基于大语言模型的肺癌中药处方推荐研究

Research on Lung Cancer Traditional Chinese Medicine Prescription Recommendation Based on Large Language Models

  • 摘要:
    目的 针对特定领域中医处方推荐问题,充分利用中医专家的肺癌临床病案来自动化生成处方,为用药规律研究、中医临床辅助决策提供参考。
    方法 通过大语言模型较强的生成能力设计一种中医处方推荐算法,将临床表现、标准化舌象、脉象等通过大模型转化为中药处方,从而将中药处方推荐任务转化为文本生成任务,利用基于GLM结构的CHATGLM3模型来加强对肺癌病案的理解,学习中医专家治疗肺癌的内在经验知识,提升模型处方生成的效果,并与传统的生成式模型进行比较。
    结果 将肺癌病案内的中医知识融入大语言模型中可有效提升模型的处方生成能力,特别是在生成中医专家常用核心药物方面,模型表现出了较高的倾向性,能够提供丰富且有价值的参考信息。肺癌中药处方推荐模型在BLEU、ROUGE、METEOR指标上取得了64.62%、55.78%、47.39%的效果,并且在前5、10、15、20味中药处方中取得67.79%、63.66%、56.76%、51.93%的准确率,优于基线模型。
    结论 肺癌中药处方推荐模型相较于传统生成式模型取得了较好的处方生成效果,表明其可以从病案中学习肺癌诊疗方面的知识,从而生成符合中医治疗原则的中药处方,也为未来辅助临床决策等提供可能的方向。

     

    Abstract:
    OBJECTIVE To address the issue of recommending traditional Chinese medicine (TCM) prescriptions in specific fields, and to fully utilize the clinical records of lung cancer from TCM experts to automatically generate prescriptions, providing reference for the study of medication rules and TCM clinical decision-making assistance.
    METHODS A TCM prescription recommendation algorithm was designed using the strong generative capabilities of large language models. This algorithm transformed clinical manifestations, standardized tongue diagnosis, and pulse diagnosis into TCM prescriptions through a large model, thereby converting the task of TCM prescription recommendation into a text generation task. The CHATGLM3 model, based on the GLM structure, was used to enhance the understanding of lung cancer cases and learn the intrinsic experiential knowledge of TCM experts in treating lung cancer, thereby improving the prescription generation effectiveness of the model. This was compared with traditional generative models.
    RESULTS The study demonstrated that integrating TCM knowledge from lung cancer cases into large language models effectively improved the model's prescription generation capabilities. Particularly in generating commonly used core medications by TCM experts, the model showed a high tendency and provided rich and valuable reference information. The lung cancer TCM prescription recommendation model achieved 64.62% in BLEU, 55.78% in ROUGE, and 47.39% in METEOR scores. It also achieved accuracies of 67.79%, 63.66%, 56.76%, and 51.93% in the top 5, 10, 15, and 20 TCM prescriptions, respectively, outperforming the baseline model.
    CONCLUSION The lung cancer TCM prescription recommendation model presented in this paper achieves better prescription generation results compared to traditional generative models. It demonstrates the model's ability to learn knowledge about lung cancer diagnosis and treatment from cases, thereby generating TCM prescriptions that align with TCM treatment principles. This also provides a potential direction for future assistance in clinical decision-making.

     

/

返回文章
返回