Quantcast

Validation of a short-scan-time imaging protocol for thallium-201 myocardial SPECT with a multifocal collimator.

Research paper by Yoriko Y Horiguchi, Tomohiro T Ueda, Tomofumi T Shiomori, Masahiko M Kanna, Hirooki H Matsushita, Tomoko T Kawaminami, Yuta Y Sudo, Shinnosuke S Kikuchi, Ryo R Sasaki, Jun J Hoshimiya, Yukiko Y Morita

Indexed on: 21 Jul '14Published on: 21 Jul '14Published in: Annals of nuclear medicine



Abstract

IQ-SPECT (Siemens AG, Munich, Germany) is a highly sensitive single-photon-emission computed tomography (SPECT) myocardial perfusion imaging (MPI) system that uses a multifocal collimator. We searched for a suitable protocol for short-time imaging by IQ-SPECT in thallium-201 (Tl-201) MPI by evaluating phantom images and also by comparing human IQ-SPECT images with conventional SPECT images as reference standards.We assessed the image quality using the normalized mean square error (NMSE) and drew up count profiles in Tl-201 SPECT images acquired with IQ-SPECT in a phantom study. We also performed Tl-201 stress myocardial SPECT/CT in 21 patients and compared delayed images acquired by using IQ-SPECT with 36 or 17 views per head with images obtained by using conventional SPECT.The NMSE of SPECT images from IQ-SPECT with 36 views was approximately one-fifth of that with 17 views. The myocardial count profile of images with 17 views was lower than those of images with 36 or 104 views in some regions. Defect scores were significantly lower, and image quality scores higher, in images from conventional SPECT than in those from IQ-SPECT with 17 views. Defect scores and image quality scores were equivalent in images from conventional SPECT and those from IQ-SPECT with 36 views. Agreement with the results of conventional SPECT in terms of coronary artery territory-based defect judgment was the best in IQ-SPECT with 36 views with computed tomography-derived attenuation correction (CTAC): the kappa values for IQ-SPECT with 36 views were 0.76 (without CTAC) and 0.83 (with CTAC), and those for IQ-SPECT with 17 views were 0.62 (without CTAC) and 0.59 (with CTAC). The difference in quantitative tracer uptake between conventional SPECT images and IQ-SPECT images was significantly greater for IQ-SPECT images with 17 views than for those with 36 views.Scanning with 36 views per head with CTAC may be appropriate for Tl-201 MPI using IQ-SPECT, because it provides images equivalent to those using conventional SPECT.