Correcting organ motion artifacts in x-ray CT systems based on tracking of motion phase by the spatial overlap correlator. II. Experimental study

• Published in: Medical physics (Lancaster) 2001-08, Vol.28 (8), p.1577-1596
• Main Authors: Dhanantwari, Amar C., Stergiopoulos, Stergios, Zamboglou, Nikolaos, Baltas, Dimos, Vogt, Hans-Georg, Karangelis, Grigoris
• Format: Article
• Language: English
• Subjects: Algorithms; Anatomy; biological organs; biomechanics; Brachytherapy; Cardiac dynamics; cardiac motion artifacts; cardiology; Coherent effects; Computed radiography; Computed tomography; Computer software; computerised tomography; Data acquisition; Heart; Humans; Image Processing, Computer-Assisted; imaging coronary calcification; Locomotion; Medical image artifacts; medical image processing; Medical imaging; Medical X‐ray imaging; Models, Statistical; Models, Theoretical; motion compensation; Movement; Phantoms, Imaging; Respiration; retrospective gating; Software; spatial overlap correlator; Time Factors; Tomography, X-Ray Computed - methods; X-Rays; x‐ray CT systems; X‐ray imaging
• ISSN: 0094-2405; 2473-4209
• DOI: 10.1118/1.1388893

This paper presents the experimental part of an investigation on tracking and eliminating organ motion artifacts in x-ray CT cardiac applications with emphasis on imaging coronary calcification. The system methodology consists of a software implementation of the spatial overlap correlator (SSOC) concept in x-ray CT scanners to track the net amplitude and phase of organ motion during the CT data acquisition process. A coherent sinogram synthesis (CSS) method is then used to identify the repeated phases of a periodic organ motion from the information provided by the SSOC process and hence synthesize a new sinogram with no motion effects. Since the SSOC scheme is capable of tracking cardiac motion, it identifies also the projection points associated with minimum amplitude cardiac motion effects. These points are used to identify a 180° plus the fan angle sinogram for image reconstruction. This leads to a retrospective gating (RG) scheme that is based on the output of the SSOC process. Performance comparison of the proposed methodology with the retrospective ECG gating using real data sets with phantoms and human patients provides a performance assessment of the merits of the proposed methods. Real results demonstrate that the new methodology eliminates the requirement for ECG gating. Moreover, the CSS and the new RG methods do not require breath holding and they can be implemented in x-ray CT scanners to image coronary calcification and the heart’s ventricles.