Locality Sensitive Deep Learning for Detection and Classification of Nuclei in Routine Colon Cancer Histology Images

被引:827
作者
Sirinukunwattana, Korsuk [1 ]
Raza, Shan E. Ahmed [1 ]
Tsang, Yee-Wah [2 ]
Snead, David R. J. [2 ]
Cree, Ian A. [2 ]
Rajpoot, Nasir M. [3 ,4 ]
机构
[1] Univ Warwick, Dept Comp Sci, Coventry CV4 7AL, W Midlands, England
[2] Univ Hosp Coventry & Warwickshire, Dept Pathol, Coventry CV2 2DX, W Midlands, England
[3] Qatar Univ, Dept Comp Sci & Engn, Doha, Qatar
[4] Univ Warwick, Dept Comp Sci, Coventry CV4 7AL, W Midlands, England
关键词
Convolutional neural network; deep learning; histology image analysis; nucleus detection; HISTOPATHOLOGY; HETEROGENEITY;
D O I
10.1109/TMI.2016.2525803
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Detection and classification of cell nuclei in histopathology images of cancerous tissue stained with the standard hematoxylin and eosin stain is a challenging task due to cellular heterogeneity. Deep learning approaches have been shown to produce encouraging results on histopathology images in various studies. In this paper, we propose a Spatially Constrained Convolutional Neural Network (SC-CNN) to perform nucleus detection. SC-CNN regresses the likelihood of a pixel being the center of a nucleus, where high probability values are spatially constrained to locate in the vicinity of the centers of nuclei. For classification of nuclei, we propose a novel Neighboring Ensemble Predictor (NEP) coupled with CNN to more accurately predict the class label of detected cell nuclei. The proposed approaches for detection and classification do not require segmentation of nuclei. We have evaluated them on a large dataset of colorectal adenocarcinoma images, consisting of more than 20,000 annotated nuclei belonging to four different classes. Our results show that the joint detection and classification of the proposed SC-CNN and NEP produces the highest average F1 score as compared to other recently published approaches. Prospectively, the proposed methods could offer benefit to pathology practice in terms of quantitative analysis of tissue constituents in whole-slide images, and potentially lead to a better understanding of cancer.
引用
收藏
页码:1196 / 1206
页数:11
相关论文
共 37 条
[1]   Improved Automatic Detection and Segmentation of Cell Nuclei in Histopathology Images [J].
Al-Kofahi, Yousef ;
Lassoued, Wiem ;
Lee, William ;
Roysam, Badrinath .
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, 2010, 57 (04) :841-852
[2]  
Cruz-Roa AA, 2013, LECT NOTES COMPUT SC, V8150, P403, DOI 10.1007/978-3-642-40763-5_50
[3]   An Integrated Region-, Boundary-, Shape-Based Active Contour for Multiple Object Overlap Resolution in Histological Imagery [J].
Ali, Sahirzeeshan ;
Madabhushi, Anant .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 2012, 31 (07) :1448-1460
[4]  
[Anonymous], 2011, JPI
[5]  
[Anonymous], 2011, J PATHOL INFORM
[6]  
Arif M, 2007, INTERNATIONAL CONFERENCE ON MACHINE VISION 2007, PROCEEDINGS, P113
[7]  
Arteta C, 2012, LECT NOTES COMPUT SC, V7510, P348, DOI 10.1007/978-3-642-33415-3_43
[8]  
Bishop C., 2006, Pattern recognition and machine learning, P423
[9]  
Ciresan D., 2012, Advances in Neural Information Processing Systems 25 (NIPS 2012), P1
[10]   Mitosis Detection in Breast Cancer Histology Images with Deep Neural Networks [J].
Ciresan, Dan C. ;
Giusti, Alessandro ;
Gambardella, Luca M. ;
Schmidhuber, Juergen .
MEDICAL IMAGE COMPUTING AND COMPUTER-ASSISTED INTERVENTION - MICCAI 2013, PT II, 2013, 8150 :411-418