Value of respiratory and patient motion correction in myocardial perfusion imaging using a CZT-based SPECT camera.

Objectives Motion detection and correction have become feasible on cadmium-zinc-telluride (CZT) based SPECT cameras. Our aim was to quantify the amount of respiratory motion (RM) and patient motion (PM), and to determine the diagnostic value of automatic motion correction.

Methods We have retrospectively included 83 consecutive patients who underwent 1-day stress-rest MPI CZT-SPECT (Discovery NM/CT 570c, GE Healthcare) and an invasive fractional flow reserve (FFR) measurement the day after. The eight minutes stress SPECT acquisitions were reformatted into 1.0 s bins for RM and into 20s bins for PM. Next, we quantified the amount of RM in the cranial-caudal direction and the amount of PM in all three directions, using a commercially available software package (MDC for Alcyone, GE Healthcare). Automatic motion correction was applied on all acquisitions. Change in total perfusion deficit (TPD) and visual stress SPECT interpretation - normal, equivocal or abnormal - between the non-corrected and RM and PM corrected images were assessed. A change of 蠅7% in TPD or change in SPECT interpretation was considered to significantly influence the diagnostic outcome. Finally, we compared the FFR measurements with the changed outcomes of the visually corrected scans to determine the possible gain in accuracy for a stress-only protocol. A cut-off value of <0.75 was used to define abnormal FFR.

Results RM and PM was detected in all patients. The mean RM was 2.5±0.4 mm (range 1.7-3.7). The maximum PM was 2.4±0.8 mm (range: 0.9-5.4), 2.8±0.9 mm (range: 0.0-5.7) and 3.4±1.5 mm (range: 1.1-8.9) in the lateral, ventral-dorsal and cranial-caudal directions, respectively. RM correction influenced the diagnostic outcome in only two patients (2.4%) based on changed TPD and in nine patients (11%) based on changed visual interpretation. However, the changed outcome corresponded in only four of these nine patients with the FFR measurement, which was considered to be an improved assessment. However, in the other five patients the change was adverse. Correction for PM did not influence the diagnostic outcomes.

Conclusions Both respiratory- and patient motion was small in patients undergoing an eight minutes acquisition on a CZT SPECT camera. Correction of this motion did not improve the diagnostic accuracy and may therefore not be necessary.