Overcorrection of iodinated contrast attenuation in SPECT-CT: Phantom studies

Abstract

Purpose: The authors investigated the effects of iodinated contrast agents on the quantification of radioactivity obtained using a combined single photon emission tomography (SPECT) and high resolution 16 slice computed tomography (CT) scanner. Methods: Two conditions were evaluated: Contrast media mixed with radioisotope and the contrast media, and the radioisotope solution confined in adjacent but distinct volumes. Phantoms containing combinations of T 99m c, I 111 n, and I 131 solutions in normal saline and contrast solutions in normal saline corresponding to the two conditions were prepared and scanned. Images were reconstructed by iterative reconstruction (ordered subset expectation maximization), with and without CT-derived attenuation maps. The reconstructed counts encoded in the reconstructed images were compared. Results: Compared to normal saline, the presence of an iodinated contrast agent resulted in the underestimation of reconstructed counts in all nonattenuation corrected SPECT images. Compared to normal saline, the presence of a contrast agent resulted in the overestimation of reconstructed counts in all attenuation corrected SPECT images. The largest underestimation of reconstructed counts was 12.8% and the largest overestimation value was 35.9%. Conclusions: High concentrations of a contrast agent caused errors in the measurement of actual radiotracer reconstructed counts in phantoms. These errors could have important consequences for the use of oral and intravenous contrast in both diagnostic (qualitative) and dosimetric (quantitative) SPECT-CT imaging. The authors' findings suggest that noncontrast imaging or alternative contrast-specific attenuation correction approaches should be considered for optimal SPECT activity quantification. © 2010 American Association of Physicists in Medicine.