Architectural distortion detection from mammograms using support vector machine

One of the leading diseases in women is breast cancer. The detection in an earlier stage is done by indicating the presence of architectural distortion (AD). An AD detection system with support vector machine is developed in this research. The 15 features are extracted from the fuzzy co-occurrence matrix and fractal dimension. The principal component analysis is also implemented to help in feature redundancy reduction. We found out that the best system for the training data set yields 91.67 % correct AD classification with 0.93 sensitivity of detecting AD and 0.91 specificity of detecting true negative. The best result of the blind test mammograms is at 100.00 % correct AD classification with approximately 16 false positive areas per image.

[1]  Benoit B. Mandelbrot,et al.  Fractal Geometry of Nature , 1984 .

[2]  K. Revathy,et al.  A New Approach for Mammogram Image Classification Using Fractal Properties , 2012 .

[3]  Julia E. E. de Oliveira,et al.  Model based approach for Detection of Architectural Distortions and Spiculated Masses in Mammograms , 2011 .

[4]  A. Retico,et al.  Mammogram Segmentation by Contour Searching and Mass Lesions Classification With Neural Network , 2004, IEEE Transactions on Nuclear Science.

[5]  R Nithya,et al.  Classification of Normal and Abnormal Patterns in Digital Mammograms for Diagnosis of Breast Cancer , 2011 .

[6]  N. Theera-Umpon,et al.  Fractal dimension estimation using modified differential box-counting and its application to mstar target classification , 2002, IEEE International Conference on Systems, Man and Cybernetics.

[7]  Franck Multon,et al.  Fall Detection With Multiple Cameras: An Occlusion-Resistant Method Based on 3-D Silhouette Vertical Distribution , 2011, IEEE Transactions on Information Technology in Biomedicine.

[8]  Sansanee Auephanwiriyakul,et al.  Breast Abnormality Detection in Mammograms Using Fuzzy Inference System , 2005, The 14th IEEE International Conference on Fuzzy Systems, 2005. FUZZ '05..

[9]  D. Sen,et al.  Image Segmentation using Global and Local Fuzzy Statistics , 2006, 2006 Annual IEEE India Conference.

[10]  Nader Karimi,et al.  Automatic Monocular System for Human Fall Detection Based on Variations in Silhouette Area , 2013, IEEE Transactions on Biomedical Engineering.

[11]  Anuradha C. Phadke,et al.  Classification of Architectural Distortion from Other Abnormalities in Mammograms , 2013 .

[12]  C. Sparrow.  The Fractal Geometry of Nature , 1984 .

[13]  Anselmo Cardoso de Paiva,et al.  Detection of Masses in Digital Mammograms using K-Means and Support Vector Machine , 2009 .

[14]  Diego Garcia-Alvarez,et al.  Fault Detection Using Principal Component Analysis ( Pca ) in a Wastewater Treatment Plant ( Wwtp ) , 2009 .

[15]  C. V. Jawahar,et al.  Fuzzy statistics of digital images , 1996, IEEE Signal Processing Letters.

[16]  Kun-Ching Wang,et al.  Based on Fuzzy Linear Discriminant Analysis for Breast Cancer Mammography Analysis , 2011, 2011 International Conference on Technologies and Applications of Artificial Intelligence.

[18]  Nirupam Sarkar,et al.  An Efficient Differential Box-Counting Approach to Compute Fractal Dimension of Image , 1994, IEEE Trans. Syst. Man Cybern. Syst..

[19]  Alan C. Bovik,et al.  Evidence based detection of spiculated masses and architectural distortions , 2005, SPIE Medical Imaging.

[20]  Rangaraj M. Rangayyan,et al.  Detection of Architectural Distortion in Prior Mammograms , 2011, IEEE Transactions on Medical Imaging.

[21]  Hiroshi Fujita,et al.  Automated detection method for architectural distortion areas on mammograms based on morphological processing and surface analysis , 2004, SPIE Medical Imaging.

[22]  Sonja Grgic,et al.  Shape analysis and classification of masses in mammographic images using neural networks , 2011, 2011 18th International Conference on Systems, Signals and Image Processing.

[23]  Robert M. Haralick,et al.  Textural Features for Image Classification , 1973, IEEE Trans. Syst. Man Cybern..