基于ASP的中医临床实践指南自动执行模型研究-《计算机技术与发展》

文章信息/Info

Title:: Research on an ASP-based Automatic Execution Model for Clinical Practice Guidelines of Chinese Medicine

作者:: 张天瑜1; 张志政2; 张魁星1; 李延军1; 1. 山东中医药大学智能与信息工程学院,山东济南 250355;
2. 东南大学计算机科学与工程学院,江苏南京 211189

Author(s):: ZHANG Tian-yu1; ZHANG Zhi-zheng2; ZHANG Kui-xing1; LI Yan-jun1; 1. School of Intelligence and Information Engineering,Shandong University of Traditional Chinese Medicine,Jinan 250355,China;
2. School of Computer Science and Engineering,Southeast University,Nanjing 211189,China

Keywords:: answer set programming; clinical practice guideline of Chinese medicine; knowledge representation; disease diagnosis; automatic execution

摘要:: 中医临床实践指南是官方发布的标准化临床指导意见,由于其文本形式内容繁多复杂,不利于进行实际推广,而目前面向中医临床实践指南自动执行的研究仍然很不完善。为此,以世界中医药学会联合会发布的《中医临床实践指南慢性阻塞性肺疾病》为例,研究基于 ASP(回答集程序)的中医临床实践指南自动执行的模型。具体地,该文提出了一种面向中医临床实践指南的疾病诊断知识表示模型,实现了基于回答集程序的模型表示和自动执行以及在特定医疗交互场景下的自动执行。根据用户提供的症状信息,能够反馈诊断结果或进一步诊断所需考虑的最小范围的其他症状。最后,将该方法与大语言模型进行了对比性实验,实验结果表明,该模型具有良好的准确性和可靠性。

Abstract:: The clinical practice guidelines for traditional Chinese medicine (TCM) are official standardized clinical guidance recommen-dations,but due to the complexity and diversity of their textual content,they are not conducive to practical promotion. Currently,research on automatic execution of TCM clinical practice guidelines for clinical practice is still far from perfect. Therefore, we take the " Guidelines for the Clinical Practice of TCM for Chronic Obstructive Pulmonary Disease" issued by the World Federation of Chinese Medicine Societies as an example and study the model of automatic execution of TCM clinical practice guidelines based on ASP (Answer Set Programming). Specifically,we propose a disease diagnosis knowledge representation model for TCM clinical practice guidelines,achieve the model representation and automatic execution based on ASP, and automatic execution in specific medical interaction scenarios. Based on the symptom information provided by the user,it can feedback the diagnosis result or the minimum range of other symptoms to be considered for further diagnosis. Finally,the experimental results of comparing the proposed method with large language models show that the model has good accuracy and reliability.