OBJECTIVE: Imperfect image resolution because of finite-sized collimator channels often causes a decrease in diagnostic performance in myocardial perfusion single-photon emission computed tomography (SPECT). This prompted us to evaluate the clinical usefulness of collimator distance dependent resolution recovery (RR) in myocardial perfusion SPECT. METHODS: We retrospectively reviewed myocardial SPECT images of 60 consecutive patients (26 men, 34 women; mean age 68 years). They consisted of 25 with coronary artery disease (CAD) and 35 without, as confirmed by the coronary angiography (CAG). The patients had undergone myocardial perfusion SPECT imaging with thallium-201 (111 MBq) under an exercise or pharmacological stress protocol. Two readers reviewed the myocardial SPECT images for the presence of CAD on a 4-point scale where 0 = normal, 1 = probably normal, 2 = probably abnormal, and 3 = abnormal. Three reading sessions were held: first, ordered subsets expectation maximization (OSEM) SPECT images, second, OSEM with RR images, and third, both OSEM and RR images. Diagnostic performance for accuracy of coronary arterial stenoses was compared for OSEM, RR and both the images. RESULTS: The sensitivity increased using RR images, but specificity decreased, and thus overall diagnostic accuracy was decreased, as compared with using OSEM images. The overall accuracy of using both the images showed the highest diagnostic performance among the three sets of image interpretations. The sensitivity, specificity and accuracy were 33, 92, 77% for OSEM, 53, 79, 72% for RR and 51, 90, 80% for both the images. CONCLUSIONS: These results suggest that RR for myocardial perfusion SPECT imaging has some potential to improve diagnostic accuracy, but both OSEM and RR image sets may have to be employed for the best diagnosis of CAD.
OBJECTIVE: Imperfect image resolution because of finite-sized collimator channels often causes a decrease in diagnostic performance in myocardial perfusion single-photon emission computed tomography (SPECT). This prompted us to evaluate the clinical usefulness of collimator distance dependent resolution recovery (RR) in myocardial perfusion SPECT. METHODS: We retrospectively reviewed myocardial SPECT images of 60 consecutive patients (26 men, 34 women; mean age 68 years). They consisted of 25 with coronary artery disease (CAD) and 35 without, as confirmed by the coronary angiography (CAG). The patients had undergone myocardial perfusion SPECT imaging with thallium-201 (111 MBq) under an exercise or pharmacological stress protocol. Two readers reviewed the myocardial SPECT images for the presence of CAD on a 4-point scale where 0 = normal, 1 = probably normal, 2 = probably abnormal, and 3 = abnormal. Three reading sessions were held: first, ordered subsets expectation maximization (OSEM) SPECT images, second, OSEM with RR images, and third, both OSEM and RR images. Diagnostic performance for accuracy of coronary arterial stenoses was compared for OSEM, RR and both the images. RESULTS: The sensitivity increased using RR images, but specificity decreased, and thus overall diagnostic accuracy was decreased, as compared with using OSEM images. The overall accuracy of using both the images showed the highest diagnostic performance among the three sets of image interpretations. The sensitivity, specificity and accuracy were 33, 92, 77% for OSEM, 53, 79, 72% for RR and 51, 90, 80% for both the images. CONCLUSIONS: These results suggest that RR for myocardial perfusion SPECT imaging has some potential to improve diagnostic accuracy, but both OSEM and RR image sets may have to be employed for the best diagnosis of CAD.