摘要
医学影像是癌症早期检测的一个重要方式.现提出一种基于更快速区域卷积神经网络(FasterRCNN)的癌症影像检测方法,通过添加拉普拉斯卷积层锐化结节边缘、突出结节轮廓,修改锚框大小,调试超参数提高了检测效果.该方法在保证精度的情况下,提高了诊断效率,为计算机辅助诊断提供了参考.
Medical imaging is an important method to detect cancers at an early stage. A cancer image detection method based on the Faster region convolution neural network （Faster RCNN） was proposed. The Laplace convolution layer was used to sharpen the edge of nodules, highlight the nodule profile, modify the size of the anchor, and debug the hyperparameter. This method improves the diagnostic efficiency and accuracy, which provides a referenc method for computer aided diagnosis.
作者
王黎
陆慧娟
叶敏超
严珂
WANG Li;LU Huijuan;YE Minchao;YAN Ke(College of Information Engineering,China Jiliang University,Hangzhou 310018,China)
出处
《中国计量大学学报》
2018年第2期136-141,共6页
Journal of China University of Metrology
基金
国家自然科学基金资助项目(No.61272315)
浙江省科技计划项目(No.2017C34003)
