Paper
18 March 2008 Landmark based compensation of patient motion artifacts in computed tomography
Yves Pauchard, Steven K. Boyd
Author Affiliations +
Abstract
We propose a new method for compensating patient motion in computed tomography. The method consists of analyzing the frequency spectrum of tracked landmark points in the acquired sinogram. These landmarks are either physically attached to the patient prior scanning, or coincide with anatomical points traceable in the sinogram. Without motion present, these extracted landmark curves represent one period of a sine wave in full-rotation tomography. Motion compensation is achieved by calculating the fundamental frequency component of the extracted landmark curves in order to obtain the motion compensated landmark curve. The extracted and compensated landmark curve pairs serve for constructing a motion map, and these are utilized to re-sort the acquired sinogram to obtain a motion compensated sinogram. Reconstructing the motion compensated sinogram using a standard reconstruction algorithm yields the motion compensated image. The proposed method is compared to a previously published motion artifact reduction method. The results show equal or improved motion compensation capabilities of the proposed method for three different types of computer simulated in-plane motion in synthetic 2D parallel-beam sinograms. The influence of inaccurate curve extraction has been simulated and results are included. Experiments aimed at demonstrating the compensation of motion artifacts in micro-computed tomography patient images are currently under way. In addition to compensation, future work will explore detection and quantification of patient motion based on frequency spectrum analysis of landmark curves, potentially providing a comprehensive tool for identifying, quantifying and correcting motion artifacts.
© (2008) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Yves Pauchard and Steven K. Boyd "Landmark based compensation of patient motion artifacts in computed tomography", Proc. SPIE 6913, Medical Imaging 2008: Physics of Medical Imaging, 69133C (18 March 2008); https://doi.org/10.1117/12.768434
Lens.org Logo
CITATIONS
Cited by 5 scholarly publications.
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
3D image processing

Computed tomography

Image segmentation

Computer simulations

Motion models

Bone

Image processing

Back to Top