Cervical Histopathology Image Clustering Using Graph Based Unsupervised Learning

被引：4

作者：

Li, Chen ^{[1
]}

Hu, Zhijie ^{[1
]}

Chen, Hao ^{[1
]}

Xue, Dan ^{[1
]}

Xu, Ning ^{[2
]}

Zhang, Yong ^{[3
]}

Li, Xiaoyan ^{[3
]}

Wang, Qian ^{[3
]}

Ma, He ^{[1
]}

机构：

[1] Northeastern Univ, MBIE Coll, Microscop Image & Med Image Anal Grp, Shenyang, Peoples R China

[2] Liaoning Shihua Univ, Fushun, Peoples R China

[3] Liaoning Hosp & Inst, Shenyang, Peoples R China

来源：

PROCEEDINGS OF THE 11TH INTERNATIONAL CONFERENCE ON MODELLING, IDENTIFICATION AND CONTROL (ICMIC2019) | 2020年 / 582卷

基金：

中国国家自然科学基金;

关键词：

Cervical cancer; Histopathology image; Clustering; Unsupervised learning; Skeletonization; Graph theory; CANCER STATISTICS; TISSUE;

D O I：

10.1007/978-981-15-0474-7_14

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

In order to apply the important topological information to solve a Cervical Histopathology Image Clustering (CHIC) problem, a Graph Based Unsupervised Learning (GBUL) approach is proposed in this paper. First, the GBUL method applies color features and k-means clustering for a first-stage "coarse" clustering. Then, a Skeletonization Based Node Generation (SBNG) approach is introduced to approximate the distribution of cervical cell nuclei. Thirdly, based on the SBNG nodes, a minimum spanning tree graph is constructed. Next, graph features and additional geometrical features are extracted based on the constructed graph. Finally, the k-means clustering is applied again for the second-stage clustering. In the experiment, a practical cervical histopathology image dataset with ten whole scanned images is tested, obtaining a promising CHIC result and showing a huge potential in the cancer risk prediction field.

引用

页码：141 / 152

页数：12

共 19 条

[1]

Alpaydin E., 2009, INTRO MACHINE LEARNI

[2]

Banu PKN, 2017, INT J MODEL IDENTIF, V27, P92, DOI 10.1504/IJMIC.2017.10003911

[3] CHARACTERIZATION AND FEATURING OF HISTOLOGICAL SECTION IMAGES [J].

CHAUDHURI, BB ;

RODENACKER, K ;

BURGER, G .

PATTERN RECOGNITION LETTERS, 1988, 7 (04) :245-252

[4] A note on the stability and discriminability of graph based features for classification problems in digital pathology [J].

Cruz-Roa, Angel ;

Xu, Jun ;

Madabhushi, Anant .

10TH INTERNATIONAL SYMPOSIUM ON MEDICAL INFORMATION PROCESSING AND ANALYSIS, 2015, 9287

[5]

Gonzalez R.C., 2004, DIGITAL IMAGE PROCES, V624

[6]

Miranda G. H. B., 2012, 2012 XXV SIBGRAPI - Conference on Graphics, Patterns and Images (SIBGRAPI 2012), P316, DOI 10.1109/SIBGRAPI.2012.51

[7] A standard tissue as a control for histochemical and immunohistochemical staining [J].

Otali, D. ;

Fredenburgh, J. ;

Oelschlager, D. K. ;

Grizzle, W. E. .

BIOTECHNIC & HISTOCHEMISTRY, 2016, 91 (05) :309-326

[8] Global cancer statistics, 2002 [J].

Parkin, DM ;

Bray, F ;

Ferlay, J ;

Pisani, P .

CA-A CANCER JOURNAL FOR CLINICIANS, 2005, 55 (02) :74-108

[9]

Ramos-Vara JA, 2011, METHODS MOL BIOL, V691, P83, DOI 10.1007/978-1-60761-849-2_5

[10] Cancer Statistics, 2017 [J].

Siegel, Rebecca L. ;

Miller, Kimberly D. ;

Jemal, Ahmedin .

CA-A CANCER JOURNAL FOR CLINICIANS, 2017, 67 (01) :7-30

← 1 2 →