Recent Advance in Exact ROI/VOI Image Reconstruction

Abstract

Over the past few years, computed tomography (CT) reconstruction theory has undergone a rapid development. Since Katsevich published his paper on exact and efficient helical cone-beam CT reconstruction in 2002, intensified research efforts have been made in this field. Various analytic algorithms have been proposed for exact reconstruction from many kinds of projection data. Among them, region-of-interest (ROI) or volume-of-interest (VOI) image reconstruction is one important class of methods in which the projections are usually transversely truncated. During the past three years, some novel algorithms have been developed for exact ROI/VOI reconstruction. Here we present an overview of some key results in this field, especially emphasizing three milestone algorithms: the super-short-scan algorithms, the backprojection-filtration (BPF) or differentiated backprojection (DBP) algorithms, and the POCS (projection onto convex sets) algorithms for interior problem. Finally, we make concluding remarks.

A reweighted total variation minimization method for few view CT reconstruction in the instant CT

“…As shown in Figl, the log-sum form is a better proxy of LO-Norm compared to Ll-Norm and it is a better object function under the condition of the sparseness of the gradient image as the prior knowledge. Furthermore, based on Majorization-Minimization (MM) algorithms, the solution of problem (4) can be given out by (5).…”

“…f(n +l ) = argmin " I V/, I subject to II Hf -p 11< c (5) 7' I VJ;(n) I +e It can be also written (5) in a reweighted TV (RwTV) minimization form as (6). In each iteration, it solves a Ll minimization problem with an update weight.…”

“…S INCE the compressed sensing (CS) theory was introduced into the medical imaging several years ago, CS based, especially total variation (TV) minimization based, iterative CT reconstruction algorithms have been widely studied and developed [1][2][3][4][5]. These algorithms are demonstrated to be quite effective to treat various problems of incomplete data, especially undersampled data.…”

Optimization of a secondary VOI protocol for lung imaging in a clinical CT scanner

“…To maintain SNR at high resolution, it is necessary to raise the photon flux through the targeted volume by concentrating the radiation dose. Volume‐of‐interest CT is an active research topic with innovative approaches in all aspects from system architecture to dynamic beam‐shaping filters to image reconstruction, some of which are summarized by review articles in the field . Some examples of experimental demonstrations include small‐animal studies and implementations on tabletop or clinical c‐arm CBCT systems …”

Iterative Image Reconstruction for Limited-Angle CT Using Optimized Initial Image

“…Computed tomography (CT) has been applied extensively in clinical diagnosis. Under some CT applications, limited-angle scan mode has deserved wide interest [ 1 ] since it has to be used in several practical cases such as large object size [ 2 ], short exposure time [ 3 ], and restricted scanning [ 4 ]. Limited-angle scan means that projection data are obtained in less than 180° angular range when X-ray tube and detector rotate around the scanned object.…”