Objective:To develop and evaluate a medical visual question answering (VQA) model for neonatal lung ultrasound (LUS) images to enhance intelligent auxiliary diagnosis of neonatal pulmonary diseases.Methods:Using data from neonates admitted to Beijing Obstetrics and Gynecology Hospital, Capital Medical University (January 2023 to December 2024), an image-question-answer dataset comprising 251 LUS images was constructed [43 pneumonia (17.1%), 42 neonatal respiratory distress syndrome (16.7%), 83 transient tachypnea (33.1%), and 83 normal (33.1%) images] with a four-tier medical question-answer framework. Building upon the Qwen2.5-VL-7B base model and integrating LoRA fine-tuning with chain-of-thought prompting, we developed the NLUS-VQA model to enhance visual-language semantic alignment and enable stepwise clinical reasoning, achieving efficient small-sample adaptation. Model performance was comprehensively assessed through natural language generation metrics (BLEU-4, ROUGE-1/2/L), qualitative evaluation of characteristic recognition, and clinical consistency analysis.Results:(1) Quantitative evaluation demonstrated that NLUS-VQA achieved scores of 22.38 (BLEU-4), 48.26 (ROUGE-1), 22.40 (ROUGE-2), and 37.20 (ROUGE-L), representing significant improvements over baseline models. (2) Qualitatively, the model exhibited strong performance in identifying lung consolidation, coalescent B-lines, and snowflake signs, with its chain-of-thought strategy enhancing clinical interpretability and answer accuracy. (3) Clinically, NLUS-VQA achieved a Cohen's Kappa coefficient of 0.78 and diagnostic accuracy of 80.8% (21/26), indicating substantial agreement with clinical experts.Conclusion:The NLUS-VQA model demonstrates robust interpretability in recognizing key sonographic patterns (e.g. lung consolidation, confluent B-lines, and snowflake signs), providing a scalable framework for small-sample medical image analysis, though diagnostic performance on complex conditions remains limited by dataset scale and minority class representation.

Objective:To develop and evaluate a medical visual question answering (VQA) model for neonatal lung ultrasound (LUS) images to enhance intelligent auxiliary diagnosis of neonatal pulmonary diseases.Methods:Using data from neonates admitted to Beijing Obstetrics and Gynecology Hospital, Capital Medical University (January 2023 to December 2024), an image-question-answer dataset comprising 251 LUS images was constructed [43 pneumonia (17.1%), 42 neonatal respiratory distress syndrome (16.7%), 83 transient tachypnea (33.1%), and 83 normal (33.1%) images] with a four-tier medical question-answer framework. Building upon the Qwen2.5-VL-7B base model and integrating LoRA fine-tuning with chain-of-thought prompting, we developed the NLUS-VQA model to enhance visual-language semantic alignment and enable stepwise clinical reasoning, achieving efficient small-sample adaptation. Model performance was comprehensively assessed through natural language generation metrics (BLEU-4, ROUGE-1/2/L), qualitative evaluation of characteristic recognition, and clinical consistency analysis.Results:(1) Quantitative evaluation demonstrated that NLUS-VQA achieved scores of 22.38 (BLEU-4), 48.26 (ROUGE-1), 22.40 (ROUGE-2), and 37.20 (ROUGE-L), representing significant improvements over baseline models. (2) Qualitatively, the model exhibited strong performance in identifying lung consolidation, coalescent B-lines, and snowflake signs, with its chain-of-thought strategy enhancing clinical interpretability and answer accuracy. (3) Clinically, NLUS-VQA achieved a Cohen's Kappa coefficient of 0.78 and diagnostic accuracy of 80.8% (21/26), indicating substantial agreement with clinical experts.Conclusion:The NLUS-VQA model demonstrates robust interpretability in recognizing key sonographic patterns (e.g. lung consolidation, confluent B-lines, and snowflake signs), providing a scalable framework for small-sample medical image analysis, though diagnostic performance on complex conditions remains limited by dataset scale and minority class representation.

OBJECTIVE：To provide reference f or the qualit y sta ndard establishment of Amaranthus retroflexus. METHODS ： Taking 7 batches of A. retroflexus medicinal materials as the research object ，the appearance properties of the medicinal materials were investigated ，and the microscopic characteristics of the medicinal powders were observed. TLC method was adopted to qualitatively identify rutin ，valine and leucine in A. retroflexus medicinal materials. According to the relevant methods of the 2015 edition of Chinese Pharmacopoeia （part Ⅳ），water content ，total ash content ，acid-insoluble ash content and water-soluble extract content were determined. HPLC method was used to determine the content of rutin in the medicinal material of A. retroflexus . The determination was performed on Agilent 5 TC-C18（2）column with mobile phase consisted of methanol- 0.3％ phosphoric acid solution（40∶60，V/V），at the flow rate of 1.0 mL/min. The detection wavelength was set at 358 nm，and the column temperature was 30 ℃. The sample size was 10 μL. RESULTS：The appearance and microstructure characteristics of the medicinal materials were consistent with the existing description. The identification results of TLC meth od showed that 7 batches of medicinal materials and each reference substance （rutin，valine，leucine）showed spots of the same color at the same position. The moisture content of 7 batches of A. retroflexus medicinal materials was 7.43％-8.72％，the total ash content was 11.82％-13.78％，the acid-insoluble ash content was 0.15％-0.55％，and the water-soluble extract content was 17.27％-24.74％. The linear range of rutin was 10-200 μg/mL（R 2＝1.000 0）. RSDs of precision test ，stability test （24 h）and repeatability test were all less than 2.0％ （n＝6）. The average recovery rates of rutin were 99.14％，97.98％ and 98.80％ in low ，medium and high concentration of samples，and RSDs were 0.97％，0.95％，0.96％（n＝3）. The contents of rutin in 7 batches of A. retrophylla were 0.314-1.102 mg/g. CONCLUSIONS：In this study ，character observation ，microscopic identification ，moisture content ，total ash content ，acid- insoluble ash content and water-soluble extract content of A. retroflexus are investigated ；TLC method was established for qualitative identification of leucine ，valine and rutin in A. retroflexus ，and the HPLC method was established for content determination of rutin. It provides reference for the quality standard establishment of A. retroflexus .

Objective:To investigate the clinical efficacy of superficial temporal artery-anterior temporal artery bypass combined with encephalo-duro-arterio-synangiosis in the treatment of Moyamoya disease.Methods:A total of 42 patients with moyamoya disease, admitted to our hospital from January 2016 to January 2019, were selected and divided into observation group and control group according to surgical management. The patients in the observation group were treated with superficial temporal artery-anterior temporal artery bypass combined with encephalo-duro-arterio-synangiosis, and the patients in the control group were treated with superficial temporal artery-angular gyrus artery bypass combined with encephalo-duro-arterio-synangiosis. The clinical data of these patients were analyzed retrospectively, and the differences of efficacy and safety between the two groups were compared.Results:There was no significant difference in operation time between the two groups ([189.16±21.23] min vs. [179.46±16.95] min, P>0.05). One d after the operation, the patients in both groups were re-examined with CT angiography, and the anastomotic vessels were unobligated. Two patients in the observation group had cerebral infarction in the operative region and one patient in the control group had cerebral infarction in the operative region; no significant difference was noted in the incidence of postoperative complications between the two groups ( P>0.05). The modified Rankin scale (mRS) scores in both groups one month after surgery were significantly decreased as compared with those one d before surgery ( P<0.05); one month after surgery, the mRS scores in observation group (0.13±0.346) were significantly lower than those in the control group (0.42±0.515, P<0.05). Conclusion:The superficial temporal artery-anterior temporal artery bypass combined with encephalo-duro-arterio-synangiosis has definite clinical efficacy in the treatment of moyamoya disease.