摘要
为了准确对肿瘤病理类型进行诊断,减轻患者因诊断造成的不便与痛楚,提出了基于变精度粗糙集理论与贝叶斯网络的肿瘤诊断方法。对930例甲状腺肿瘤病例使用变精度粗糙集与贝叶斯网络方法导出诊断规则,并与粗糙集对比后发现,基于变精度粗糙集与贝叶斯网络的诊断规则准确性高于粗糙集理论的诊断准确性。基于变精度粗糙集与贝叶斯网络的肿瘤诊断模型对提高肿瘤诊断水平具有更好的临床使用价值,为肿瘤的诊断与分类提供了一种新思路。
To accurately diagnose the type of tumor pathology,reduce patients＇ inconvenience and pain caused by diagnosis,a tumor diagnosis method is proposed based on variable rough set theory and Bayesian network.For the 930 cases of thyroid cancer using variable precision rough set and Bayesian network approach to export diagnostic rules,and in comparison with the rough set,it concludes that diagnosis rule accuracy based on variable rough set theory and Bayesian network is higher than that of rough set.Cancer diagnosis model based on VPRS and Bayesian network has better clinical value in improving cancer diagnosis level,which provides a new idea for tumor diagnosis and classification.
出处
《统计与信息论坛》
CSSCI
北大核心
2015年第4期71-77,共7页
Journal of Statistics and Information
基金
国家自然科学基金项目<单病种质量管理体系的数字化模型研究>(71363063)
关键词
肿瘤病理类型
肿瘤诊断
粗糙集
贝叶斯网络
知识简约
tumor pathological type
tumor diagnosis
variable precision rough set
Bayesian network
knowledge reduction
