RT Journal Article
SR Electronic
T1 Relationships between gated SPECT and cardiac magnetic resonance measurement of myocardial wall motion
JF Journal of Nuclear Medicine
JO J Nucl Med
FD Society of Nuclear Medicine
SP 1685
OP 1685
VO 51
IS supplement 2
A1 Nichols, Kenneth
A1 Van Tosh, Andrew
A1 Wang, Yi
A1 Palestro, Christopher
A1 Reichek, Nathaniel
YR 2010
UL http://jnm.snmjournals.org/content/51/supplement_2/1685.abstract
AB 1685 Objectives Gradient surface methods such as “Quantitative Gated SPECT” software (“QGS,” Cedars-Sinai, Los Angeles, CA) report regional left ventricular (LV) myocardial wall motion (WM), based on tracking endocardial borders. This study was conducted to establish relationships between QGS versus cardiac magnetic resonance (CMR) WM measurements. Methods Data were acquired prospectively for 20 pts (age 60±11 years; 95% males) evaluated following myocardial infarction. All pts had 99mTc-MIBI SPECT analyzed by QGS, and gated True-FISP CMR analyzed by Medis “MASS” software. QGS formed segmental polar maps of the number of standard deviations below normal limits of WM, using limits supplied by the manufacturer, and reported the total number of segments with abnormal WM per pt. QGS also computed summed stress scores (SSS), summed rest scores (SRS) and summed difference scores (SDS). For CMR data, an experienced observer manually drew endocardial outlines, from which regional LV WM was computed. Abnormal CMR WM per segment was defined as WM < 2 mm (Radiology 2009; Proceedings of the 2009 Annual Scientific Meeting; P 581).The total number of segments per pt with abnormal CMR WM was determined. Results A wide range of perfusion patterns was present among pts (SSS = 14±8, SRS = 8±7, and SDS= 5±4). 15% of all segments had abnormally low CMR WM. On a segment by segment basis abnormal QGS WM predicted abnormal CMR WM (ROC area = 84±4%), with sensitivity = 88% and specificity = 72%. Total number of segments per pt with abnormal QGS WM correlated significantly with abnormal CMR WM (r=0.78, p=0.0002). Contingency table analysis demonstrated significant association of QGS WM with CMR WM (p<0.0001), with χ2= 111.7, although McNemar’s Δ = 22% (p<0.0001) indicated that QGS overestimated frequency of abnormal WM. Conclusions QGS WM measurements correlated significantly with CMR values, and demonstrated 88% sensitivity in detecting segments with abnormal WM