25) for CNLBP is associated with the distribution of surrounding pressure-sensitive points and the pathological characteristics of the condition.
Humans ; Acupuncture Points ; Low Back Pain/physiopathology* ; Male ; Female ; Middle Aged ; Adult ; Aged ; Acupuncture Therapy ; Young Adult ; Pressure

Recent advances in large models demonstrate significant prospects for transforming the field of medical imaging. These models, including large language models, large visual models, and multimodal large models, offer unprecedented capabilities in processing and interpreting complex medical data across various imaging modalities. By leveraging self-supervised pretraining on vast unlabeled datasets, cross-modal representation learning, and domain-specific medical knowledge adaptation through fine-tuning, large models can achieve higher diagnostic accuracy and more efficient workflows for key clinical tasks. This review summarizes the concepts, methods, and progress of large models in medical imaging, highlighting their potential in precision medicine. The article first outlines the integration of multimodal data under large model technologies, approaches for training large models with medical datasets, and the need for robust evaluation metrics. It then explores how large models can revolutionize applications in critical tasks such as image segmentation, disease diagnosis, personalized treatment strategies, and real-time interactive systems, thus pushing the boundaries of traditional imaging analysis. Despite their potential, the practical implementation of large models in medical imaging faces notable challenges, including the scarcity of high-quality medical data, the need for optimized perception of imaging phenotypes, safety considerations, and seamless integration with existing clinical workflows and equipment. As research progresses, the development of more efficient, interpretable, and generalizable models will be critical to ensuring their reliable deployment across diverse clinical environments. This review aims to provide insights into the current state of the field and provide directions for future research to facilitate the broader adoption of large models in clinical practice.
Humans ; Diagnostic Imaging/methods* ; Precision Medicine/methods* ; Image Processing, Computer-Assisted/methods*

BACKGROUND: Doxorubicin hydrochloride (DOX) is extensively used in the treatment of various tumors. However, its clinical application is limited due to dose-dependent cardiotoxicity. Currently, few effective strategies exist to mitigate or eliminate DOX-induced cardiomyopathy (DIC). Although ferroptosis is implicated in DIC and its inhibition partially alleviates the condition, the direct targets of DOX in the progression of cardiotoxicity remain unclear. This study aimed to discover the direct targets of DOX in ferroptosis-mediated DIC. METHODS: A DOX pulldown assay was performed to identify proteins specifically binding to DOX in murine hearts, followed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) to identify candidate proteins. A cardiac injury mouse model was established by DOX treatment. Based on this, multiple ferroptosis biomarkers were detected by flow cytometry, quantitative real-time polymerase chain reaction, western blotting, immunochemistry, etc. Besides, specific activator and inhibitor of signaling pathways were applied to illuminate molecular mechanisms. RESULTS: Glutathione S-transferase P1 (GSTP1) was identified as a DOX target. GSTP1 activity was inhibited in DOX-treated cardiomyocytes, while its overexpression significantly alleviated DIC. Moreover, GSTP1 overexpression inhibited acyl-CoA synthetase long-chain family member 4 (ACSL4)-dependent ferroptosis. Mechanistically, GSTP1 overexpression suppressed c-Jun N-terminal kinase (JNK) phosphorylation, thereby reducing reactive oxygen species (ROS) production and inhibiting ferroptosis in DIC. CONCLUSIONS This study identifies the DOX/GSTP1/JNK axis as a critical pathway mediating ACSL4-dependent ferroptosis in DIC. GSTP1 is highlighted as a potential key mediator of ferroptosis and a promising therapeutic target for DIC.

This study aims to systematically characterize the targeted effects of Quanduzhong Capsules on cartilage lesions in knee osteoarthritis by integrating spatial transcriptomics data mining and animal experiments validation, thereby elucidating the related molecular mechanisms. A knee osteoarthritis model was established using Sprague-Dawley(SD) rats, via a modified Hulth method. Hematoxylin and eosin(HE) staining was employed to detect knee osteoarthritis-associated pathological changes in knee cartilage. Candidate targets of Quanduzhong Capsules were collected from the HIT 2.0 database, followed by bioinformatics analysis of spatial transcriptomics datasets(GSE254844) from cartilage tissues in clinical knee osteoarthritis patients to identify spatially specific disease genes. Furthermore, a "formula candidate targets-spatially specific genes in cartilage lesions" interaction network was constructed to explore the effects and major mechanisms of Quanduzhong Capsules in distinct cartilage regions. Experimental validation was conducted through immunohistochemistry using animal-derived biospecimens. The results indicated that Quanduzhong Capsules effectively inhibited the degenerative changes in the cartilage of affected joints in rats, which was associated with the regulation of Quanduzhong Capsules on the thioredoxin-interacting protein(TXNIP)-NOD-like receptor family pyrin domain containing 3(NLRP3)-bone morphogenetic protein receptor type 2(BMPR2)-fibronectin 1(FN1)-matrix metallopeptidase 2(MMP2) signal axis in the articular cartilage surface and superficial zones, subsequently inhibiting cartilage matrix degradation leading to oxidative stress and inflammatory diffusion. In summary, this study clarifies the spatially specific targeted effects and protective mechanisms of Quanduzhong Capsules within pathological cartilage regions in knee osteoarthritis, providing theoretical and experimental support for the clinical application of this drug in the targeted therapy on the inflamed cartilage.
Animals ; Osteoarthritis, Knee/metabolism* ; Drugs, Chinese Herbal/administration & dosage* ; Rats, Sprague-Dawley ; Rats ; Male ; Humans ; Capsules ; Female ; Disease Models, Animal

Objective To establish a chromatography-tandem mass spectrometry method for detecting chlorfenapyr and its metabolite tralopyril in blood,and to investigate the toxicokinetics in rats.Methods Chlorfenapyr(8 mg/kg)was administered orally to rats,and blood samples were collected from rats'canthus vein at 5 min,15 min,30 min,1 h,3 h,6 h,12 h,24 h and 48 h after administration.The blood samples were extracted using 100 μL of 5%formic acid solution and 400 μL of acetonitrile.Chlorfena-pyr was qualitatively and quantitatively detected by triple quadrupole gas chromatography-tandem mass spectrometry(GC-MS/MS)and tralopyril was detected by triple quadrupole liquid chromatography-tandem mass spectrometry(LC-MS/MS).The DAS 3.0 software was used to fit the toxicokinetic equa-tions and calculate the toxicokinetic parameters.Results Chlorfenapyr was detectable from 5 min to 24 h with a peak time of 1 h.Tralopyril was detectable from 15 min to 48 h with a peak time of 3 h.The toxicokinetic process of chlorfenapyr in rat blood conformed to a first-order absorption one-compartment open model,with the toxicokinetic equation described as C=e-0.265t-e-0.175t.Tralopyril con-formed to the first-order absorption three-compartment model,and the toxicokinetic equation was C=47 361.069e-2.209t-35 404.962e-1.486t+11 956.363e-0.512t.In the equations,C stands for the concentration of the target substance in the blood,e is the natural constant(≈2.718 28),and t stands for time.Conclu-sion This study optimized the detection method for chlorfenapyr and its metabolite tralopyril in blood.The toxicokinetic equations and parameters of chlorfenapyr and tralopyril can provide a reference for the estimation of oral intake time of chlorfenapyr.

Cerebral blood flow directly affects the metabolism of substances and neural activity in the brain, and is closely associated with the occurrence and development of cerebral small vessel disease (CSVD). Multiple studies have revealed that various imaging biomarkers in patients with CSVD, such as lacunar infarction, enlarged perivascular spaces, cerebral microbleeds, cerebral atrophy, and white matter hyperintensities, are closely associated with cerebral autoregulation (CA) function. Therefore, understanding the regulatory mechanism of CA in patients with CSVD is of great significance for delaying the further development of CSVD, improving cerebral ischemia and cognitive impairment. This article reviews the correlation and mechanism between CA and CSVD.

Objective:To explore the key components and mechanism of Qufeng Ningfei Powder in treating asthma through qualitative analysis of its blood components, combined with network pharmacology and molecular docking techniques.Methods:The blood components of Qufeng Ningfei Powder were qualitatively analyzed using UPLC-QE-Orbitrap-MS technology. R language was employed to analyze chip data from the GEO database, obtaining a list of differentially expressed genes. SwissTargetPrediction was utilized to predict the targets of drug components. Asthma-related targets were searched through databases such as OMIM, GeneCards, and TTD. The intersection of drug and disease targets was identified using Venn online analysis tool, constructing a "drug-component-target-disease" network to screen potential core components. A protein-protein interaction network (PPI) of core targets was constructed using STRING platform and Cytoscape software. GO function enrichment and KEGG pathway analysis were conducted using DAVID database to validate potential mechanism. Molecular docking was performed to verify the interaction between key components and core targets.Results:A total of 64 components were identified, from which 53 active components were screened, corresponding to 609 targets. Further searching disease databases revealed 96 target genes related to asthma, with an intersection of 6 genes between drug and asthma differential genes. Core target gene IL6 and its corresponding core compound were determined through network topology analysis. Molecular docking confirmed the binding of the main active components of Qufeng Ningfei Powder with the core target protein IL6.Conclusion:The blood components of Qufeng Ningfei Powder may regulate IL-17 through IL6, counteract airway remodeling, oxidative stress, and airway hyperresponsiveness, and thus treat asthma.