期刊
PATTERN RECOGNITION LETTERS
卷 125, 期 -, 页码 1-6出版社
ELSEVIER
DOI: 10.1016/j.patrec.2019.03.022
关键词
Deep learning; Smart pattern recognition; Transfer learning; Breast cancer
资金
- FCT - Fundacao para a Ciencia e a Tecnologia [UID/EEA/50 0 08/2019]
- MCTIC under the Centro de Referencia em Radiocomunicacoes - CRR project of the Instituto Nacional de Telecomunicacoes (Inatel), Brazil [01250.075413/2018-04]
- Brazilian National Council for Research and Development (CNPq) [309335/2017-5]
- RNP
Breast cancer is among the leading cause of mortality among women in developing as well as under-developing countries. The detection and classification of breast cancer in the early stages of its development may allow patients to have proper treatment. In this article, we proposed a novel deep learning framework for the detection and classification of breast cancer in breast cytology images using the concept of transfer learning. In general, deep learning architectures are modeled to be problem specific and is performed in isolation. Contrary to classical learning paradigms, which develop and yield in isolation, transfer learning is aimed to utilize the gained knowledge during the solution of one problem into another related problem. In the proposed framework, features from images are extracted using pre-trained CNN architectures, namely, GoogLeNet, Visual Geometry Group Network (VGGNet) and Residual Networks (ResNet), which are fed into a fully connected layer for classification of malignant and benign cells using average pooling classification. To evaluate the performance of the proposed framework, experiments are performed on standard benchmark data sets. It has been observed that the proposed framework outclass all the other deep learning architectures in terms of accuracy in detection and classification of breast tumor in cytology images. (C) 2019 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据