Content-based image retrieval of centroblast cells and noncentroblast cells in Follicular lymphoma

In this study content-based image retrieval is applied to hematoxylin and eosin H&E stained Follicular lymphoma centroblast cell images and K-nearest neighbour classifier is used with multi-texton histogram features. With developed method, it is aimed to assist pathologists in their diagnosis of follicular lymphoma disease. The purpose of this project is the classification of centroblast cells and non-centroblast cells with a microscopic content based image retrieval method. Follicular lymphoma database is composed of 218 centroblast and 218 non-centroblast cells. The experiments were conducted by creating 10%-90% and 20%-80% data sets as training and test. The best average accuracies were 93.8% and 86.3% in the 10%-90% and 20%-80% data sets, respectively. In this work, two different feature extraction methods were employed and the classification results were compared with each other. The classification with multi-texton histogram features outperforms the classification accuracy of our group's previous work by 22.4% and 25.1% when 20%-80% and 10%-90% test-training datasets were used, respectively.

[1]  Jamshid Shanbehzadeh,et al.  Image retrieval based on shape similarity by edge orientation autocorrelogram , 2003, Pattern Recognit..

[2]  Hui Kong,et al.  Automated detection of cells from immunohistochemically-stained tissues: application to Ki-67 nuclei staining , 2012, Medical Imaging.

[3]  Hiroyuki Shimada,et al.  A Combined Computerized Classification System for Whole-slide Neuroblastoma Histology: Model-based Structural Features , 2009 .

[4]  N. Otsu.  A threshold selection method from gray level histograms , 1979 .

[5]  Jingyu Yang,et al.  Image retrieval based on the texton co-occurrence matrix , 2008, Pattern Recognit..

[6]  Hui Kong,et al.  Partitioning Histopathological Images: An Integrated Framework for Supervised Color-Texture Segmentation and Cell Splitting , 2011, IEEE Transactions on Medical Imaging.

[7]  Jun Kong,et al.  Computerized microscopic image analysis of follicular lymphoma , 2008, SPIE Medical Imaging.

[8]  Joel H. Saltz,et al.  Histopathological Image Analysis Using Model-Based Intermediate Representations and Color Texture: Follicular Lymphoma Grading , 2009, J. Signal Process. Syst..

[9]  M. Gurcan,et al.  Computer-aided classification of centroblast cells in follicular lymphoma. , 2010, Analytical and quantitative cytology and histology.

[10]  B. Julesz,et al.  Texton gradients: The texton theory revisited , 2004, Biological Cybernetics.

[11]  Lei Zhang,et al.  Contents lists available at ScienceDirect Pattern Recognition , 2022 .

[12]  Jun Kong,et al.  Computer-aided prognosis of neuroblastoma on whole-slide images: Classification of stromal development , 2009, Pattern Recognit..

[13]  Metin Nafi Gürcan,et al.  Content-Based Microscopic Image Retrieval System for Multi-Image Queries , 2012, IEEE Transactions on Information Technology in Biomedicine.