TY - JOUR
T1 - Motion detection and amelioration in a dedicated cardiac solid-state CZT SPECT device
AU - Kennedy, John A.
AU - William Strauss, H.
N1 - Publisher Copyright:
© 2016, International Federation for Medical and Biological Engineering.
PY - 2017/4/1
Y1 - 2017/4/1
N2 - A solid-state cadmium zinc tellurium (CZT) dedicated multipinhole cardiac camera which acquires all views simultaneously has been introduced for myocardial SPECT acquisition. We report a method to detect and ameliorate patient motion artifacts in myocardial perfusion imaging (MPI) studies recorded with this device. To detect motion, a myocardial phantom study was recorded, and at mid-scan, the phantom was moved stepwise along each of 6 orthogonal directions, causing MPI artifacts. Using QPS software (Cedars-Sinai) and an in-house normal database, displacements giving artifactual perfusion defects (total perfusion deficit score, TPD, >5 %) were all 1.5 cm or greater (11.2 ± 1.3 % for 1.5 cm). List mode data were reframed into 10-s steps, and the norm of the changes in center of mass among the 19 projections (32 × 32 matrix, pixel size 2.46 mm) was used as a motion index. Rejection of misregistered data gave artifact-free reconstructions (TPD = 1.0 ± 0.8 %) in phantom scans and reduced blur in a rest/stress clinical study. Blur on the patient’s stress scan was consistent with increased motion compared to rest (motion index of 4.4 vs. 3.0 pixels, respectively). For CZT cameras that acquire data from multiple views simultaneously, motion during MPI can cause clinically significant artifacts. Reframing acquisitions into discrete time intervals enables the detection of motion and its amelioration, improving diagnostic accuracy.
AB - A solid-state cadmium zinc tellurium (CZT) dedicated multipinhole cardiac camera which acquires all views simultaneously has been introduced for myocardial SPECT acquisition. We report a method to detect and ameliorate patient motion artifacts in myocardial perfusion imaging (MPI) studies recorded with this device. To detect motion, a myocardial phantom study was recorded, and at mid-scan, the phantom was moved stepwise along each of 6 orthogonal directions, causing MPI artifacts. Using QPS software (Cedars-Sinai) and an in-house normal database, displacements giving artifactual perfusion defects (total perfusion deficit score, TPD, >5 %) were all 1.5 cm or greater (11.2 ± 1.3 % for 1.5 cm). List mode data were reframed into 10-s steps, and the norm of the changes in center of mass among the 19 projections (32 × 32 matrix, pixel size 2.46 mm) was used as a motion index. Rejection of misregistered data gave artifact-free reconstructions (TPD = 1.0 ± 0.8 %) in phantom scans and reduced blur in a rest/stress clinical study. Blur on the patient’s stress scan was consistent with increased motion compared to rest (motion index of 4.4 vs. 3.0 pixels, respectively). For CZT cameras that acquire data from multiple views simultaneously, motion during MPI can cause clinically significant artifacts. Reframing acquisitions into discrete time intervals enables the detection of motion and its amelioration, improving diagnostic accuracy.
KW - Cadmium zinc telluride
KW - Motion correction
KW - Myocardial perfusion imaging
KW - SPECT instrumentation
KW - Ultrafast cardiac gamma camera
UR - http://www.scopus.com/inward/record.url?scp=84978869489&partnerID=8YFLogxK
U2 - 10.1007/s11517-016-1548-z
DO - 10.1007/s11517-016-1548-z
M3 - Article
AN - SCOPUS:84978869489
SN - 0140-0118
VL - 55
SP - 663
EP - 671
JO - Medical and Biological Engineering and Computing
JF - Medical and Biological Engineering and Computing
IS - 4
ER -