A new method for image reconstruction in computed tomography (CT) using QR-Decomposition: Image quality assessment

Abstract

This paper describes the results obtained with a new method for medical image reconstruction with computed tomography (CT): QR-Decomposition. QR-Decomposition is a model based (MB) algorithm like maximum likelihood expectation maximization (MLEM) but not iterative. It can be classified as a model-based direct reconstruction (MBDiR) algorithm. The QR-Decomposition algorithm takes advantage of the benefits of the MB approach, but only requires a matrix vector multiplication and backward substitution for image reconstruction. Noise power spectrum (NPS) of three dimensional (3D) images is analyzed and compared using QR-Decomposition standard filtered backprojection (FBP) and maximum likelihood expectation maximization (MLEM). 3D CT reconstructed images show that QR-decomposition process achieves competitive advantages compared to FBP and MLEM images reconstructed with the same voxel size.