基于语义网络的肾系疾病古籍本体知识库构建及方证规律挖掘

Chronic kidney disease (CKD) has high prevalence and mortality and causes a heavy economic burden. How to effectively delay the progress of CKD has always been a hot research topic in the field of kidney disease. Traditional Chinese medicine has an advantage on prevention and control of CKD through multiple targets therapy, due to the characteristics of holistic concept and treatment based on syndrome differentiation. There are a large amount of kidney disease treatment related provisions recorded in TCM ancient books, which bearing for practical experience of traditional Chinese medicine during thousands of years. Ancient book mining has always been the source of Chinese scholars inheriting work. Our previous research found that there is a large amount of kidney disease treatment recorded in the ancient books. However, troubles like lack of digital ancient book resources, difficulty to find tacit knowledge, and divorce between formulas mining and clinical performance hampered ancient literature research. The semantic network and ontology in the field of artificial intelligence technology combined with disease knowledge contribute to the concept of tacit knowledge expression and mutual relationship to be explicit, and could help to build ancient books database based on knowledge discovery. Therefore, this study intends to construct ancient Chinese medical ontology knowledge database of kidney disease based on semantic network. During the study, ontologies of kidney disease syndrome, symptoms, drugs, prescriptions will be established, standardized and semantic tagged according to the characteristics of kidney disease. Semantic relation network will be built to explicit the relationships among these ontologies. The objective of this study is to explore the syndrome and therapeutic principles of kidney disease in ancient books, which would show the real way that ancient doctors did syndrome differentiation and treatment of renal disease.

慢性肾脏病患病率和病死率高，医疗费用昂贵。如何有效延缓慢性肾脏病进展一直是肾病领域的研究热点。中医整体观念和辨证论治的特点使中医药防治慢性肾脏病具有多靶点的治疗优势。承载中医学知识的古籍一直是学者传承中医经验的活水源头。课题组前期研究发现，古籍中记载了大量肾病治疗相关条文，但是大量隐性知识难以发掘，古籍挖掘无法深入到方证规律的探索。人工智能领域的语义网络和本体技术与疾病专业领域知识相结合，有助于隐性知识的概念表达和相互关系显性化，构建基于知识发现的专病古籍数据库。目前尚没有适用于肾系疾病古籍挖掘的本体知识库。因此，本研究拟构建基于语义网络的肾系疾病中医古籍本体知识库，根据肾系疾病特点对古籍知识概念进行规范化及语义标注，通过语义关系网络的实现，从“以病带方”，“以证施方”“以症用药”三个维度，探索肾系疾病辨证施方、方随证变的微调规律，将古代医家辨证论治肾系疾病的思维与用药经验真实展现。

中医药治疗慢性肾脏病具有多靶点，多途径的治疗优势，已成为公认的治疗慢性肾脏病新药开发的资源宝库。中医古籍是传统医学传承、创新、发展的源头活水。但是由于古籍数字化资源的发展水平参差不齐，古籍中蕴含的大量隐性知识得不到充分挖掘。本研究与国内顶尖中医药信息技术团队紧密合作，借助现代信息技术，通过知识本体构建和语义网络关联，结合肾病专业领域知识，构建了基于语义网络的肾系疾病中医古籍本体知识库及辅助挖掘系统。在此基础上，课题组以肾系疾病“水肿”，“癃闭”，“关格”“肾风”，“肾劳”，“肾消”，“下消”，“消肾”的古籍条文为研究对象，完成病、证、症状、药物、处方、治则治法、医籍、医家本体构建。通过形成以疾病为纲，方证关系、药症关系为核心的语义关联网络，采用复杂网络分析，从“以病带方”，“以证施方”“以症用药”三个维度，挖掘肾系疾病方证规律，探索辨证施方、方随证变的微调规律。研究通过构建“疾病-方剂”，“疾病-中药”关联进行以病带方的核心方药挖掘；构建“治则治法-朝代”关联，得出治则治法和选方用药随历史朝代变迁的规律特点；构建“治则治法-医家”，“方剂-医家”关联，得出名医医家治疗肾系疾病的治则治法及选方用药特点；构建“证候-中药”，“证候-方剂”关联，进行以证施方的辨证规律挖掘。构建“症状-中药”关联,提取核心症状群及相应方药群，并在此基础上初步探索随症加减的用药规律。. 本研究将现代信息技术应用于中医古籍研究领域，再现古代医家辨证施治的思维过程。对比现有教材专著，研究结果丰富了现代中医家治疗肾系疾病的选方用药，开拓了辨证思维，提高古籍方药在慢性肾脏病防治中运用的精准度和灵活度。为防治慢性肾脏病的中药新药开发提供临床应用指导，为慢性肾脏病中医药防治研究开辟新的航道。. 本研究获得软件著作权1项，发表相关SCI论文2篇，EI论文2篇，中文核心论文1篇。参编著作1部。培养青年骨干人才1名，博士研究生1名，硕士研究生2名。