Abstract

Over the past several years,there have been substantial improvements in the area of three-dimensional (3D) cone-beam Computed Tomography (CT) imaging systems. Nevertheless, more improvement is needed to detect and mitigate motion artifacts in the clinical follow-up of neurological patients with multiple sclerosis, tumors, and stroke, etc., in which failure to detect motion artifacts often leads to misdiagnosis of disease. In this paper, we propose a marker-based innovative approach to detect and mitigate motion artifacts in 3D cone-beam brain CT systems without using any external motion tracking sensors. Motion is detected by comparing the motion-free ideal marker projections and the corresponding measured marker projections. Once motion is detected, motion parameters (six degrees-of-freedom of motion) are estimated using a numerical optimization technique. Artifacts, caused by motions, are mitigated in the back projection stage of the 3D reconstruction process by correcting the position of every reconstruction voxel according to the estimated motion parameters. We refer to this algorithm as the MB_FDK (Marker-based Feldkemp–Davis–Kress) algorithm. MB_FDK has been evaluated on a modified 3D Shepp–Logan phantom with a range of simulated motion. Simulation results demonstrate a quantitative and qualitative validation of motion detection and artifact mitigation techniques.

Abstract