Blind image restoration by anisotropic regularization
IEEE Transactions on Image Processing
IEEE Transactions on Image Processing
| Hi-index | 0.00 |
Ultrasound B-scan exhibits shadowing and enhancement artifacts due to acoustic wave propagation and spatially varying scatter attenuation across layers of tissues. These artifacts hide underlying echo signals that are truly clinically indicative of diseases. Attenuation compensation estimates and corrects for shadowing and enhancement artifacts, which improves the quality of ultrasound imaging. Block-based attenuation compensation methods, widely employed in commercial scanners, produce results with resolutions limited by the block size. To obtain higher spatial resolution (as desired for quantitative analysis), we present a backscatter-contour-attenuation (BCA) joint estimation model for attenuation compensation in pulse-echo imaging using a set of self-consistent partial differential equations and a contour evolution model. The problem is posed as reconstructing sources of information from observations. We derive the joint estimation model from minimizing a cost functional of separated attributes with region-based isotropic regularizations. A three-step alternating minimization method is adopted towards a tractable numerical solution. Detailed numerical methods are described. The efficacy of the proposed approach is demonstrated using simulated and real images.