患者网上挂号时常有挂错科室的现象,因此需要科室推荐应用,功能类似线下医院的护士台预诊。然而,由于医院科室设置不尽相同,患者各项特征和科室之间的关系也不明确,给自动科室推荐带来挑战。因此,该文首先定义了带权重的知识图谱,用于...患者网上挂号时常有挂错科室的现象,因此需要科室推荐应用,功能类似线下医院的护士台预诊。然而,由于医院科室设置不尽相同,患者各项特征和科室之间的关系也不明确,给自动科室推荐带来挑战。因此,该文首先定义了带权重的知识图谱,用于描述症状、疾病以及性别等特征与科室和医院之间复杂的量化关系。其次,利用区域信息平台的电子健康档案(electronic health records,EHR)数据,获取多家医院的疾病—科室信息。在融合国际疾病编码(international classification of diseases,ICD)、医疗网站中的症状—疾病数据后,用搜索引擎结果补充权重关系,形成可用的知识图谱。图谱目前包含了38家医院,6110个科室,6220个症状,60736个症状相关疾病关系。当患者输入基于自然语言描述的症状与疾病后,通过该文设计的预滤噪的BERT实体识别模型与部位制导的医疗实体归一化算法,识别并归一化患者主诉中的症状词、疾病词和部位词。最后,基于该文设计的基于权重的联合症状预测疾病概率算法(weight-based disease prediction algorithm based on multiple symptoms,WBDPMS),联合多个症状预测可能的相关疾病,以此来实现通过主诉推荐最合适的医院及科室。实验结果表明,准确率达到0.88。展开更多
To improve the performance of the K-shortest paths search in intelligent traffic guidance systems, this paper proposes an optimal search algorithm based on the intelligent optimization search theory and the metaphor m...To improve the performance of the K-shortest paths search in intelligent traffic guidance systems, this paper proposes an optimal search algorithm based on the intelligent optimization search theory and the metaphor mechanism of vertebrate immune systems. This algorithm, applied to the urban traffic network model established by the node-expanding method, can expediently realize K-shortest paths search in the urban traffic guidance systems. Because of the immune memory and global parallel search ability from artificial immune systems, K shortest paths can be found without any repeat, which indicates evidently the superiority of the algorithm to the conventional ones. Not only does it perform a better parallelism, the algorithm also prevents premature phenomenon that often occurs in genetic algorithms. Thus, it is especially suitable for real-time requirement of the traffic guidance system and other engineering optimal applications. A case study verifies the efficiency and the practicability of the algorithm aforementioned.展开更多
文摘患者网上挂号时常有挂错科室的现象,因此需要科室推荐应用,功能类似线下医院的护士台预诊。然而,由于医院科室设置不尽相同,患者各项特征和科室之间的关系也不明确,给自动科室推荐带来挑战。因此,该文首先定义了带权重的知识图谱,用于描述症状、疾病以及性别等特征与科室和医院之间复杂的量化关系。其次,利用区域信息平台的电子健康档案(electronic health records,EHR)数据,获取多家医院的疾病—科室信息。在融合国际疾病编码(international classification of diseases,ICD)、医疗网站中的症状—疾病数据后,用搜索引擎结果补充权重关系,形成可用的知识图谱。图谱目前包含了38家医院,6110个科室,6220个症状,60736个症状相关疾病关系。当患者输入基于自然语言描述的症状与疾病后,通过该文设计的预滤噪的BERT实体识别模型与部位制导的医疗实体归一化算法,识别并归一化患者主诉中的症状词、疾病词和部位词。最后,基于该文设计的基于权重的联合症状预测疾病概率算法(weight-based disease prediction algorithm based on multiple symptoms,WBDPMS),联合多个症状预测可能的相关疾病,以此来实现通过主诉推荐最合适的医院及科室。实验结果表明,准确率达到0.88。
基金This work was supported by the Natural Science Foundation of Shandong Province(No.Y2005G12)National Natural ScienceFoundation of China(No.60674062)and the Information Industry Foundation of Shandong Province(No.2006R00046).
文摘To improve the performance of the K-shortest paths search in intelligent traffic guidance systems, this paper proposes an optimal search algorithm based on the intelligent optimization search theory and the metaphor mechanism of vertebrate immune systems. This algorithm, applied to the urban traffic network model established by the node-expanding method, can expediently realize K-shortest paths search in the urban traffic guidance systems. Because of the immune memory and global parallel search ability from artificial immune systems, K shortest paths can be found without any repeat, which indicates evidently the superiority of the algorithm to the conventional ones. Not only does it perform a better parallelism, the algorithm also prevents premature phenomenon that often occurs in genetic algorithms. Thus, it is especially suitable for real-time requirement of the traffic guidance system and other engineering optimal applications. A case study verifies the efficiency and the practicability of the algorithm aforementioned.
