This paper outlines the development of artificial intelligence （AI） and its applications in traditional Chinese medicine （TCM） research， and elucidates the roles and advantages of large language models， knowledge graphs， and natural language processing in advancing syndrome identification， prescription generation， and mechanism exploration. Using spleen-stomach diseases as an example， it demonstrates the empowering effects of AI in classical literature mining， precise clinical syndrome differentiation， efficacy and safety prediction， and intelligent education， highlighting an upgraded research paradigm that evolves from data-driven and knowledge-driven approaches to intelligence-driven models. To address challenges related to privacy protection and regulatory compliance in cross-institutional data collaboration， a "trusted federated evidence-based analysis platform for TCM spleen-stomach diseases" is proposed， integrating blockchain-based smart contracts， federated learning， and secure multi-party computation. The deep integration of AI with privacy-preserving computing is reshaping research and clinical practice in TCM spleen-stomach diseases， providing feasible pathways and a technical framework for building a high-quality， trustworthy TCM big-data ecosystem and achieving precision syndrome differentiation.

Objective To explore the feasibility and reference value of allogeneic lung transplantation and postoperative monitoring in miniature pigs for lung transplantation research. Methods Two miniature pigs (R1 and R2) underwent left lung allogeneic transplantation. Complement-dependent cytotoxicity tests and blood cross-matching were performed before surgery. The main operative times and partial pressure of arterial oxygen (PaO₂) after opening the pulmonary artery were recorded during surgery. Postoperatively, routine blood tests, biochemical blood indicators and inflammatory factors were detected, and pathological examinations of multiple organs were conducted. Results The complement-dependent cytotoxicity test showed that the survival rate of lymphocytes between donors and recipients was 42.5%-47.3%, and no agglutination reaction occurred in the cross-matching. The first warm ischemia times of D1 and D2 were 17 min and 10 min, respectively, and the cold ischemia times were 246 min and 216 min, respectively. Ultimately, R1 and R2 survived for 1.5 h and 104 h, respectively. Postoperatively, in R1, albumin (ALB) and globulin (GLB) decreased, and alanine aminotransferase increased; in R2, ALB, GLB and aspartate aminotransferase all increased. Urea nitrogen and serum creatinine increased in both recipients. Pathological results showed that in R1, the transplanted lung had partial consolidation with inflammatory cell infiltration, and multiple organs were congested and damaged. In R2, the transplanted lung had severe necrosis with fibrosis, and multiple organs had mild to moderate damage. The expression levels of interleukin-1β and interleukin-6 increased in the transplanted lungs. Conclusions The allogeneic lung transplantation model in miniature pigs may systematically evaluate immunological compatibility, intraoperative function and postoperative organ damage. The data obtained may provide technical references for subsequent lung transplantation research.

Objective: To propose an improved method for constructing the internal quality control (IQC) framework for ELISA assays and validate its efficacy by statistically analyzing IQC data from nine blood center laboratories. Methods: 1) IQC data was collected from nine blood centers and analyzed using a domestic HBsAg ELISA detection kit as an example. 2) Differences between IQC values across batches within Blood Center 1 were assessed. 3) Statistical analyses were performed on batch usage, number of batches used, days of use, number of QC points, batch-specific means, and coefficients of variation (CV) across all nine centers. 4) Using the improved construction method for IQC framework, provisional and permanent frames were established for batches within Blood Center 1 and Blood Center 9, followed by outlier determination. Results: 1) Statistically significant differences were observed in IQC data between batches within Blood Center 1 (P<0.01). It is recommended that both the control material/reagents and the control chart framework be replaced simultaneously. 2) There were substantial differences among 9 blood centers regarding the control material/reagent lot numbers used, the number of QC runs per batch, and the QC values for identical lots. Therefore, individual laboratories should establish their own IQC chart frameworks. 3) The improved IQC framework construction method for ELISA assays is as follows: provisional frames are established via frame-shifting, using the pre-experimental mean and cumulative coefficient of variation (CV) from the preceding batch. For batches used >20 days with >20 QC points, permanent frames are constructed by aggregating in-control data accumulated over ≥20 days with ≥20 points to calculate cumulative mean and standard deviation. The provisional and permanent frames constructed by this method identified all 26 extreme outliers across Blood Centers 1 and 9 as out-of-control. Among the 218 general outliers, 10 were classified as normal by the provisional frames, while the remainder were designated as warnings or out-of-control. This method effectively monitors assay stability. Conclusion: Based on the statistical analysis of IQC practices across blood centers of varying scales, combined with the inherent characteristics of ELISA assays and the batch-to-batch instability of reagents/QC materials, it is recommended to reconstruct QC charts upon lot changes. The proposed method—utilizing frame-shifting for provisional frames and establishing permanent frames based on cumulative data—is applicable to blood center laboratories of differing sizes and effectively monitors the stability of the ELISA assay process.

The common marmoset (Callithrix jacchus), a primate species belonging to the genus Callithrix in the family Callitrichidae, has multiple advantages including small body size, short reproductive cycle, early sexual maturity, 1-3 offspring per litter, and ease of experimental manipulation. It also exhibits behavioral characteristics such as vocal communication and strict monogamy. Common marmosets present unique advantages in neuroscience and brain disease research and are widely used in biomedical fields including neuroscience, gene editing, and ethology. Experimental common marmosets (hereinafter referred to as "experimental marmosets") require management of breeding and use to ensure traceable pedigrees and prevention of inbreeding. Existing management systems developed for rodents or Macaca often cannot effectively meet the practical needs of experimental marmoset management in key aspects such as pedigree tracking and inbreeding warning. To address this issue, the Laboratory Animal Center of the Center for Excellence in Brain Science and Intelligence Technology, Chinese Academy of Sciences (hereinafter referred to as "the Center") designed and developed a second-generation information management system for marmosets in 2021 based on its first-generation information management system, and the second-generation system was launched in 2023. This second-generation system adopts a B/S architecture, is modular, and features front-end and back-end separation. It uses mainstream technical frameworks, including Spring Boot, MySQL, and Redis, and integrates 16 functional modules such as basic information, pedigree tree, breeding and experimental selection database, microchip identification management, and ethical supervision. Compared with the first-generation system, the second-generation system optimizes the pedigree tree and basic information modules. It also adds new functions, including purpose-based database partitioning, animal status tracking, abnormal alerts, automatic microchip identification task generation, inbreeding warning, ethical document upload and review, and record management. Together, these form a systematic and refined management and service system covering the entire breeding and experimental use of experimental marmosets. From 2023 to 2025, the practical application at the center shows that it has significantly improved the management efficiency and service level of the experimental marmoset management unit, and can provide a reference for the development of information management systems in other institutions using marmosets.

Objective: To investigate the incidence, characteristics, and influencing factors of resolution discrepancies within the minipool (MP) testing model across Chinese blood station laboratories in 2024. Methods: A nationwide, multicenter, cross-sectional study was conducted, including 334 blood station laboratories that reported nucleic acid reactive data among enzyme immunoassay non-reactive samples. Of these, 296 laboratories adopted the pool resolution model, with a total of 12 536 273 samples tested. Systematic analysis was performed on resolution data, focusing on the MP-NAT reactivity rate, the pool resolution concordance rate, and the resolution discrepancy rate. Subgroup analyses were conducted based on reagent types, viral targets, and Ct values. Potential causes were further explored through laboratory surveys and re-examination of raw amplification curves. Results: In 2024, the national average MP-NAT reactivity rate was 0.15%. The overall pool resolution concordance rate was 57.86%, which showed a gradual decline as Ct values increased across all reagents. The national average resolution discrepancy rate was 0.081‱(102/12 536 273), with 17.91%(53/296) of laboratories reporting at least one discrepancy. Nine reagent types were associated with these events, exhibiting reagent-specific patterns. For Reagent A2, the predominant discrepancy was HBV reactive pools resolving as HIV (36.36%); for Reagent D1, HBV pools frequently resolved as HCV (38.89%); and for Reagent E, the most common pattern was HIV pools resolving as HBV (48.00%). These resolution discrepancies were strongly associated with high Ct values: the median pool Ct for HBV exceeded 38, while those for HCV and HIV both exceeded 40. Investigations across 16 laboratories revealed that most discrepant samples exhibited “tailing” amplification curves, with some cases linked to cross-contamination or reagent batch-specific issues. Conclusion: While the incidence of resolution discrepancies in the MP-NAT model remains low in China, variations exist across different reagents and laboratories. These discrepancies are closely associated with low viral load, reagent performance, and laboratory operational practices.

Objective Magnetoreception, the remarkable ability of diverse animals to sense and utilize the geomagnetic field for orientation and navigation, remains a molecularly unresolved mystery in sensory biology. The putative magnetoreceptor (MagR, previously known as IscA1) is a highly conserved iron-sulfur protein implicated in both magnetoreception and iron metabolism; however, the functional diversity among its cross-species homologs remains poorly understood. Cellular morphology is a key genetically determined trait that can be altered through genetic or environmental modifications—a process known as cell morphology engineering. Constructing engineered cells with specific morphological features and magnetic sensitivity to achieve remote, non-invasive magnetic modulation represents a crucial goal in this field with significant application potential. Therefore, this study aims to systematically investigate the effects of MagR heterologous expression on bacterial morphology and magnetic sensing capabilities, screen for MagR-based magnetically sensitive morphology engineering pathways, and reveal the underlying molecular mechanisms. Methods We systematically screened 28 MagR homologous genes from diverse prokaryotic and animal taxa to evaluate their expression and corresponding phenotypic effects in Escherichia coli (E. coli). To compare the differential magnetic responses among bacteria expressing various recombinant MagR proteins, we utilized high-throughput automated bright-field microscopic imaging and scanning electron microscopy (SEM). Furthermore, comprehensive biochemical and biophysical characterizations of iron and iron-sulfur cluster binding were performed using Ferrozine colorimetric assays, electron paramagnetic resonance (EPR) spectroscopy, ultraviolet-visible (UV-Vis) absorption, and circular dichroism (CD) spectroscopy. Additionally, 100 mT static magnetic field (SMF) exposure experiments were conducted to assess magnetically tunable phenotypes, while the intrinsic magnetic properties of purified MagR proteins were directly measured using a superconducting quantum interference device (SQUID) magnetometer. Results Our results demonstrated that the heterologous expression of MagR homologs induced varying degrees of bacterial filamentation. From this comprehensive screen, two distinct morphological patterns were identified: hydra (Hydra vulgaris) MagR (hyMagR) promoted uniform cell elongation and filamentation, exhibiting robust magnetic sensitivity manifested as significantly enhanced filamentation under the 100 mT SMF. In contrast, pigeon (Columba livia) MagR (clMagR) induced only low-frequency, extreme filamentation (sporadically exceeding 80 μm) with a relatively weaker magnetic morphological response. Mechanistically, our data unambiguously proved that these phenotypic differences are primarily driven by distinct iron redox preferences rather than total cellular iron accumulation. Specifically, hyMagR preferentially binds ferrous iron (Fe²⁺), whereas clMagR favors ferric iron (Fe³⁺) and forms more stable iron-sulfur clusters. Intriguingly, although SQUID magnetometry showed that purified clMagR exhibited approximately five-fold higher mass magnetic susceptibility than hyMagR, its cellular magnetic response was weaker. We hypothesize that the Fe²⁺-preferred intracellular environment associated with hyMagR overexpression primes the cell for enhanced generation of reactive oxygen species (ROS) via the Fenton reaction. Exposure to an SMF synergizes with this primed redox state, triggering the bacterial SOS response and upregulating cell division inhibitors to efficiently induce uniform filamentation. Conclusion Our findings identify the Fe²⁺/Fe³⁺ redox state as a critical determinant of MagR-mediated morphological remodeling and magnetic responsiveness. This discovery suggests a potential strategy for engineering magnetically responsive cellular systems for synthetic biology applications, and provides a plausible framework, which potentially combines intrinsic protein magnetism with redox-state modulation, for further investigating the evolutionary mechanisms of MagR-mediated magnetoreception.

Objective Magnetoreception, the remarkable ability of diverse animals to sense and utilize the geomagnetic field for orientation and navigation, remains a molecularly unresolved mystery in sensory biology. The putative magnetoreceptor (MagR, previously known as IscA1) is a highly conserved iron-sulfur protein implicated in both magnetoreception and iron metabolism; however, the functional diversity among its cross-species homologs remains poorly understood. Cellular morphology is a key genetically determined trait that can be altered through genetic or environmental modifications—a process known as cell morphology engineering. Constructing engineered cells with specific morphological features and magnetic sensitivity to achieve remote, non-invasive magnetic modulation represents a crucial goal in this field with significant application potential. Therefore, this study aims to systematically investigate the effects of MagR heterologous expression on bacterial morphology and magnetic sensing capabilities, screen for MagR-based magnetically sensitive morphology engineering pathways, and reveal the underlying molecular mechanisms. Methods We systematically screened 28 MagR homologous genes from diverse prokaryotic and animal taxa to evaluate their expression and corresponding phenotypic effects in Escherichia coli (E. coli). To compare the differential magnetic responses among bacteria expressing various recombinant MagR proteins, we utilized high-throughput automated bright-field microscopic imaging and scanning electron microscopy (SEM). Furthermore, comprehensive biochemical and biophysical characterizations of iron and iron-sulfur cluster binding were performed using Ferrozine colorimetric assays, electron paramagnetic resonance (EPR) spectroscopy, ultraviolet-visible (UV-Vis) absorption, and circular dichroism (CD) spectroscopy. Additionally, 100 mT static magnetic field (SMF) exposure experiments were conducted to assess magnetically tunable phenotypes, while the intrinsic magnetic properties of purified MagR proteins were directly measured using a superconducting quantum interference device (SQUID) magnetometer. Results Our results demonstrated that the heterologous expression of MagR homologs induced varying degrees of bacterial filamentation. From this comprehensive screen, two distinct morphological patterns were identified: hydra (Hydra vulgaris) MagR (hyMagR) promoted uniform cell elongation and filamentation, exhibiting robust magnetic sensitivity manifested as significantly enhanced filamentation under the 100 mT SMF. In contrast, pigeon (Columba livia) MagR (clMagR) induced only low-frequency, extreme filamentation (sporadically exceeding 80 μm) with a relatively weaker magnetic morphological response. Mechanistically, our data unambiguously proved that these phenotypic differences are primarily driven by distinct iron redox preferences rather than total cellular iron accumulation. Specifically, hyMagR preferentially binds ferrous iron (Fe²⁺), whereas clMagR favors ferric iron (Fe³⁺) and forms more stable iron-sulfur clusters. Intriguingly, although SQUID magnetometry showed that purified clMagR exhibited approximately five-fold higher mass magnetic susceptibility than hyMagR, its cellular magnetic response was weaker. We hypothesize that the Fe²⁺-preferred intracellular environment associated with hyMagR overexpression primes the cell for enhanced generation of reactive oxygen species (ROS) via the Fenton reaction. Exposure to an SMF synergizes with this primed redox state, triggering the bacterial SOS response and upregulating cell division inhibitors to efficiently induce uniform filamentation. Conclusion Our findings identify the Fe²⁺/Fe³⁺ redox state as a critical determinant of MagR-mediated morphological remodeling and magnetic responsiveness. This discovery suggests a potential strategy for engineering magnetically responsive cellular systems for synthetic biology applications, and provides a plausible framework, which potentially combines intrinsic protein magnetism with redox-state modulation, for further investigating the evolutionary mechanisms of MagR-mediated magnetoreception.

Global childhood and adolescent obesity has become a pressing public health challenge, imposing significant burdens on human health. Obesity is an independent risk factor for insulin resistance, which in turn serves as a critical initiating event for multiple chronic metabolic diseases. The Developmental Origins of Health and Disease (DOHaD) theory highlights the existence of a plastic ^”window period^” during early life, spanning pregnancy and lactation. Maternal nutritional status during this window period profoundly influences offspring metabolic health, with intergenerational transmission of gut microbiota acting as a key mediating pathway. This review summarizes current evidence on how maternal nutrition during the ^”window period^” shapes maternal and offspring gut microbial ecosystems and explores the relationship between these alterations and metabolic risks of obesity and other metabolic disorders in offspring. Based on emerging research, it has been found that maternal nutritional intake during this critical window period modulates early colonization of offspring gut microbiota through multiple pathways, thereby programming long-term metabolic trajectories. These findings suggest that targeting gut microbiota as a preventive strategy during the ^“window period^” may offer novel approaches for combating metabolic disorders, while also providing mechanistic insights into potential microbiota-modulating interventions. This perspective could inform future research directions and clinical applications in metabolic disease prevention.

With the emergence of deep learning techniques based on convolutional neural networks, artificial intelligence (AI) has driven transformative developments in the field of medical image analysis. Recently, large language models (LLMs) such as ChatGPT have also started to achieve distinction in this domain. Increasing research shows the undeniable role of AI in reshaping various aspects of medical image analysis, including processes such as image enhancement, segmentation, detection in image preprocessing, and postprocessing related to medical diagnosis and prognosis in clinical settings. However, despite the significant progress in AI research, studies investigating the recent advances in AI technology in the aforementioned aspects, the changes in research hotspot trajectories, and the performance of studies in addressing key clinical challenges in this field are limited. This article provides an overview of recent advances in AI for medical image analysis and discusses the methodological profiles, advantages, disadvantages, and future trends of AI technologies.
Artificial Intelligence ; Humans ; Image Processing, Computer-Assisted/methods* ; Neural Networks, Computer ; Deep Learning ; Diagnostic Imaging/methods*

Objective:To investigate the inhibitory effect of astragaloside Ⅳ(AS-Ⅳ)on cisplatin(CDDP)-induced liver injury in the mice,and to elucidate its possible mechanism.Methods:Forty male C57BL/6 mice with body weights of 18-22 g were randomly divided into control group,model group,AS-Ⅳ group and adenosine 5'-monophosphate-activated protein kinase(AMPK)inhibitor(Compound C)+AS-Ⅳ group.The mice in control group and model group were gavaged with the same volume of normal saline,and the drug was administered continuously for 9 d.The mice in AS-Ⅳ group and Compound C+AS-Ⅳ group were given AS-Ⅳ aqueous solution(150 mg·kg-1·d-1),respectively.On the 6th day of experiment,the mice in Compound C+AS-Ⅳ group were intraperitoneally injected with Compound C(20 mg·kg-1),and on the 7th day,except for control group,the mice in other groups were intraperitoneally injected with 20 mg·kg-1 CDDP to establish the mouse liver injury models,and the mice were sacrificed 48 h later.Serum and liver tissues were collected,and the levels of aspartate aminotransferase(AST)and alanine aminotransferase(ALT)in the serum of the mice,as well as the activities of superoxide dismutase(SOD)and catalase(CAT)and the levels of malondialdehyde(MDA)in the liver tissue of the mice in various groups were detected by kits.The pathomorphology of liver tissue of the mice in various groups were detected by HE staining.The expression levels of glutathione peroxidase 4(GPX4),ferritin heavy chain 1(FTH1)and ferroptosis inhibitory protein 1(FSP1)proteins in liver tissue of the mice in various groups were detected by immunohistochemical staining,and the expression levels of nuclear factor-E2-related factor 2(Nrf2),heme oxygenase-1(HO-1)and AMPK proteins in liver tissue of the mice in various groups were detected by Western blotting method.Results:Compared with control group,the levels of AST and ALT in serum of the mice in model group were increased(P＜0.01),the activities of SOD and CAT in the liver tissue were significantly decreased(P＜0.01),and the MDA level was increased(P＜0.01);compared with model group,the levels of AST and ALT in serum of the mice in AS-Ⅳ group were decreased(P＜0.01),the MDA level in the liver tissue was decreased(P＜0.01),and the activities of SOD and CAT were increased(P＜0.01);compared with AS-Ⅳ group(P＜0.01),the levels of AST and ALT in serum of the mice in Compound C+AS-Ⅳ group were increased(P＜0.01),the level of MDA in liver tissue was increased(P＜0.05),and the activities SOD and CAT were decreased(P＜0.01).The HE staining results showed that compared with control group,the liver damage degree of the mice in model group was enhanced,the hepatocyte arrangement was disordered,and some hepatocyte edema were increased;compared with model group,the liver morphology of the mice in AS-Ⅳ group returned to normal;compared with AS-Ⅳ group,the hepatocyte arrangement of the mice in Compound C+AS-Ⅳ group was disordered and the edges were blurred.The immunohistochemistry results showed that compared with control group,the expression levels of GPX4,FTH1 and FSP1 proteins in liver tissue of the mice in model group were decreased(P＜0.05);compared with model group,the expression levels of GPX4,FTH1 and FSP1 proteins in liver tissue of the mice in AS-Ⅳ group were increased(P＜0.05);compared with AS-Ⅳ group,the expression levels of GPX4,FTH1 and FSP1 proteins in liver tissue of the mice in Compound C+AS-Ⅳ group were decreased(P＜0.05 or P＜0.01).The Western blotting results showed that compared with control group,the expression levels of Nrf2,HO-1 and AMPK proteins in liver tissue of the mice in model group were decreased(P＜0.01);compared with model group,the expression levels of Nrf2,HO-1 and AMPK proteins in liver tissue of the mice in AS-Ⅳgroup were increased(P＜0.01);compared with AS-Ⅳ group,the expression levels of Nrf2,HO-1 and AMPK proteins in liver tissue of the mice in Compound C+AS-Ⅳ group were decreased(P＜0.01).Conclusion:AS-Ⅳ can alleviate the CDDP-induced liver injury,and its mechanism may be related to the regulation of AMPK/Nrf2/HO-1 signal pathway and ferroptosis by AS-Ⅳ.