Optimizing lung volume segmentation by texture classification

Accurate and automated Lung Field (LF) segmentation in volumetric computed tomography protocols is highly challenged by the presence of pathologies affecting lung borders, also affecting the performance of computer-aided diagnosis (CAD) schemes. In this work, a three-dimensional LF segmentation algorithm adapted to interstitial lung disease patterns (ILD) patterns is presented. The algorithm employs k-means clustering followed by a filling operation to obtain an initial LF order estimate. The final LF border is obtained by an iterative support vector machine neighborhood labeling of border pixels based on 3D texture features. The proposed method is evaluated on a dataset of 10 cases spanning a range of ILD patterns such as ground glass, reticular, and honeycombing. The accuracy of the method is assessed using area overlap and shape differentiation metrics (dmean, drms, and dmax), by comparing automatically derived lung borders to manually traced ones by a radiologist, and further compared to a Gray Level Thresholding-based (GLT-based) method. The proposed method demonstrated the highest segmentation accuracy, overlap=0.942, dmean=1.835 mm, drms=1.672 mm, and dmax =4.255 mm, which is statistically significant (two-tailed student's t test for paired data, p<0.0001) with respect to all metrics considered as compared to the the GLT-based method overlap=0.836, dmean=2.324 mm, drms=3.890 mm,and dmax=2.946 mm. The proposed segmentation method could be used as an initial stage of a CAD scheme for ILD patterns.

[1]  Lena Costaridou,et al.  Texture classification-based segmentation of lung affected by interstitial pneumonia in high-resolution CT. , 2008, Medical physics.

[2]  Michael F. McNitt-Gray,et al.  Method for segmenting chest CT image data using an anatomical model: preliminary results , 1997, IEEE Transactions on Medical Imaging.

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

[4]  S. Armato,et al.  Automated lung segmentation for thoracic CT impact on computer-aided diagnosis. , 2004, Academic radiology.

[5]  Lena Costaridou,et al.  Towards quantification of interstitial pneumonia patterns in lung multidetector CT , 2008, 2008 8th IEEE International Conference on BioInformatics and BioEngineering.

[6]  B. van Ginneken,et al.  Computer-aided diagnosis in high resolution CT of the lungs. , 2003, Medical physics.

[7]  Soumik Ukil,et al.  Smoothing lung segmentation surfaces in three-dimensional X-ray CT images using anatomic guidance. , 2005, Academic radiology.

[8]  L. Costaridou,et al.  Combining 2D wavelet edge highlighting and 3D thresholding for lung segmentation in thin-slice CT. , 2007, The British journal of radiology.

[9]  Huichuan Duan,et al.  3D computerized segmentation of lung volume with computed tomography. , 2006, Academic radiology.

[10]  Qiang Li,et al.  Automated segmentation of lungs with severe interstitial lung disease in CT. , 2009, Medical physics.

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

[12]  K. Doi,et al.  Quantitative computerized analysis of diffuse lung disease in high-resolution computed tomography. , 2003, Medical physics.

[13]  Richard A. Robb,et al.  High resolution multidetector CT-aided tissue analysis and quantification of lung fibrosis. , 2007 .

[14]  Vladimir Cherkassky,et al.  The Nature Of Statistical Learning Theory , 1997, IEEE Trans. Neural Networks.

[15]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[16]  Bram van Ginneken,et al.  Toward automated segmentation of the pathological lung in CT , 2005, IEEE Transactions on Medical Imaging.

[17]  Lena Costaridou,et al.  Texture-Based Identification and Characterization of Interstitial Pneumonia Patterns in Lung Multidetector CT , 2010, IEEE Transactions on Information Technology in Biomedicine.