摘要
乳腺癌亚型的分类对乳腺癌的诊断和预后起着关键性的作用。针对目前还缺乏识别乳腺癌病理图像的深度学习模型,提出一种基于改进的深度卷积神经网络的乳腺癌病理图像识别模型。通过构建小尺寸病理图像数据集,训练深度卷积神经网络,达到对乳腺癌分子亚型进行智能识别的目的。将该模型与其他经典的深度学习模型进行对比,在乳腺癌分子亚型的分类工作上,该模型有更加好的识别性能。
The classification of breast cancer subtypes plays a key role in the diagnosis and prognosis of breast cancer.In view of the lack of deep learning model for breast cancer pathological image recognition,an improved deep convolution neural network model for breast cancer path⁃ological image recognition is proposed.By constructing small-scale pathological image data set and training deep convolution neural net⁃work,the purpose of intelligent recognition of breast cancer molecular subtypes is achieved.Compared with other classical deep learning models,this model has better recognition performance in the classification of breast cancer molecular subtypes.
作者
黄军豪
廖天驰
HUANG Jun-hao;LIAO Tian-chi(College of Information Engineering,Sichuan Agricultural University,Ya'an 625014;College of Science,Sichuan Agricultural University,Ya'an 625014)
出处
《现代计算机》
2020年第22期3-8,共6页
Modern Computer
基金
四川农业大学创新训练计划项目(No.201910626061)。
关键词
乳腺癌亚型
分类
卷积神经网络
深度学习
Timer Breast Cancer Subtypes
Classification
Convolutional Neural Network
Deep Learning
