logo

A quantitative assessment of heart phantom motion and its effect on myocardial perfusion SPECT images

A quantitative assessment of heart phantom motion and its effect on myocardial perfusion SPECT images

SHI Hong-Cheng
CHEN Shao-Liang
Zhang Xin
YAO Zhi-Feng
LIU Wen-Guan
ZHU Wei-Min
Nuclear Science and TechniquesVol.13, No.2pp.84-91Published in print 01 May 2002
22200

In order to study the image characteristics of motion artifacts and to determine the relations of motion artifacts with varied motion types, and the imaging timings, frames, distances and directions during SPECT acquisition, a myocardial phantom filled with pertechnetate solution was used to simulate the patient motion. In nonreturning pattern, the simulation motion was timed at the 0°, -45° and -90° positions during the rotation of the detector over a 180° are from +45° right anterior oblique to -135° left posterior oblique. Simulation motion was performed by moving the phantom ±5mm, ±10mm and ±20mm along X- (from left to right), Y- (from head to caudal) and Z-axis (from back to ventral) respectively. In returning pattern the acquired 30 projections were divided into three equal parts. The simulation motion was timed at the middle 1-7 projections of each part and performed by moving the phantom ±5, ±10, ±15, ±20, ±25, ±30 and ±50 mm along X-, Y- and Z-axis respectively. Each image was compared with normal image and assessed by three experienced observers without knowledge of the phantom motion. Logistic regression analysis was used to determine the relationship of motion artifacts with the affecting factors. No significant artifacts can be found when the phantom was moved slightly, no matter which motion pattern, direction and timing were taken. The characteristics of motion artifacts showed a radioactive marker dot in inferior wall firstly when the phantom was moved along X-axis. Septal and lateral wall became "hot" symmetrically when the phantom was moved along Y-axis. And nodular hot could be found in anterior wall when the phantom was moved along Z-axis. At last the "lumpy" and "defect" areas existed alternately and formed a triangle respectively. The presence of motion artifacts was related to motion directions, distance and affected frames, but was independent of motion timing. The characteristics of motion artifacts could be found when the phantom was moved along different axis. Motion distance contributed more to the appearance of motion artifacts than other related factors, this was especially clear when motion was along Y-axis.

Heart phantomMyocardial perfusion SPECTMotion artifacts
References
1 Botvinick E H, Zhu Y Y, O'Connell W J et al. J Nucl Med, 1993, 34(2): 303-310
2 O'Connor M K, Kanal K M, Gebhard M W et al. J Nucl Med, 1998, 39(10): 2027-2034
3 Shi H C, Li W G, Chen S L et al. Chin J Nucl Med, 2000, 20(3): 109-111
4 DePuey E G, Garcia E V. J Nucl Med, 1989, 30(4): 441-449
5 Eisner R L, Noever T, Nowak D et al. J Nucl Med, 1987, 28(1): 97-101
6 Matsumoto N, Berman D S, Kavanagh P B et al. J Nucl Med, 2001, 42(5): 687-694
7 Eisner R, Churchwell A, Nover T et al. J Nucl Med, 1988, 29(1): 91-97
8 Friedman J, Train K V, Maddahi J et al. J Nucl Med, 1989, 30(10): 1718-1722