摘要
乳腺肿瘤的计算机辅助诊断对乳腺肿瘤的诊断和治疗有着重要意义。本文提出一种基于深度置信网络(Deep Belief Network, DBN)的乳腺肿瘤辅助诊断方法。将病人的细胞核图像参数作为深度置信网络的输入,对病人乳腺肿瘤恶性与良性进行判断,并与传统的基于支持向量机、概率神经网络和随机森林模型进行比较。实验结果表明,基于深度置信网络的乳腺肿瘤辅助诊断方法能够进行准确的判断,并且具有较高的准确率。
Computer-aided diagnosis of breast tumors is of great significance in the diagnosis and treatment of breast tumors. This paper presents a method of breast cancer diagnosis based on Deep Belief Network(DBN). Taking the patient’s nuclear image parameters as input to the deep belief network to judge the malignancy and benign of the patient’s breast tumor. Compared with traditional models based on support vector machine, probabilistic neural network and random forest. The experimental results show that the method of breast cancer diagnosis based on depth belief network can make accurate judgment, and has a better accuracy.
作者
吕文豪
雷菊阳
LV Wen-hao;LEI Ju-yang(College of Mechanical and Automotive Engineering, Shanghai University of Enginering Science, Shanghai 201620, China)
出处
《软件》
2019年第6期157-159,共3页
Software
关键词
深度置信网络
支持向量机
概率神经网络
随机森林
乳腺肿瘤辅助诊断
Deep belief network
Support vector machine
Probabilistic neural network
Random forest
Breast tumor assistant diagnosis