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 prepare a peptide fluorescent probe based on aggregation-induced emission and to investigate its application in the detection of early caries.Methods:Eight aspartate-serine-serine (DSS) were combined with aggregation-induced emission material to prepare peptide fluorescent probes, and an artificial demineralization model was established in vitro. The samples were immersed in the peptide fluorescent probe solution for 1 min, and a fluorescence imaging system was applied to examine the tooth samples and collect images and fluorescence data. Scanning electron microscopy was also applied to observe the phenotype of the teeth, and electron microscopy was applied to detect the calcium-phosphorus ratio on the enamel surface of the teeth. Polarized light microscopy was also applied to observe the enamel area of the teeth. Results:The fluorescence intensity of demineralized teeth was clearly observed to be lower than that of normal teeth in the peptide fluorescent probe-treated area, and the difference was statistically significant ( P < 0.05). The results of scanning electron microscopy showed that the enamel surface of the demineralized group had more irregular pores, while the enamel surface of the undemineralized group was flatter with only some irregular accumulation of flakes. The results of polarized light microscopy showed that a clear birefringence could be observed in the enamel region of normal teeth, while a black area or the disappearance of the birefringence effect accompanied by a partial black dark shadow could be observed in the enamel region of demineralized teeth. Conclusions:An aggregation-induced luminescence-based peptide fluorescent probe was successfully prepared, which can precisely localize the enamel and show some application value in early caries detection.

Circular RNAs(circRNAs)are involved in various biological processes and disease pathogenesis.However,only a small number of functional circRNAs have been identified among hundreds of thousands of circRNA species,partly because most current methods are based on cir-cular junction counts and overlook the fact that a circRNA is formed from the host gene by back-splicing(BS).To distinguish the expression difference originating from BS or the host gene,we pre-sent differentially expressed back-splicing(DEBKS),a software program to streamline the discov-ery of differential BS events between two rRNA-depleted RNA sequencing(RNA-seq)sample groups.By applying to real and simulated data and employing RT-qPCR for validation,we demon-strate that DEBKS is efficient and accurate in detecting circRNAs with differential BS events between paired and unpaired sample groups.

To investigate the accuracy and feasibility of Brix value on monitoring gastric residual volume (GRV) in patients with enteral nutrition. Fifty patients with enteral nutrition via nasogastric tube were enrolled. The GRV was measured by both ultrasonography and Brix value. The results were compared according to the methods. The Pearson correlation coefficients showed that GRV measured by these two ways was positively correlated (r=0.986, P<0.05). Moreover paired sample t-test showed that the discrepancy was not statistically significant (P>0.05) between different measurements. The consistency was analyzed by Bland-Altman graph, showing that the two measurements were consistent. Brix value is recommended to measure GRV due to its convenience and easy operation.

Objective Collagenous gastritis is a rare entity, characterized by the deposition of a subepithelial collagen band with an inflammatory infiltrate in the mucosa.This report describes the first case of collagenous gastritis occurring in a young Chinese woman and reviews the literatures.Methods The patient underwent the gastroscopy screening, and the biopsy specimens were treated with HE staining, Masson staining, Congo red staining and Warthin-Starry staining.Patients' clinical data was discussed and followed up.Results A twenty-year-old girl had intermittent epigastric pain for 4 years, abdominal distention, hiccup and weight loss for two months.The gastric endoscopy revealed diffuse white nodular appearance of the mucosa in angular incisure and antrum.Pathologic examination of the gastric biopsies from the antrum and angular showed a subepithelial collagen deposition with moderate infiltrates of lymphoplasma cells and eosinophils of the lamina propria.The collagen band measured up to 120.3 μm (mean 43.8 μm).Prednisone 20 mg/d for 4 weeks led to clinical remission and weight gain.Conclusion There are about 40 cases in literatures to date, and the cause and pathogenesis of collagenous gastritis remain unknown.According to the clinical and pathological characteristics, the patient in this article is the subtype of collagenous gastritis that occurring in children and young adults.Specific therapy has not been established, the gluten-free diet and glucocorticosteroid may be helpful to relieve symptoms in collagenous gastritis patients.