Comparison of Pleomorphic and Structural Features Used for Breast Cancer Malignancy Classification

Malignancy of a cancer is one of the most important factors that are taken into consideration during breast cancer. Depending on the malignancy grade the appropriate treatment is suggested. In this paper we make use of the Bloom-Richardson grading system, which is widely used by pathologists when grading breast cancer malignancy. Here we discuss the use of two categories of cells features for malignancy classification. The features are divided into polymorphic features that describe nuclei shapes, and structural features that describe cells ability to form groups. Results presented in this work, show that calculated features present a valuable information about cancer malignancy and they can be used for computerized malignancy grading. To support that argument classification error rates are presented that show the influence of the features on classification. In this paper we compared the performance of Support Vector Machines (SVMs) with three other classifiers. The SVMs presented here are able to assign a malignancy grade based on pre-extracted features with accuracy up to 94.24% for pleomorphic features and with an accuracy 91.33% when structural features were used.

[1]  J. Sethian,et al.  An overview of level set methods for etching, deposition, and lithography development , 1997 .

[2]  Chunming Li,et al.  Level set evolution without re-initialization: a new variational formulation , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[3]  Adam Krzyzak,et al.  Classification of Breast Cancer Malignancy Using Cytological Images of Fine Needle Aspiration Biopsies , 2008, Int. J. Appl. Math. Comput. Sci..

[4]  Lasse Riis Østergaard,et al.  Active Surface Approach for Extraction of the Human Cerebral Cortex from MRI , 2006, MICCAI.

[5]  Paul L. Rosin,et al.  A Convexity Measurement for Polygons , 2002, BMVC.

[6]  E. Oja.  Simplified neuron model as a principal component analyzer , 1982, Journal of mathematical biology.

[7]  H. Bloom,et al.  Histological Grading and Prognosis in Breast Cancer , 1957, British Journal of Cancer.

[8]  Jiangwen Deng,et al.  A fast level set method for segmentation of low contrast noisy biomedical images , 2002, Pattern Recognit. Lett..

[9]  Olvi L. Mangasarian,et al.  Nuclear feature extraction for breast tumor diagnosis , 1993, Electronic Imaging.

[10]  J. Sethian,et al.  Fronts propagating with curvature-dependent speed: algorithms based on Hamilton-Jacobi formulations , 1988 .

[11]  Adam Krzyzak,et al.  Fast and Robust Clinical Triple-Region Image Segmentation Using One Level Set Function , 2006, MICCAI.

[12]  Teuvo Kohonen,et al.  The self-organizing map , 1990 .

[13]  S Friedman,et al.  Prognostic value of histologic grade nuclear components of Scarff‐Bloom‐Richardson (SBR). An improved score modification based on a multivariate analysis of 1262 invasive ductal breast carcinomas , 1989, Cancer.

[14]  Dana H. Ballard,et al.  Generalizing the Hough transform to detect arbitrary shapes , 1981, Pattern Recognit..

[15]  Martin Rumpf,et al.  An Adaptive Level Set Method for Medical Image Segmentation , 2001, IPMI.

[16]  David G. Stork,et al.  Pattern Classification , 1973 .

[17]  Andrew P. Bradley,et al.  The use of the area under the ROC curve in the evaluation of machine learning algorithms , 1997, Pattern Recognit..

[18]  W. Eric L. Grimson,et al.  A shape-based approach to the segmentation of medical imagery using level sets , 2003, IEEE Transactions on Medical Imaging.

[19]  David G. Stork,et al.  Pattern Classification (2nd ed.) , 1999 .

[20]  Nello Cristianini,et al.  The Kernel-Adatron Algorithm: A Fast and Simple Learning Procedure for Support Vector Machines , 1998, ICML.