Objective: To develop QingNangTCM, a specialized large language model (LLM) tailored for expert-level traditional Chinese medicine (TCM) question-answering and clinical reasoning, addressing the scarcity of domain-specific corpora and specialized alignment. Methods: We constructed QnTCM_Dataset, a corpus of 100 000 entries, by integrating data from ShenNong_TCM_Dataset and SymMap v2.0, and synthesizing additional samples via retrieval-augmented generation (RAG) and persona-driven generation. The dataset comprehensively covers diagnostic inquiries, prescriptions, and herbal knowledge. Utilizing P-Tuning v2, we fine-tuned the GLM-4-9B-Chat backbone to develop QingNangTCM. A multi-dimensional evaluation framework, assessing accuracy, coverage, consistency, safety, professionalism, and fluency, was established using metrics such as bilingual evaluation understudy (BLEU), recall-oriented understudy for gisting evaluation (ROUGE), metric for evaluation of translation with explicit ordering (METEOR), and LLM-as-a-Judge with expert review. Qualitative analysis was conducted across four simulated clinical scenarios: symptom analysis, disease treatment, herb inquiry, and failure cases. Baseline models included GLM-4-9B-Chat, DeepSeek-V2, HuatuoGPT-II (7B), and GLM-4-9B-Chat (freeze-tuning). Results: QingNangTCM achieved the highest scores in BLEU-1/2/3/4 (0.425/0.298/0.137/0.064), ROUGE-1/2 (0.368/0.157), and METEOR (0.218), demonstrating a balanced and superior normalized performance profile of 0.900 across the dimensions of accuracy, coverage, and consistency. Although its ROUGE-L score (0.299) was lower than that of HuatuoGPT-II (7B) (0.351), it significantly outperformed domain-specific models in expert-validated win rates for professionalism (86%) and safety (73%). Qualitative analysis confirmed that the model strictly adheres to the “symptom-syndrome-pathogenesis-treatment” reasoning chain, though occasional misclassifications and hallucinations persisted when dealing with rare medicinal materials and uncommon syndromes. Conclusion Combining domain-specific corpus construction with parameter-efficient prompt tuning enhances the reasoning behavior and domain adaptation of LLMs for TCM-related tasks. This work provides a technical framework for the digital organization and intelligent utilization of TCM knowledge, with potential value for supporting diagnostic reasoning and medical education.

Objective: To characterize longitudinal trajectories of testicular development in boys and breast development in girls, so as to provide reference data for understanding patterns of pubertal sexual maturation. Methods: Based on the Shanghai Pudong New Area Cohort Study on Growth, Development and Health in Children and Adolescents, a baseline survey was conducted in 2020 using a mult stage cluster random sampling method. A total of 2 184 children who completed all follow ups during the primary school period from 13 elementary schools in Pudong New Area,Shanghai,with annual follow ups during 2021-2025. Testicular volume and Tanner stage of breast development were assessed by professional physicians using standardized visual inspection and palpation. The age distribution of testicular volume and breast development was fitted by using cumulative link mixed models and Turnbull s nonparametric maximum likelihood estimation method. Results: Median ages for testicular volumes of 2, 3, 4 and 5 mL in boys were 7.07, 9.24, 10.29, and 11.57 years old, respectively. Median ages for Tanner breast stages Ⅱ, Ⅲ, Ⅳ, and Ⅴ in girls were 8.55 , 10.17, 11.18, and 13.78 years old, respectively. Based on overweight and obesity, stratified analysis showed that earlier pubertal onset among overweight/obesity children, and the key milestones for pubertal initiation were testicular volume reaching 4 mL in boys and breast Tanner II in girls for 10.29, 10.83; 8.18, 9.00 years. Conclusion Overweight and obesity are associated with earlier pubertal initiation,but there are certain gender and developmental stage specific patterns.

ObjectiveTo decipher the mechanism by which Zuoguiwan （ZGW） treat hyperthyroidism in rats with kidney-Yin deficiency based on the dopamine receptor D4 （DRD4）/nicotinamide adenine dinucleotide phosphate （NADPH） oxidase 4 （NOX4） signaling pathway. MethodsThe rat model of kidney-Yin deficiency was induced by unilateral intramuscular injection of dexamethasone （0.35 mg·kg^-1）. After successful modeling， the rats were randomized into model， methimazole （positive control， 5 mg·kg^-1）， low-， medium-， and high-dose （1.85， 3.70， 7.40 g·kg^-1， respectively） ZGW， and normal control groups. After 21 days of continuous gavage， the behavioral indexes and body weight changes of rats were evaluated. The pathological changes of the renal tissue were observed by hematoxylin-eosin staining. The serum levels of thyroid hormones ［triiodothyronine （T₃）， thyroxine （T₄）， thyroid-stimulating hormone （TSH）］， renal function indexes ［serum creatine （Scr） and blood urea nitrogen （BUN）］， energy metabolism markers ［cyclic adenosine monophosphate （cAMP） and cyclic guanosine monophosphate （cGMP）］， and oxidative stress-related factors ［superoxide dismutase （SOD）， malondialdehyde （MDA）， and NADPH）］ were measured by enzyme-linked immunosorbent assay （ELISA）. Western blot was employed to analyze the expression of DRD4， NOX4， mitochondrial respiratory chain complex proteins ［NADH：ubiquinone oxidoreductase subunit S4 （NDUFS4） and cytochrome C oxidase subunit 4 （COX4）］， and inflammation-related protein ［tumor necrosis factor-alpha （TNF-α）， interleukin-6 （IL-6）， p38 mitogen-activated protein kinase （MAPK）］ pathway in the renal tissue. ResultsCompared with the normal group， the model group showed mental malaise， body weight decreases （P<0.01）， inflammatory cell infiltration in the renal tissue， a few residual parotid glands in the thyroid， elevations in serum levels of T₃， T₄， Scr， BUN， cAMP， cAMP/cGMP， MDA， and NADPH （P<0.01）， down-regulation in protein levels of TSH， SOD， and DRD4 （P<0.05， P<0.01）， and up-regulation in expression of NOX4， p-p38 MAPK/p38 MAPK， and inflammatory factors （P<0.01）. Compared with the model group， ZGW increased the body weight （P<0.05， P<0.01）， reduced the infiltration of renal interstitial inflammatory cells， restored the thyroid structure and follicle size， lowered the serum levels of T₃， T₄， Scr， BUN， cAMP， cAMP/cGMP， MDA and NADPH （P<0.05， P<0.01）， up-regulated the expression of TSH， SOD and DRD4 （P<0.05， P<0.01）， and down-regulated the expression of NOX4， p-p38 MAPK/p38 MAPK， and inflammatory factors （P<0.05， P<0.01）. Moreover， high-dose ZGW outperformed methimazole （P<0.05）. ConclusionBy activating DRD4， ZGW can inhibit the expression of NOX4 mediated by the p38 MAPK pathway， reduce oxidative stress and inflammatory response， thereby ameliorating the pathological state of hyperthyroidism due to kidney-Yin deficiency. This study provides new molecular mechanism support for the clinical application of ZGW.

Objective To investigate the species of sandflies and the prevalence of Leishmania infections in sandflies from selected areas of northern and northwestern China, so as to provide insights into identification of leishmaniasis vectors and assessment of epidemiological trends of leishmaniasis in China. Methods Sandfly samples were collected from Mentougou District of Beijing Municipality, Xiangning County in Linfen City of Shanxi Province, Ejin Banner in Alxa League of Inner Mongolia Autonomous Region, and Payzawat County of Karamay District of Karamay City, Gaochang District of Turpan City in Xinjiang Uygur Autonomous Region from July 2023 to July 2024. Approximately 100 intact female sandfly samples were randomly selected from each site and the species of sandflies was identified according to morphological characteristics and molecular assays. Female sandflies originating from the same habitat were grouped into pools of 10 individuals. Leishmania infection was detected using polymerase chain reaction (PCR) assay targeting the internal transcribed spacer 1 (ITS-1) gene, and the prevalence of Leishmania infection was calculated in sandflies from different sampling sites using the minimum infection rate (MIR) method. In addition, positive amplicons were sequenced and subjected to phylogenetic analysis. Results A total of 6 155 sandflies were collected from different environments at sampling sites across the six aforementioned regions from July 2023 to July 2024. Phlebotomus chinensis (96.00%) was the dominant sandfly species in Mentougou District, Beijing Municipality, with a small proportion of Ph. sergenti (4.00%), and only Ph. chinensis was found in Xiangning County, Linfen City, Shanxi Province. Ph. wui was the only sandfly species detected in Ejin Banner, Alxa League, Inner Mongolia Autonomous Region, and Payzawat County, Kashgar City, Xinjiang Uygur Autonomous Region, and Ph. caucasicus (97.70%) was the dominant sandfly species in Karamay District, Karamay City, Xinjiang Uygur Autonomous Region, with a small proportion of Ph. wui (2.30%), while Ph. alexandri was the only species in Gaochang District, Turpan City, Xinjiang Uygur Autonomous Region. A total of 40, 60, 34, 18, 18, and 22 pools of sandfly samples were tested from Mentougou District in Beijing Municipality, Xiangning County in Linfen City of Shanxi Province, Ejin Banner in Alxa League of Inner Mongolia Autonomous Region, Payzawat County in Kashgar City, Karamay District in Karamay City, and Gaochang District in Turpan City of Xinjiang Uygur Autonomous Region, respectively. L. infantum was detected in Ph. chinensis samples from Mentougou District in Beijing Municipality, and Xiangning County of Linfen City in Shanxi Province, with MIR of 0.25% to 1.00%, and L. donovani was detected in Ph. wui from Ejin Banner in Alxa League of Inner Mongolia Autonomous Region, and Payzawat County in Kashgar City of Xinjiang Uygur Autonomous Region, with MIR of 0.56% to 0.88%; however, no Leishmania infection was detected in Ph. caucasicus from Karamay District in Karamay City or Ph. alexandri from Gaochang District in Turpan City of Xinjiang Uygur Autonomous Region. Phylogenetic analysis showed that the Leishmania ITS-1 gene sequences obtained from Mentougou District in Beijing Municipality and Xiangning County in Linfen City of Shanxi Province were clustered into the same clade with the reference sequences of L. infantum ITS-1 gene, while the Leishmania ITS-1 gene sequences obtained from Ejin Banner in Alxa League of Inner Mongolia Autonomous Region and Payzawat County in Kashgar City of Xinjiang Uygur Autonomous Region were clustered into the same clade with the reference sequences of L. donovani ITS-1 gene. Conclusions There are variations in sandfly species in selected areas of northern and northwestern China, and variations in the species of Leishmania infecting sandflies. Improved surveillance of sandfly vectors and targeted control strategies with adaptations to geographical features and leishmaniasis vectors are recommended.

ObjectiveTo investigate the effect of lower extremity exoskeleton robots on balance and walking function of patients with post-stroke cerebellar ataxia. MethodsA total of 60 patients with post-stroke cerebellar ataxia in Beijing Bo'ai Hospital from October, 2022 to October, 2024 were selected, and randomly divided into control group (n = 30) and exoskeleton group (n = 30) randomly. Both groups were given conventional exercise training, including trunk control training, rotation axis training and Frenkel training; the exoskeleton group received additional training with lower limb exoskeleton robots, for four weeks. Before and after treatment, the Gait Watch three-dimensional gait analyzer and the Holden Functional Ambulation Classification (HFAC) were used to evaluate the walking spatiotemporal parameters such as walking speed, walking frequency and step length deviation, as well as the walking ability. Berg Balance Scale (BBS) and the International Cooperative Ataxia Rating Scale (ICARS) were used to access the balance and ataxia functions, respectively. ResultsAfter treatment, the walking speed, walking frequency and step length deviation of both groups improved (|t| > 19.676, P < 0.001), the BBS score improved (|t| > 29.032, P < 0.001), and the ICARS scores decreased (t > 33.192, P < 0.001) in both groups, and they were better in the exoskeleton group than in the control group (|t| > 2.284, P < 0.05). There was no significant difference in the improvement rate of HFAC between two groups (P > 0.05). ConclusionLower extremity exoskeleton robots can effectively improve the balance and walking function of patients with post-stroke cerebellar ataxia.

Objective: To explore the value of contrast-enhanced computed tomography (CT) radiomics combined with machine learning algorithms in the preoperative prediction of perineural invasion (PNI) in oral squamous cell carcinoma (OSCC), aiming to provide evidence for assisting clinical treatment decision-making. Methods​: This study was approved by the Ethics Committee of the Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine. A total of 250 OSCC patients confirmed by postoperative pathology were included, comprising 128 PNI-positive and 122 PNI-negative cases. The dataset was randomly divided into training (n=175), validation (n=38), and independent testing (n=37) sets in a ratio of 7:1.5:1.5. Regions of interest were delineated on preoperative images, and radiomic features were extracted. After dimensionality reduction and feature selection using methods like Least Absolute Shrinkage and Selection Operator (LASSO) regression, multiple machine learning models, including support vector machine (SVM), random forest, Light gradient boosting machine (LightGBM), and a Stacking ensemble model, were constructed. Model performance was evaluated using metrics such as the area under the receiver operating characteristic curve (AUC), sensitivity, calibration curves, and decision curve analysis (DCA). Model interpretability was analyzed using Shapley additive explanations (SHAP) and grouped permutation feature importance analysis. Results ​: ​ Among the 250 samples analyzed, the LightGBM model based on radiomics demonstrated the best performance on the independent test set, with an AUC of 0.781, outperforming models like SVM (AUC = 0.730) and Random Forest (AUC = 0.691), as well as clinical models (AUCs ranging 0.549-0.711). The LightGBM model showed good calibration (Brier score 0.198), and DCA indicated high clinical net benefit over a wide threshold probability range. Paired DeLong tests revealed no statistically significant differences in AUC between the ensemble (Stacking) model and the corresponding best-performing radiomics-based model. SHAP analysis and grouped permutation feature importance analysis further indicated that the primary discriminative information for the model came from radiomic texture features. Conclusion ​ The LightGBM model based on contrast-enhanced CT radiomics demonstrated good discriminative ability for preoperative prediction of PNI in OSCC. In the independent test set, it achieved the highest AUC. This model holds promise as a non-invasive auxiliary tool for preoperative risk assessment. Given the limited sample size of the independent test set, these results require further validation in larger cohorts and external datasets.

In recent years， repetitive transcranial magnetic stimulation （rTMS） has garnered significant attention as a therapeutic approach for sleep disorders in the elderly. However， the prevailing rTMS protocols are predominantly developed based on normative neurophysiological data derived from young adults and fail to incorporate individualized parameters tailored to the brain characteristics of the elderly. To address this gap， the consensus development group synthesized the latest evidence from 2010 to 2025 and established a standardized rTMS protocol specifically for elderly patients with sleep disorders. Adhering to the Appraisal of Guidelines for Research and Evaluation II （AGREE II） framework， systematically screened randomized controlled trials （RCTs） and systematic reviews regarding rTMS in the treatment of sleep disorders across various conditions. Meanwhile， the Grading of Recommendations Assessment， Development and Evaluation （GRADE） system was employed to rigorously grade the quality of evidence and the strength of recommendations. This consensus guideline delineates precise rTMS protocols for the management of sleep disorders in the elderly， highlights the adjustment of stimulation intensity according to scalp-cortex distance recommends either MRI‑guided neuronavigation or the Beam F3/F4 heuristic approach for accurate target localization， thereby providing precise rTMS intervention protocol for sleep disorders in the elderly， aiming to enhance clinical efficacy while ensuring treatment safety. ［Funded by National Key Research and Development Program （number， 2023YFC3603200）； General Program of Shenzhen Science and Technology Innovation Commission （number， JCYJ20240813112859008， JCYJ20240813112900002）； Youth Program of Shenzhen Kangning Hospital （number， KN2023A004）； www.guidelines-registry.cn number， PREPARE-2026CN530］

ObjectiveThis study employs a special conformational engineering (CE) technology to construct an α-chymotrypsin-like active center, which includes a catalytic triad, an oxyanion hole, and a substrate-binding site, on silver nanoparticles (AgNPs), thereby creating an AgNP-based artificial hydrolase with high catalytic activity. This study provides a new approach for the design of highly efficient artificial enzymes and enzyme-mimicking. MethodsAgNPs were chosen as the scaffold to build the an α-chymotrypsin-like active center. A special CE procedure enables the designed peptide, Triad5, to adopt an α‑helical conformation on AgNPs, with the key catalytic residues located on one side of the α-helix forming a catalytic active center with a catalytic triad, an oxyanion hole, and a substrate-binding site. The CE procedure consists of three steps, including conformation induction via trifluoroethanol (TFE), conformation stabilization on AgNPs via Ag-S bonds, and TFE removal via lyophilization. Circular dichroism (CD) spectra were used to confirm the formation and stabilization of the α-helix conformation. Mutations of the key residues combined with stopped-flow kinetic experiments were used to demonstrate the indispensability of each key residue and the synergistic effects among the catalytic triad, the oxyanion hole, and the substrate-binding site. ResultsCD spectra show that the designed Triad5 alone is in random coil conformation; when conjugated on AgNPs, Triad5 still remains largely unstructured; but after the CE treatment, Triad5 adopts a typical α-helical conformation on AgNPs as designed, thus produces an AgNP-based artificial hydrolase, Silverzyme. Silverzyme exhibits extremely high hydrolytic activity towards p-nitrophenyl acetate (p-NPA), with an extremely high catalytic turnover number per active site of 3.5 s^-1, which is even higher than that of α-chymotrypsin. As a comparison, the AgNP-Triad5 conjugate without CE treatment shows much lower catalytic activity than Silverzyme, highlighting the important role of the right conformation of the active center for the catalytic activity and the power of the CE treatment. When the key residues of the catalytic triad of Silverzyme were mutated to alanine, the overall catalytic efficiency of this mutant dropped by about 2 orders of magnitude, unambiguously demonstrating the key role of the designed catalytic triad. Similarly, when the residues for the oxyanion hole were deleted, the mutant with the intact catalytic triad also showed significantly decreased catalytic activity, highlighting the indispensable role of the oxyanion hole for the catalytic activity. Unexpectedly, when both the catalytic triad and the oxyanion hole were kept intact, a slight change of the binding site also resulted in significantly decreased catalytic activity, indicating that the designed binding site is at the right position to align the substrate in the right orientation in the active center for catalytic hydrolysis. These results confirm the synergy among the catalytic triad, the oxyanion hole, and the substrate-binding site, indicating successful mimicking of the active center of α-chymotrypsin. Moreover, Silverzyme shows better thermal stability than α-chymotrypsin, and can even hydrolyze the tough non-activated ester diethyl phthalate, a priority pollutant by the United States Environmental Protection Agency (USEPA). ConclusionThis study successfully mimicked the complex catalytic active center of α-chymotrypsin using a conformational engineering strategy, and produced a highly active artificial hydrolase with a well-defined structure and catalytic mechanism. The findings highlight the significant potential of conformational engineering.

Background Attention-deficit/hyperactivity disorder (ADHD) is one of the most common neurodevelopmental disorders in children and adolescents. Current research suggests that environmental factors, such as per- and polyfluoroalkyl substances (PFAS), may be associated with an increased risk of ADHD in offspring. However, most epidemiological evidence originates from non-Chinese populations, with limited research investigating the association between prenatal PFAS and ADHD-like symptoms in Chinese school-aged children. Objective To examine the association between prenatal PFAS exposure and ADHD-like symptoms in 7-year-old children. Methods Based on the Shanghai Birth Cohort, this study included 488 mother-child pairs. PFAS were measured in maternal serum during the second trimester using ultra-performance liquid chromatography-tandem mass spectrometry. The analysis focused on eight specific compounds from three PFAS categories [perfluorooctanoic acid (PFOA), perfluorooctanesulfonic acid (PFOS), and perfluorohexanesulfonic acid (PFHxS)]. The hyperactivity subscale of the Strengths and Difficulties Questionnaire (SDQ) was used to assess ADHD-like symptoms in children at age 7. Basic demographic information was collected via questionnaires and medical records. Negative binomial regression models were used to evaluate the relationship between individual exposures to the three typical PFAS categories (PFOA, PFOS, and PFHxS) and children's ADHD-like symptoms, while the bayesian kernel machine regression (BKMR) mixture model was employed to assess the overall effect of PFAS mixture exposure on ADHD-like symptoms. Results The detection rates of the eight target PFAS in maternal second-trimester serum were all above 85%, with PFOA exhibiting the highest median concentration (10.21 ng·mL⁻¹). Overall, 14.75% of the children exhibited ADHD-like symptoms. Results from the negative binomial regression models showed that second-trimester exposures to PFOA (IRR=1.15, 95%CI: 1.05, 1.26), n-PFOS (IRR=1.08, 95%CI: 1.01, 1.15), 6m-PFOS (IRR=1.13, 95%CI: 1.05, 1.21), and 1m-PFOS (IRR=1.11, 95%CI: 1.03, 1.18) were associated with an increased risk of ADHD-like symptoms in 7-year-old children. Stratified analyses suggested potentially statistically significant associations in girls [e.g., for PFOA, boys vs. girls: 1.09 (0.96, 1.23) vs. 1.25 (1.09, 1.44), P-int=0.31]. The BKMR mixture model indicated that the risk of ADHD-like symptoms trended upward with increasing PFAS mixture concentrations, although this association was statistically significant only among girls (P < 0.05). Conclusion Prenatal PFAS exposure may be associated with an increased risk of ADHD-like symptoms in 7-year-old children, and this adverse association appears to be more prominent in girls.

Objective To investigate the protective effect and underlying mechanisms of HydroMg, a novel sustained-release hydrogen donor, in a model of doxorubicin (DOX)-induced chronic cardiotoxicity. Methods H9C2 cells were treated with 1 μmol/L DOX and/or 1 μg/mL HydroMg, and mitochondrial membrane potential, reactive oxygen species (ROS) levels, and apoptosis were measured. In the animal experiment, mice were divided into control, HydroMg, DOX, and DOX＋HydroMg groups. The HydroMg group received intraperitoneal injection of HydroMg (100 mg/kg). Chronic cardiotoxicity was established by intraperitoneal injection of DOX (5 mg/kg every week) for four consecutive weeks; In the DOX＋HydroMg group, HydroMg (100 mg/kg) was administered intraperitoneally 4–5 h prior to each DOX injection. Cardiac function and remodeling were evaluated by echocardiography and heart weight index (HWI). The long-term survival of mice were analyzed. Subsequently, transcriptome sequencing was performed on myocardial tissues from the DOX group to identify key molecular alterations compared with controls. Results At the cellular level, HydroMg treatment effectively reversed DOX-induced mitochondrial dysfunction and oxidative stress, as evidenced by restored membrane potential, reductions in total ROS and mitochondrial superoxide levels by approximately 30% and 9%, respectively, and a decrease in the apoptosis rate from 34.13% to 18.27%. In the animal model, compared with the DOX group, HydroMg intervention increased left ventricular ejection fraction (LVEF) and fractional shortening (LVFS) by 9.62% and 5.82%, respectively, attenuated the reduction in HWI (4.41 mg/mm vs 3.53 mg/mm), and improved the 4-week survival rate (80% vs 60%). Transcriptomic analysis revealed that the core molecular features of DOX-induced cardiotoxicity, including widespread suppression of oxidative phosphorylation and antioxidant pathways. Conclusions HydroMg protects against DOX-induced chronic cardiotoxicity by alleviating oxidative stress and improving mitochondrial function, and provides a potential therapeutic strategy against DOX-related cardiotoxicity.