应用数据挖掘技术获取周围型肺癌临床和CT诊断规则的初步研究

目的探讨数据挖掘技术对周围型肺癌影像诊断规则提取的价值。方法收集58例经过临床病理证实的周围型肺癌病例,对其临床及CT表现属性进行标准化认定,输入数据库,分别采用自主开发的基于关联规则知识发现程序与通用数据分析工具ROSETTA中的粗糙集约简算法和遗传分类算法对58例周围型肺癌临床及影像学数据进行挖掘对比研究。结果由Johnson’s Algorithm粗糙集约简算法产生诊断规则51条,由ROSETTA遗传分类算法所产生的诊断规则有5千多条,基于关联规则的挖掘算法所产生的诊断规则有123条。这3种不同的数据挖掘方法产生的最重要的诊断规则基本上都将性别、年龄、位置、毛刺、形状、毛玻璃样密度等属性作为诊断周围型肺癌的主要依据。结论数据挖掘技术在医学影像诊断和鉴别诊断中具有潜在的应用价值。

Preliminary Study on Clinical and CT Diagnostic Rule of Peripheral Lung Cancer Based on Data Mining Technique

Objective To discuss the extraction value about the imaging diagnostic rules of peripheral lung cancer by using data mining technique.Methods 58 cases of peripheral lung cancer confirmed by clinical pathology were collected,the data were imported into the database after the standardization of the clinical and CT findings attributes.The data were studied comparatively based on the knowledge discovery process in association with the rough set reduction algorithm and genetic algorithm of the generic data analysis tool(ROSETTA),respectively.Results The diagnostic rules generated by the rough set reduction algorithm of Johnson's Algorithm,the genetic classification algorithm of ROSETTA and the mining algorithm generates were 51,over 5000 and 123 respectively.The main items for the diagnosis of peripheral lung cancer generated by three data mining methods basically were gender and age of patients,the location,burr sign,shape and ground-glass density of lesions.Conclusion The data mining technology in medical imaging diagnosis and differential diagnosis is of the potential value.