Improving abdomen tumor low-dose CT images using dictionary learning based patch processing and unsharp filtering

Reducing patient radiation dose, while maintaining a high-quality image, is a major challenge in Computed Tomography (CT). The purpose of this work is to improve abdomen tumor low-dose CT (LDCT) image quality by using a two-step strategy: a first patch-wise non linear processing is first applied to suppress the noise and artifacts, that is based on a sparsity prior in term of a learned dictionary, then an unsharp filtering aiming to enhance the contrast of tissues and compensate the contrast loss caused by the DL processing. Preliminary results show that the proposed method is effective in suppressing mottled noise as well as improving tumor detectability.

[1]  Qianjin Feng,et al.  Improving low-dose abdominal CT images by Weighted Intensity Averaging over Large-scale Neighborhoods. , 2011, European journal of radiology.

[2]  D. Donoho,et al.  Maximal Sparsity Representation via l 1 Minimization , 2002 .

[3]  Joseph F. Murray,et al.  Dictionary Learning Algorithms for Sparse Representation , 2003, Neural Computation.

[4]  Emanuele Trucco,et al.  Computer and Robot Vision , 1995 .

[5]  Armando Manduca,et al.  Noise reduction to decrease radiation dose and improve conspicuity of hepatic lesions at contrast-enhanced 80-kV hepatic CT using projection space denoising. , 2012, AJR. American journal of roentgenology.

[6]  Avinash C. Kak,et al.  Principles of computerized tomographic imaging , 2001, Classics in applied mathematics.

[7]  Wufan Chen,et al.  Joint-MAP Tomographic Reconstruction with Patch Similarity Based Mixture Prior Model , 2011, Multiscale Model. Simul..

[8]  Shutao Li,et al.  An Efficient Dictionary Learning Algorithm and Its Application to 3-D Medical Image Denoising , 2012, IEEE Transactions on Biomedical Engineering.

[9]  Jing Wang,et al.  Multiscale Penalized Weighted Least-Squares Sinogram Restoration for Low-Dose X-Ray Computed Tomography , 2008, IEEE Transactions on Biomedical Engineering.

[10]  Terrence J. Sejnowski,et al.  Learning Overcomplete Representations , 2000, Neural Computation.

[11]  David J. Field,et al.  Emergence of simple-cell receptive field properties by learning a sparse code for natural images , 1996, Nature.

[12]  M. Kalra,et al.  Strategies for CT radiation dose optimization. , 2004, Radiology.

[13]  Michael Elad,et al.  Optimally sparse representation in general (nonorthogonal) dictionaries via ℓ1 minimization , 2003, Proceedings of the National Academy of Sciences of the United States of America.

[14]  L. Tanoue.  Computed Tomography — An Increasing Source of Radiation Exposure , 2009 .

[15]  Noriyuki Moriyama,et al.  Improvement of image quality of low radiation dose abdominal CT by increasing contrast enhancement. , 2010, AJR. American journal of roentgenology.

[16]  Michael Elad,et al.  Image Denoising Via Sparse and Redundant Representations Over Learned Dictionaries , 2006, IEEE Transactions on Image Processing.

[17]  Mehran Yazdi,et al.  Artifacts in Spiral X-ray CT Scanners: Problems and Solutions , 2007 .

[18]  Linda G. Shapiro,et al.  Computer and Robot Vision , 1991 .

[19]  Kazuo Awai,et al.  Improvement of Low-Contrast Detectability in Low-Dose Hepatic Multidetector Computed Tomography Using a Novel Adaptive Filter: Evaluation With a Computer-Simulated Liver Including Tumors , 2006, Investigative radiology.

[20]  P. Cochat,et al.  Et al , 2008, Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.