/sup 131/I tumor quantification: a new background-adaptive method

The authors examine the accuracy of focal activity /sup 131/I quantification using either one of two new methods which don't explicitly employ scatter/penetration correction or using the dual-energy-window "scatter" correction method in an iterative background-adaptive mode. All imaging is carried out using circular orbits with a Picker Prism 3000XP triple-headed camera. Reconstruction by the SAGE iterative algorithm includes attenuation correction. Volumes of interest are set by hand drawing outlines on CT scans. These scans are fused to the SPECT scans. To provide the necessary calibrations, an elliptical phantom with a central 6 cm diameter sphere is imaged at two different values of radius of rotation with the water surrounding the sphere containing either no background activity or a known level. By taking account of the influence of the radius of rotation on resolution, all three methods achieve better than 11% accuracy in an inhomogeneous test phantom. However, the new method that employs a background-and-radius-adaptive conversion factor but doesn't use dual-energy-window scatter correction has the best overall accuracy: 2.8% in the phantom and 8.4% in a patient. Moreover, it yields a patient SPECT time activity curve which is smooth. It may be a good, practical technique since it has better than 10% accuracy and a reasonable noise level.