DCCL: A Benchmark for Cervical Cytology Analysis

被引：25

作者：

Zhang, Changzheng ^{[1
]}

Liu, Dong ^{[4
]}

Wang, Lanjun ^{[2
]}

Li, Yaoxin ^{[1
]}

Chen, Xiaoshi ^{[1
]}

Luo, Rui ^{[3
]}

Che, Shuanlong ^{[4
]}

Liang, Hehua ^{[4
]}

Li, Yinghua ^{[4
]}

Liu, Si ^{[4
]}

Tu, Dandan ^{[1
]}

Qi, Guojun ^{[3
]}

Luo, Pifu ^{[4
]}

Luo, Jiebo ^{[5
]}

机构：

[1] Huawei, Shenzhen, Peoples R China

[2] Huawei Canada, Markham, ON, Canada

[3] Futurewei, Bellevue, WA USA

[4] KingMed Diagnost Co Ltd, Guangzhou, Peoples R China

[5] Univ Rochester, Rochester, NY 14627 USA

来源：

MACHINE LEARNING IN MEDICAL IMAGING (MLMI 2019) | 2019年 / 11861卷

关键词：

Cervical cancer screening; Liquid-based cytology; Deep learning;

D O I：

10.1007/978-3-030-32692-0_8

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Medical imaging analysis has witnessed impressive progress in recent years thanks to the development of large-scale labeled datasets. However, in many fields, including cervical cytology, a large well-annotated benchmark dataset remains missing. In this paper, we introduce by far the largest cervical cytology dataset, called Deep Cervical Cytological Lesions (referred to as DCCL). DCCL contains 14,432 image patches with around 1,200x2,000 pixels cropped from 1,167 whole slide images collected from four medical centers and scanned by one of the three kinds of digital slide scanners. Besides patch level labels, cell level labels are provided, with 27,972 lesion cells labeled based on The 2014 Bethesda System and the bounding box by six board-certified pathologists with eight years of experience on the average. We also use deep learning models to generate the baseline performance for lesion cell detection and cell type classification on DCCL. We believe this dataset can serve as a valuable resource and platform for researchers to develop new algorithms and pipelines for advanced cervical cancer diagnosis and prevention.

引用

页码：63 / 72

页数：10

共 19 条

[1]

[Anonymous], 2016, CoRR abs/1512.00567, DOI DOI 10.1109/CVPR.2016.308

[2] Pap Smear Image Classification Using Convolutional Neural Network [J].

Bora, Kangkana ;

Chowdhury, Manish ;

Mahanta, Lipi B. ;

Kundu, Malay K. ;

Das, Anup K. .

TENTH INDIAN CONFERENCE ON COMPUTER VISION, GRAPHICS AND IMAGE PROCESSING (ICVGIP 2016), 2016,

[3] Semi-Automatic Segmentation and Classification of Pap Smear Cells [J].

Chen, Yung-Fu ;

Huang, Po-Chi ;

Lin, Ker-Cheng ;

Lin, Hsuan-Hung ;

Wang, Li-En ;

Cheng, Chung-Chuan ;

Chen, Tsung-Po ;

Chan, Yung-Kuan ;

Chiang, John Y. .

IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS, 2014, 18 (01) :94-108

[4] The Pascal Visual Object Classes (VOC) Challenge [J].

Everingham, Mark ;

Van Gool, Luc ;

Williams, Christopher K. I. ;

Winn, John ;

Zisserman, Andrew .

INTERNATIONAL JOURNAL OF COMPUTER VISION, 2010, 88 (02) :303-338

[5]

He K., 2016, CVPR, DOI [10.1109/CVPR.2016.90, DOI 10.1109/CVPR.2016.90]

[6]

Huang G, 2017, PROC CVPR IEEE, P4700, DOI [DOI 10.1109/CVPR.2017.243, 10.1109/CVPR.2017.243]

[7]

Jantzen J., 2005, P NISIS 2005 NAT INS

[8]

Jemal A, 2011, CA-CANCER J CLIN, V61, P134, DOI [10.3322/caac.20115, 10.3322/caac.21492, 10.3322/caac.20107]

[9] Microsoft COCO: Common Objects in Context [J].

Lin, Tsung-Yi ;

Maire, Michael ;

Belongie, Serge ;

Hays, James ;

Perona, Pietro ;

Ramanan, Deva ;

Dollar, Piotr ;

Zitnick, C. Lawrence .

COMPUTER VISION - ECCV 2014, PT V, 2014, 8693 :740-755

[10]

LIN TY, 2017, PROC CVPR IEEE, P936, DOI DOI 10.1109/CVPR.2017.106

← 1 2 →