An efficient iterative CBCT reconstruction approach using gradient projection sparse reconstruction algorithm.

Heui Chang Lee,Jin Sung Kim,H Harold Li,Sasa Mutic,Bongyong Song,Justin C Park,James J Jung

doi:10.18632/oncotarget.13567

Abstract

The purpose of this study is to develop a fast and convergence proofed CBCT reconstruction framework based on the compressed sensing theory which not only lowers the imaging dose but also is computationally practicable in the busy clinic. We simplified the original mathematical formulation of gradient projection for sparse reconstruction (GPSR) to minimize the number of forward and backward projections for line search processes at each iteration. GPSR based algorithms generally showed improved image quality over the FDK algorithm especially when only a small number of projection data were available. When there were only 40 projections from 360 degree fan beam geometry, the quality of GPSR based algorithms surpassed FDK algorithm within 10 iterations in terms of the mean squared relative error. Our proposed GPSR algorithm converged as fast as the conventional GPSR with a reasonably low computational complexity. The outcomes demonstrate that the proposed GPSR algorithm is attractive for use in real time applications such as on-line IGRT.

Highlights

In recent years, the introduction of cone-beam computed tomography (CBCT) system in radiotherapy procedure has enabled a precise patient positioning prior to the treatment for an on-line targeted radiation delivery [13]
The computational efficiency of gradient projection for sparse reconstruction (GPSR)-Prop largely surpassed that of GPSR-Conv and it is presented in the latter part of this results section
Our proposed GPSR approach involves 1) one forward and backward projection which are the minimum requirement for solving any iterative CBCT reconstruction problems and 2) one extra forward projection to ensure the convergence of the cost function to the global minimum

Summary

Introduction

The introduction of cone-beam computed tomography (CBCT) system in radiotherapy procedure has enabled a precise patient positioning prior to the treatment for an on-line targeted radiation delivery [13]. In order to reduce the imaging dose of CBCT, we need to either 1) minimize the number of x-ray projections based on the anatomy of the patient, and/or 2) reduce the current level of x-ray tube (mAs setting) of each imaging session [5]. The total variation (TV) method has been useful in CT reconstruction by exploiting the small variability in x-ray attenuation across the body www.impactjournals.com/oncotarget tissues [8, 9, 11,12,13,14] This theory has offered a promising solution to the CT reconstruction problems in general as it allows to maximally utilize the projection data through an iterative process. The remaining challenge is in developing an algorithm that handles the computation in an efficient manner while guaranteeing its convergence to the desired 3D image

Objectives

Methods

Results

Discussion

Conclusion