基于国际头痛诊断标准的原发性头痛辅助决策系统

针对原发性头痛在我国发病率较高但诊断准确率较低的问题，提出了一种将文本化的国际头痛诊断标准转换为计算机可执行推理的临床知识建模和知识库构建方法，该方法首先将基于诊断标准绘制的诊断思维流程图转换为规范化的临床知识表达模型，再将临床知识表达模型通过规则映射技术转换为计算机推理诊断所用的规则，形成知识库。并以此为基础开发了覆盖完整头痛诊断流程的原发性头痛辅助决策系统。临床评估显示，该系统可正确识别出91.3%的偏头痛、87.2%的紧张型头痛和90.0%的丛集性头痛病人，对常见的原发性头痛具有较高的诊断准确率。

Assistant decision-making system based on international diagnostic criteria for primary headache disorders

In order to overcome the problem of high morbidity and low diagnostic accuracy of primary headaches in China, this paper proposed a clinical knowledge modeling and knowledge base constrcuting method that translating text-based diagnostic criteria to computer execuable knowledge. This method first translated diagnostic criteria based flow chart of diagnostic thoughts to clinical knowledge representation model, then translated clinical knowledge representation model to reasoning rules by rules-mapping technique. On the basis of the above method, an assistant decision-making system for primary headaches was developed. Clinical evaluation showed that this system can recognize 91.3% of migraine, 87.2% of tension-type headache, 90.0% of cluster headache, and therefore achieve a relatively high success rate for common primary headaches.