The anisotropic Gaussian kernel for SVM classification of HRCT images of the lung

High-resolution computed tomography (HRCT) produces lung images with a high level of detail which makes it suitable for diagnosis of diffuse lung diseases. Segmentation of abnormal lung patterns is a necessary stage in the construction of a computer-aided diagnosis system. We interpret lung patterns as textures and develop a texture classification technique for segmentation of lung patterns. The wavelet transform is used to extract texture features and then the support vector machines (SVM) machine learning algorithm is applied to texture classification. The parameters of the SVM play a crucial role in the performance of the algorithm. We apply gradient-based optimization of the radius/margin bound of a generalization error to tune parameters of the SVM algorithm. We compare the performance of isotropic and anisotropic Gaussian kernels and study the applicability of the radius/margin bound to tuning parameters of the SVM algorithm on the problem of lung pattern classification.

[1]  Michael Unser,et al.  Texture classification and segmentation using wavelet frames , 1995, IEEE Trans. Image Process..

[2]  Anthony Widjaja,et al.  Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond , 2003, IEEE Transactions on Neural Networks.

[3]  C.-C. Jay Kuo,et al.  Texture analysis and classification with tree-structured wavelet transform , 1993, IEEE Trans. Image Process..

[4]  Ingrid Daubechies,et al.  Ten Lectures on Wavelets , 1992 .

[5]  Bernhard Schölkopf,et al.  The connection between regularization operators and support vector kernels , 1998, Neural Networks.

[6]  Stéphane Mallat,et al.  A Theory for Multiresolution Signal Decomposition: The Wavelet Representation , 1989, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  Tomaso A. Poggio,et al.  Regularization Networks and Support Vector Machines , 2000, Adv. Comput. Math..

[8]  E. Hoffman,et al.  Computer recognition of regional lung disease patterns. , 1999, American journal of respiratory and critical care medicine.

[9]  Kunio Doi,et al.  Automatic detection of abnormalities in chest radiographs using local texture analysis , 2002, IEEE Transactions on Medical Imaging.

[10]  S. Sathiya Keerthi,et al.  Efficient tuning of SVM hyperparameters using radius/margin bound and iterative algorithms , 2002, IEEE Trans. Neural Networks.

[11]  Paul Scheunders,et al.  Wavelets for texture analysis, an overview , 1997 .

[12]  Jelena Kovacevic,et al.  Wavelet families of increasing order in arbitrary dimensions , 2000, IEEE Trans. Image Process..

[13]  Sayan Mukherjee,et al.  Choosing Multiple Parameters for Support Vector Machines , 2002, Machine Learning.

[14]  M. McNitt-Gray,et al.  The effects of co-occurrence matrix based texture parameters on the classification of solitary pulmonary nodules imaged on computed tomography. , 1999, Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society.

[15]  Ioannis Pitas,et al.  Segmentation of ultrasonic images using Support Vector Machines , 2003, Pattern Recognit. Lett..

[16]  S. Sathiya Keerthi,et al.  Evaluation of simple performance measures for tuning SVM hyperparameters , 2003, Neurocomputing.

[17]  A F Laine,et al.  Wavelets in temporal and spatial processing of biomedical images. , 2000, Annual review of biomedical engineering.

[18]  D K Smith,et al.  Numerical Optimization , 2001, J. Oper. Res. Soc..

[19]  Nello Cristianini,et al.  Dynamically Adapting Kernels in Support Vector Machines , 1998, NIPS.

[20]  Miodrag Popović,et al.  Characterization of visually similar diffuse diseases from B-scan liver images using nonseparable wavelet transform , 1998, IEEE Transactions on Medical Imaging.

[21]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .