A comparison of left ventricular volumes and ejection fraction by various commercially-available nuclear cardiology software on a CZT SPECT camera [abstract] Abstract uri icon

abstract

  • Introduction: Nuclear cardiology is known to provide accurate left ventricular ejection fraction (LVEF), left ventricular end-diastolic volume (LVEDV), and left ventricular end-systolic volume (LVESV) calculations. It is less established how much variation exists between the different software that are used for LVEF, LVEDV, and LVESV calculation. Methods: 290 patients having a nuclear cardiology stress test between 1/12/15 and 3/1/15 were identified. LVEDV, LVESV, and LVEF data from 4 nuclear cardiology software programs (Myometrics; 4DM; EC Tool Box; Cedars) were compared using a correlated analysis of variance (MANOVA). 141 of the patients had an echocardiogram within one-year of their nuclear study, which allowed for a comparison with the 4 nuclear cardiology LVEF calculations. Results: Of the 290 patients, the average age was 63 ± 11 years, and 147 (51%) were female. 100 (34%) patients had a history of coronary artery disease, with 69 (24%) having a previous percutaneous coronary intervention, and 20 (7%) having a previous coronary artery bypass surgery. 131 (45%) patients performed exercise stress testing, with the remaining having pharmacologic stress studies. There were 79 (27%) patients that had an abnormal study with infarction or ischemia identified. The LVEF, LVEDV, and LVESV calculations differed significantly (MANOVA\0.001) between the 4 nuclear cardiology software programs. All 4 had statistically significant linear correlations with one another (r [ 0.87; P\0.001). Pair-wise comparison found Myometrics and Cedars to have an insignificant difference in LVEF (p-value = 0.31) and LVEDV (P value = 0.15), while 4DM and Cedars had an insignificant difference in LVESV (P value = 0.84). Compared to the LVEF estimated by echocardiography all 4 nuclear cardiology software’s LVEF calculations were significantly higher. In a subgroup of 44 (14%) patients with at least one method suggesting reduced LVEF (\50%), only echo and 4DM had insignificant differences in LVEF. Myometrics, 4DM, Cedars, and echocardiography had respective LVEF of: 43 ± 13; 46 ± 13; 40 ± 11; and 46 ± 14, while EC Tool Box reported an LVEF of 55 ± 13. Conclusion: Nuclear cardiology software yield statistically significant different calculations of LVEF, LVEDV, and LVESV. EC Tool Box calculated a significantly higher LVEF that in the setting of reduced LVEF may be of clinical significance.

publication date

  • 2017