Detecting changes in serial myocardial perfusion SPECT: A simulation study
- 32 Downloads
New algorithms were evaluated for their efficacy in detecting and quantifying serial changes in myocardial perfusion from single photon emission computed tomography (SPECT).
Methods and Results
We generated 72 simulations with various left ventricular positions, sizes, count rates, and perfusion defect severities using the nonuniform rational B-splines (NURBs)-based CArdiac Torso (NCAT) phantom. Images were automatically aligned by use of both full linear and rigid transformations and quantified for perfusion by use of the CEqual program. Changes within a given perfusion defect were compared by use of a Student t test before and after registration. Registration approaches were compared by use of receiver operating characteristic analysis. Changes of 5% were not detected well in single patients with or without alignment. Changes of 10% and 15% could be detected with false-positive rates of 15% and 10%, respectively, in single studies if alignment was performed before perfusion analysis. Alignment also reduced the number of studies necessary to demonstrate a significant perfusion change (P <.05) in groups of patients by about half.
Comparison of mean uptake by t values in SPECT perfusion defects can be used to detect 10% and greater differences in serial perfusion studies of single patients. Image alignment is necessary to optimize automatic detection of perfusion changes in both single patients and groups of patients.
Key WordsSingle photon emission computed tomography t values myocardial perfusion
Unable to display preview. Download preview PDF.
- 2.Neglia D, Sambuceti G, Giorgetti A, Bartoli M, Salvadori P, Sorace O, et al. Effects of long-term treatment with verapamil on left ventricular function and myocardial blood flow in patients with dilated cardiomyopathy without overt heart failure. J Cardiovasc Pharmacol 2000;36:744–50.PubMedCrossRefGoogle Scholar
- 11.Terry JA, Tsui BMW, Perry JR, Hendricks JL, Gullberg GT. The design of a mathematical phantom of the upper human torso for use in 3-D SPECT imaging research. In: Mikulecky DC, Clarke AM, editors. Biomedical engineering: opening new doors. New York: New York University Press; 1990. p. 185–90.Google Scholar
- 13.Faber TL, Garcia EV, Lalush DS, et al. Simulating patient-specific heart shape and motion using SPECT perfusion images with the NCAT phantom. In: Mun SK, editor. Visualization, display, and image guided procedures. Bellingham, Wash: SPIE 4319. 2001. p. 22–6.Google Scholar
- 14.Chen J, Galt JR, Valentine JD, Faber TL, Garcia EV. Modelling SPECT acquisition and processing of changing radiopharmaceutical distribution. Nuclear Science Symposium and Medical Imaging Conference. Piscataway, NJ: IEEE Press. p. 1336–70 1996.Google Scholar
- 16.Neter J, Kutner MH, Nachtsheim CJ, Wasserman W. Applied linear statistical models. Chicago: Richard D. Irwin, Inc; 1996. p. 3–94.Google Scholar
- 19.Slomka PJ, Hurwith GA, Stephenson J, Cradduck T. Automated alignment and sizing of myocardial stress and rest scans to three-dimensional normal templates using an image registration algorithm. J Nucl Med 1996;36:1115–22.Google Scholar
- 22.Faber TL, Galt JR, Chen J, Tsui BMW, Garcia EV. Detecting changes in myocardial perfusion. In: Lemke HU, Vannier MW, Inamura K, Forman AG, Doi K, Reiber JHC, editors. Computerassisted radiology and surgery. Berlin: Springer-Verlag; 2002. p. 879–83.Google Scholar