Comparison of a new whole body CZT SPECT-CT camera vs. dedicated cardiac CZT for myocardial perfusion imaging using an anthropomorphic torso phantom.

Introduction: We compared a new whole body CZT SPECT-CT camera (DNM 670CZT) to a cardiac dedicated camera (DNM 530c, GE Healthcare) for myocardial perfusion imaging using anthropomorphic torso phantom.

Methods: We used an anthropomorphic torso phantom with a cardiac insert filled with a solution of ^99mTc (mediastinum, liver, heart), with and without a defect mimicking a normal SPECT and an antero septal infarction. A set of list-mode acquisitions was performed. Acquisition time was set to 30 sec/projection for DNM 670CZT (15 projections/head over 180°, with a total acquisition time of 450 sec) and to 480 sec for DNM 530c. Images for different collection times were subsequently reframed to 7.5, 5, 2.5 and 1.5 min for both cameras. All images were reconstructed using a commercially available workstation, according to the manufacturer recommendations. Segmental uptake was evaluated using a 17-segment model. The impact of the type of camera and acquisition time were assessed successively for the phantom without and with cardiac defect using a linear model. The concordance for segmental uptake was assessed between the reference acquisition time (7.5 min) and each post-hoc collection time (5, 2.5 and 1.5 min) using Lin’s concordance correlation coefficient (CCC).

Results: On the normal SPECT, there was a significant impact of acquisition time (p<0.001) and camera type (p = 0.02). Using the DNM670CZT yielded a poor CCC for acquisition time below 2.5 min (CCC=0.27). Using the DNM530c, CCC demonstrated an almost perfect concordance between all acquisition times (all CCC>0.98). On the phantom with abnormal perfusion, there was no impact of acquisition time and camera type on the assessment of segmental uptake (p=ns). Lin’s CCC between all acquisition time ranged from 0.89 to 0.97 using the DNM670CZT and from 0.96 to 0.97 using the DNM530c.

Conclusions: This study suggested that DNM670CZT may be more sensitive to acquisition time than the cardiac dedicated DNM530c for myocardial perfusion imaging.