Original Article
Optimal reproducibility of gated sestamibi and thallium myocardial perfusion study left ventricular ejection fractions obtained on a solid-state CZT cardiac camera requires operator input

https://doi.org/10.1007/s12350-012-9561-6Get rights and content

Abstract

Aim

To evaluate the reproducibility of serial re-acquisitions of gated Tl-201 and Tc-99m sestamibi left ventricular ejection fraction (LVEF) measurements obtained on a new generation solid-state cardiac camera system during myocardial perfusion imaging and the importance of manual operator optimization of left ventricular wall tracking.

Methods

Resting blinded automated (auto) and manual operator optimized (opt) LVEF measurements were measured using ECT toolbox (ECT) and Cedars-Sinai QGS software in two separate cohorts of 55 Tc-99m sestamibi (MIBI) and 50 thallium (Tl-201) myocardial perfusion studies (MPS) acquired in both supine and prone positions on a cadmium zinc telluride (CZT) solid-state camera system. Resting supine and prone automated LVEF measurements were similarly obtained in a further separate cohort of 52 gated cardiac blood pool scans (GCBPS) for validation of methodology and comparison. Appropriate use of Bland-Altman, chi-squared and Levene’s equality of variance tests was used to analyse the resultant data comparisons.

Results

For all radiotracer and software combinations, manual checking and optimization of valve planes (+/− centre radius with ECT software) resulted in significant improvement in MPS LVEF reproducibility that approached that of planar GCBPS. No difference was demonstrated between optimized MIBI/Tl-201 QGS and planar GCBPS LVEF reproducibility (P = .17 and P = .48, respectively). ECT required significantly more manual optimization compared to QGS software in both supine and prone positions independent of radiotracer used (P < .02).

Conclusions

Reproducibility of gated sestamibi and Tl-201 LVEF measurements obtained during myocardial perfusion imaging with ECT toolbox or QGS software packages using a new generation solid-state cardiac camera with improved image quality approaches that of planar GCBPS however requires visual quality control and operator optimization of left ventricular wall tracking for best results. Using this superior cardiac technology, Tl-201 reproducibility also appears at least equivalent to sestamibi for measuring LVEF.

Introduction

The availability of solid-state cardiac detector systems for thallium and sestamibi myocardial perfusion imaging is increasing; however, there is little published data evaluating the reproducibility of serial re-acquisitions of left ventricular ejection fraction (LVEF) measurements on such systems. Prior studies have demonstrated supine or prone patient positioning has no effect on LVEF measurements obtained on equilibrium gated cardiac blood pool scanning (GCBPS) or gated myocardial perfusion studies (MPS) performed on a standard gamma camera, with results in either postures considered interchangeable.1,2 Thus, serial acquisition of prone and supine MPS or GCBPS studies can be used to measure the reproducibility of LVEF measurement for these and any other methods used to measure this important parameter.

The aim of this prospective study is to evaluate the reproducibility of serial re-acquisitions of gated thallium and sestamibi LVEF measurements obtained on a solid-state cardiac camera system obtained during myocardial perfusion imaging and to determine the importance of manual operator optimization of left ventricular wall tracking. Results were then compared with the same protocol using planar GCBPS LVEF whose day-to-day reproducibility is well validated.

Section snippets

Solid-State Cardiac CZT SPECT MPS Analysis

Resting automated (auto) and manual operator optimized (opt) LVEF measurements were obtained using ECT toolbox (ECT) and Cedars-Sinai Medical Center QGS software analysis during the resting component of a stress/rest myocardial perfusion protocol in the supine and prone positions in separate cohorts of 55 Tc-99m sestamibi (MIBI) (8 MBq/kg) and 50 thallium (Tl-201) (0.8 MBq/kg + 15 MBq), 8 minute/8 frame gated MPS obtained within 10-15 minutes of radiotracer injection on a General Electric

Manual Optimization and Tracking of the Left Ventricular Wall

ECT software required significantly more manual optimization of valve planes than QGS software in both supine and prone positions regardless of radiotracer used (P < .0001 to .02 for all combinations) (Table 1). Tl-201 generally required less manual intervention and optimization of valve planes than MIBI, especially when QGS software was used (Table 1). For ECT software, 10/110 (9.1%) of supine and prone MIBI pairs and 2/100 (2%) of supine and prone Tl-201 pairs required manual optimization of

Discussion

This study demonstrates reproducibility of serial re-acquisitions of automated ECG-gated Tl-201 and MIBI LVEF measurements obtained on a new generation solid-state cardiac camera system during myocardial perfusion imaging using either ECT toolbox or QGS software is inferior to that obtained with planar GCBPS imaging. There is, however, significant improvement in reproducibility when valve planes and tracking of the left ventricular wall are manually checked and optimized when required prior to

Conclusion

Reproducibility of gated sestamibi and Tl-201 LVEF measurements obtained during myocardial perfusion imaging with ECT toolbox or QGS software packages using a new generation solid-state cardiac camera with improved image quality approaches that of planar GCBPS however requires visual quality control and operator optimization of left ventricular wall tracking for best results. With new solid-state cardiac technology, Tl-201 also appears at least equivalent to sestamibi for reproducibly measuring

Disclosure

All authors have nothing to declare.

References (9)

There are more references available in the full text version of this article.
View full text