An iterative tomosynthesis reconstruction using total variation combined with non-local means filtering.

Metin Ertas,Aydin Akan,Mustafa Kamasak,Isa Yildirim

doi:10.1186/1475-925x-13-65

Metin Ertas, Aydin Akan + Show 2 more

Open Access

https://doi.org/10.1186/1475-925x-13-65

Copy DOI

Abstract

BackgroundAfter the release of compressed sensing (CS) theory, reconstruction algorithms from sparse and incomplete data have shown great improvements in diminishing artifacts of missing data. Following this progress, both local and non-local regularization induced iterative reconstructions have been actively used in limited view angle imaging problems.MethodsIn this study, a 3D iterative image reconstruction method (ART + TV)NLM was introduced by combining local total variation (TV) with non-local means (NLM) filter. In the first step, TV minimization was applied to the image obtained by algebraic reconstruction technique (ART) for background noise removal with preserving edges. In the second step, NLM is used in order to suppress the out of focus slice blur which is the most existent image artifact in tomosynthesis imaging. NLM exploits the similar structures to increase the smoothness in the image reconstructed by ART + TV.ResultsA tomosynthesis system and a 3D phantom were designed to perform simulations to show the superior performance of our proposed (ART + TV)NLM over ART and widely used ART + TV methods. Visual inspections show a significant improvement in image quality compared to ART and ART + TV.ConclusionsRMSE, Structure SIMilarity (SSIM) value and SNR of a specific layer of interest (LOI) showed that by proper selection of NLM parameters, significant improvements can be achieved in terms of convergence rate and image quality.

Highlights

After the release of compressed sensing (CS) theory, reconstruction algorithms from sparse and incomplete data have shown great improvements in diminishing artifacts of missing data
Numerous iterative algorithms have been applied to tomographic imaging such as; expectation-maximization (EM) [2], projection onto convex sets (POCS) [3], algebraic reconstruction technique (ART) [2,4,5], simultaneous algebraic reconstruction technique Simultaneous algebraic reconstruction technique (SART) [6]
A widely used sparse image reconstruction algorithm ART + total variation (TV) was modified with Non Local Means (NLM) filter to reduce the out-of-focus slice blur in tomosynthesis system

Summary

Methods

Algebraic Reconstruction Technique (ART) ART is one of the simplest and most commonly used iterative reconstruction techniques [17]. X where Y is the measured data, X^ is the image to be estimated and A is the system matrix which can be considered as the weighting matrix. Weighting parameter is calculated by using the Siddon’s algorithm [18] This algorithm calculates the contribution of voxels to the corresponding radiological path of a ray. This data is used as system matrix in (1). By CS, it was mathematically proven that an image or signal can be recovered from a highly undersampled data. This theory originated a new word “sparsity” for digital information processing. The denoising process is repeated pixel by pixel for the entire image and formulated as:

Results

Conclusions

Background

À ÁÀ Á

TV minimization

NLM filtering