Chunting Tan

Background: Pneumonia is a common infectious disease, and antibiotic treatment in hospitalized patients must balance efficacy, safety, and resistance risk. However, antibiotic selection and dose adjustment still rely heavily on clinician experience. Although large language models (LLMs) are promising for clinical reasoning, their direct use for antibiotic selection and dose recommendation is limited by hallucinations and weak adherence to clinical constraints. Objective: This study aimed to develop and externally validate a constrained LLM-based clinical decision support pipeline for antibiotic selection and dose recommendation in hospitalized patients with pneumonia. Methods: We conducted a multicenter retrospective study using electronic health record narratives, antibiotic orders, and laboratory indicators of hepatic and renal function from 331 hospitalized patients with pneumonia from 2 hospitals in China. The development cohort included 233 patients, and the external validation cohort included 98 patients. The pipeline integrated dual-branch retrieval (similar-case vector retrieval plus guideline-based knowledge graph retrieval), clinician-defined rule constraints, and hybrid-context reasoning. DeepSeek-V3, GLM-4.6, and GPT-4o were evaluated using F ₁ -score and Jaccard accuracy. Results: On the internal test set, the full pipeline using DeepSeek-V3 achieved the best performance, with an F ₁ -score of 0.8110 (95% CI 0.7371-0.8762) and Jaccard accuracy of 0.7624 (95% CI 0.6810-0.8386) for antibiotic selection and an F ₁ -score of 0.7538 (95% CI 0.6671-0.8329) and Jaccard accuracy of 0.7076 (95% CI 0.6145-0.7938) for joint antibiotic selection plus dosing recommendation. On the external validation set, performance remained high, with an F ₁ -score of 0.8605 (95% CI 0.7891-0.9252) and Jaccard accuracy of 0.8571 (95% CI 0.7857-0.9184) for antibiotic selection, and an F ₁ -score of 0.8503 (95% CI 0.7789-0.9150) and Jaccard accuracy of 0.8469 (95% CI 0.7755-0.9133) for antibiotic selection plus dosing recommendation. The system also provided traceable evidence and rule trigger information to support clinician review. Conclusions: A constrained, retrieval-augmented LLM pipeline improved the consistency and interpretability of antibiotic selection and dose recommendation for hospitalized patients with pneumonia and provided preliminary evidence of cross-site generalizability.

Background: The rapid advancements in natural language processing, particularly the development of large language models (LLMs), have opened new avenues for managing complex clinical text data. However, the inherent complexity and specificity of medical texts present significant challenges for the practical application of prompt engineering in diagnostic tasks. Objective: This paper explores LLMs with new prompt engineering technology to enhance model interpretability and improve the prediction performance of pulmonary disease based on a traditional deep learning model. Methods: A retrospective dataset including 2965 chest CT radiology reports was constructed. The reports were from 4 cohorts, namely, healthy individuals and patients with pulmonary tuberculosis, lung cancer, and pneumonia. Then, a novel prompt engineering strategy that integrates feature summarization (F-Sum), chain of thought (CoT) reasoning, and a hybrid retrieval-augmented generation (RAG) framework was proposed. A feature summarization approach, leveraging term frequency–inverse document frequency (TF-IDF) and K-means clustering, was used to extract and distill key radiological findings related to 3 diseases. Simultaneously, the hybrid RAG framework combined dense and sparse vector representations to enhance LLMs’ comprehension of disease-related text. In total, 3 state-of-the-art LLMs, GLM-4-Plus, GLM-4-air (Zhipu AI), and GPT-4o (OpenAI), were integrated with the prompt strategy to evaluate the efficiency in recognizing pneumonia, tuberculosis, and lung cancer. The traditional deep learning model, BERT (Bidirectional Encoder Representations from Transformers), was also compared to assess the superiority of LLMs. Finally, the proposed method was tested on an external validation dataset consisted of 343 chest computed tomography (CT) report from another hospital. Results: Compared with BERT-based prediction model and various other prompt engineering techniques, our method with GLM-4-Plus achieved the best performance on test dataset, attaining an F₁-score of 0.89 and accuracy of 0.89. On the external validation dataset, F₁-score (0.86) and accuracy (0.92) of the proposed method with GPT-4o were the highest. Compared to the popular strategy with manually selected typical samples (few-shot) and CoT designed by doctors (F₁-score=0.83 and accuracy=0.83), the proposed method that summarized disease characteristics (F-Sum) based on LLM and automatically generated CoT performed better (F₁-score=0.89 and accuracy=0.90). Although the BERT-based model got similar results on the test dataset (F₁-score=0.85 and accuracy=0.88), its predictive performance significantly decreased on the external validation set (F₁-score=0.48 and accuracy=0.78). Conclusions: These findings highlight the potential of LLMs to revolutionize pulmonary disease prediction, particularly in resource-constrained settings, by surpassing traditional models in both accuracy and flexibility. The proposed prompt engineering strategy not only improves predictive performance but also enhances the adaptability of LLMs in complex medical contexts, offering a promising tool for advancing disease diagnosis and clinical decision-making.

Background: We evaluated the effects of drug therapy on T lymphocyte subsets and their associations with recurrent chronic bronchitis (CB) attacks. Methods: A total of 162 CB patients treated from April 2020 to April 2021 were selected. All patients underwent anti-infective, cough-relieving, and phlegm-eliminating treatment, as detailed in Clinical Pathway for Chronic Bronchitis. They were divided into a recurrent attack group (n=95) and a non-recurrent attack group (n=67). Changes in T lymphocyte subsets at different time points of treatment and their associations with the number of attacks were analyzed. Associated factors were analyzed in a multivariate logistic regression model, and their predictive value was validated using a nomogram prediction model and receiver operating characteristic (ROC) curves. Results: The number of attacks at 1 year after treatment was positively correlated with CD8+ and negatively correlated with CD3+, CD4+, and CD4+/CD8+. A history of smoking, CD3+, CD4+, CD8+, and CD4+/CD8+ were independent risk factors for recurrent attacks. The nomogram prediction model showed that the total risk value corresponding to smoking history, low CD3+, CD4+, and CD4+/CD8+, and elevated CD8+ was 0.86, and the concordance index and area under the ROC curve of the model was 0.896 (95% CI: 0.782-0.997, P<0.05) and 0.816 (95% CI: 0.803-0.847, P<0.001), respectively. There was an association between T lymphocyte subsets and recurrent attacks before and after treatment of CB. Conclusions: Low CD3+, CD4+/CD8+ and CD4+, elevated CD8+, and smoking history were risk factors for recurrent attack.

Summary: The mesothelium is a squamous monolayer that ensheathes internal organs, lines the body cavity, and the diaphragm. It serves as a protective barrier, coated in glycocalyx, and secretes lubricants to facilitate tissue movement. How the mesothelium forms is poorly understood. Here, we investigate Myrf , a transcription factor gene expressed in the mesothelium, because it carries variants in patients with Congenital Diaphragmatic Hernia (CDH), a disorder that affects the diaphragm, lung, and other organs. In mice, inactivation of Myrf early in organogenesis resulted in CDH and defective mesothelial specification, compromising its function as a signaling center for lung growth. Inactivation of Myrf later led to enhanced mesothelium differentiation into mesenchymal cell types through partial epithelial-to-mesenchymal transition (EMT), resulting in a unique accumulation of smooth muscle encasing the lung. In this role, MYRF functions in parallel with YAP/TAZ. Together, these findings establish MYRF as a critical regulator of mesothelium development, and when mutated, causes CDH.

SCIEPublish is an international open-access journal publishing service provider run by SCIE Publishing Limited, dedicated to supporting and inspiring scientists or institutions/societies who aspire to publish high-quality journals.

Idiopathic pulmonary fibrosis (IPF) is a progressive and chronic disease that significantly impacts patient quality of life, and its incidence is on the rise. The pathogenesis of IPF remains poorly understood. Alveolar type 2 (AT2) cells are crucial in the onset and progression of IPF, yet the specific mechanisms involved are not well defined. Lon protease 1 (LONP1), known for its critical roles in various diseases, has an unclear function in IPF. Our research investigated the impact of Lonp1 gene deletion on AT2 cell functionality and its subsequent effect on IPF development. We generated a bleomycin-induced pulmonary fibrosis mouse model with a targeted Lonp1 knockout in AT2 cells and assessed the consequences on AT2 cell function and fibrosis progression. Additionally, we constructed the MLE12 cells with stable Lonp1 knockdown and utilized transcriptome sequencing to identify pathways altered by the Lonp1 knockdown. Our results indicated that mice with AT2 cell-specific Lonp1 knockout exhibited more severe fibrosis compared to controls. These mice exhibited a reduction in AT2 and AT1 cell populations, along with an increase in p53- and p21-positive AT2 cells. Lonp1 knockdown in MLE12 cells led to the upregulation of aging-associated pathways, with fibroblast growth factor 2 (Fgf2) gene emerging as a central gene interconnecting these pathways. Therefore, loss of Lonp1 appears to promote AT2 cell aging and exacerbate bleomycin-induced pulmonary fibrosis. Fgf2 emerges as a pivotal downstream gene associated with cellular senescence. This study uncovers the role of the Lonp1 gene in pulmonary fibrosis, presenting a novel target for investigating the pathological mechanisms and potential therapeutic approaches for IPF.

Idiopathic pulmonary fibrosis (IPF) is a progressive and irreversible pulmonary interstitial disease that seriously affects the patient’s quality of life and lifespan. The pathogenesis of IPF has not been clarified, and its treatment is limited to pirfenidone and nintedanib, which only delays the decline of lung function. Alveolar epithelial type 2 (AT2) cells are indispensable in the regeneration and lung surfactant secretion of alveolar epithelial cells. Studies have shown that AT2 cell dysfunction initiates the occurrence and progression of IPF. This review expounds on the AT2 cell dysfunction in IPF, involving senescence, apoptosis, endoplasmic reticulum stress, mitochondrial damage, metabolic reprogramming, and the transitional state of AT2 cells. This article also briefly summarizes potential treatments targeting AT2 cell dysfunction.