CPMI-ChatGLM: parameter-efficient fine-tuning ChatGLM with Chinese patent medicine instructions.

Chinese patent medicine (CPM) is a typical type of traditional Chinese medicine (TCM) preparation that uses Chinese herbs as raw materials and is an important means of treating diseases in TCM. Chinese patent medicine instructions (CPMI) serve as a guide for patients to use drugs safely and effectively. In this study, we apply a pre-trained language model to the domain of CPM. We have meticulously assembled, processed, and released the first CPMI dataset and fine-tuned the ChatGLM-6B base model, resulting in the development of CPMI-ChatGLM. We employed consumer-grade graphics cards for parameter-efficient fine-tuning and investigated the impact of LoRA and P-Tuning v2, as well as different data scales and instruction data settings on model performance. We evaluated CPMI-ChatGLM using BLEU, ROUGE, and BARTScore metrics. Our model achieved scores of 0.7641, 0.8188, 0.7738, 0.8107, and - 2.4786 on the BLEU-4, ROUGE-1, ROUGE-2, ROUGE-L and BARTScore metrics, respectively. In comparison experiments and human evaluation with four large language models of similar parameter scales, CPMI-ChatGLM demonstrated state-of-the-art performance. CPMI-ChatGLM demonstrates commendable proficiency in CPM recommendations, making it a promising tool for auxiliary diagnosis and treatment. Furthermore, the various attributes in the CPMI dataset can be used for data mining and analysis, providing practical application value and research significance.

GastroBot: a Chinese gastrointestinal disease chatbot based on the retrieval-augmented generation.

Introduction: Large Language Models (LLMs) play a crucial role in clinical information processing, showcasing robust generalization across diverse language tasks. However, existing LLMs, despite their significance, lack optimization for clinical applications, presenting challenges in terms of illusions and interpretability. The Retrieval-Augmented Generation (RAG) model addresses these issues by providing sources for answer generation, thereby reducing errors. This study explores the application of RAG technology in clinical gastroenterology to enhance knowledge generation on gastrointestinal diseases. Methods: We fine-tuned the embedding model using a corpus consisting of 25 guidelines on gastrointestinal diseases. The fine-tuned model exhibited an 18% improvement in hit rate compared to its base model, gte-base-zh. Moreover, it outperformed OpenAI's Embedding model by 20%. Employing the RAG framework with the llama-index, we developed a Chinese gastroenterology chatbot named "GastroBot," which significantly improves answer accuracy and contextual relevance, minimizing errors and the risk of disseminating misleading information. Results: When evaluating GastroBot using the RAGAS framework, we observed a context recall rate of 95%. The faithfulness to the source, stands at 93.73%. The relevance of answers exhibits a strong correlation, reaching 92.28%. These findings highlight the effectiveness of GastroBot in providing accurate and contextually relevant information about gastrointestinal diseases. During manual assessment of GastroBot, in comparison with other models, our GastroBot model delivers a substantial amount of valuable knowledge while ensuring the completeness and consistency of the results. Discussion: Research findings suggest that incorporating the RAG method into clinical gastroenterology can enhance the accuracy and reliability of large language models. Serving as a practical implementation of this method, GastroBot has demonstrated significant enhancements in contextual comprehension and response quality. Continued exploration and refinement of the model are poised to drive forward clinical information processing and decision support in the gastroenterology field.

Identifying strategic human resource management ability in the clinical departments of public hospitals in China: a modified Delphi study.

OBJECTIVES: Chinese public hospitals are managed like a bureaucracy, which is divided into two levels of hospital and departmental management. Improving strategic human resource management ability (SHRMA) within clinical departments can improve department performance and service quality, which is an important way for public hospitals to obtain an advantage in a diversified competitive medical market. However, there is a lack of specialised evaluation tools for SHRMA in clinical departments to support this effort. Therefore, this study aims to develop an index for evaluating the SHRMA of clinical departments in public hospitals. STUDY DESIGN AND SETTING: The Delphi technique was carried out with 22 experts, and an evaluation index of the SHRMA in the clinical departments of public hospitals was constructed. The weight of each indicator was calculated by the intuitive fuzzy analytic hierarchy process. RESULTS: The SHRMA index constructed in this study for the clinical departments in public hospitals includes 5 first-level indicators, 13 second-level indicators and 36 third-level indicators. The first-level indicators are distributed in weight among human resource maintenance (0.204), human resource planning (0.201), human resource development (0.200), human resource stimulation (0.198) and human resource absorption (0.198). The top three weighted indicators on the second level are job analysis and position evaluation (0.105), career management (0.103) and salary incentivisation (0.100). CONCLUSIONS: The index constructed in this study is scientific and feasible and is expected to provide an effective tool for the quantitative evaluation of SHRMA in the clinical departments of public hospitals in China.

Effects of "audience effects"on animal mate choice: A review.

The information tranfered among individual animals can be shared by a network, which is consisted of the sender, the receiver, and the extra bystander of the communication signals. The bystanders can read and use the signal that is not sent directly to them and make use of it to interfere with the sender and the receiver, which is known as "audience effects" in the research area of animal behaviors. The processes of mate choice and mating of animals occur mainly in the network that is composed of the particular species. Increasing evidence show that the audience effects play an important role in regulating mating preference and mating strategy, resulting in changes in species evolution. Here, we review the role of audience effects on animal mate choice and evolution by clarifying the definition and functional explanations of audience effects, the factors contributing to audience effects, as well as the different impacts of audience effects on males and females. It would provide novel ideas to study the impacts of audience effects on mate choice and species evolution in the future.

Cognitive impairment in diffuse axonal injury patients with favorable outcome.

Background and purpose: Traumatic brain injury (TBI), especially the severe TBI are often followed by persistent cognitive sequalae, including decision-making difficulties, reduced neural processing speed and memory deficits. Diffuse axonal injury (DAI) is classified as one of the severe types of TBI. Part of DAI patients are marginalized from social life due to cognitive impairment, even if they are rated as favorable outcome. The purpose of this study was to elucidate the specific type and severity of cognitive impairment in DAI patients with favorable outcome. Methods: The neurocognition of 46 DAI patients with favorable outcome was evaluated by the Chinese version of the Montreal Cognitive Assessment Basic (MoCA-BC), and the differences in the domains of cognitive impairment caused by different grades of DAI were analyzed after data conversion of scores of nine cognitive domains of MoCA-BC by Pearson correlation analysis. Results: Among the 46 DAI patients with favorable outcome, eight had normal cognitive function (MoCA-BC ≥ 26), and 38 had cognitive impairment (MoCA-BC < 26). The MoCA-BC scores were positively correlated with pupillary light reflex (r = 0.361, p = 0.014), admission Glasgow Coma Scale (GCS) (r = 0.402, p = 0.006), and years of education (r = 0.581, p < 0.001). Return of consciousness (r = -0.753, p < 0.001), Marshall CT (r = -0.328, p = 0.026), age (r = -0.654, p < 0.001), and DAI grade (r = -0.403, p = 0.006) were found to be negatively correlated with the MoCA-BC scores. In patients with DAI grade 1, the actually deducted scores (Ads) of memory (r = 0.838, p < 0.001), abstraction (r = 0.843, p < 0.001), and calculation (r = 0.782, p < 0.001) were most related to the Ads of MoCA-BC. The Ads of nine cognitive domains and MoCA-BC were all proved to be correlated, among patients with DAI grade 2. However, In the DAI grade 3 patients, the highest correlation with the Ads of MoCA-BC were the Ads of memory (r = 0.904, p < 0.001), calculation (r = 0.799, p = 0.006), orientation (r = 0.801, p = 0.005), and executive function (r = 0.869, p = 0.001). Conclusion: DAI patients with favorable outcome may still be plagued by cognitive impairment, and different grades of DAI cause different domains of cognitive impairment.