Abstract
The retention index (R.I.) in Thalliun-201-chloride (^<201>Tl-Cl) brain tumor single photon emission computed tomography (SPECT) imaging analysis is useful in assessing grades of malignancy. When the T/N ratio that influences the R.I. is calculated, it is ideal to draw an ROI on the same slices and coordinates of both early and delayed transaxial images. However, in many institutions, the image slice and ROI are determined manually by operators (i.e., by the conventional method). The Co-registration function within iNRT software enables an ROI to be automatically drawn. In this study, we obtained the R.I. from six brain tumor patients employing two different methods (i.e., Co-registration and the conventional method). Each R.I. was obtained from the same personal computer that was used to transfer the transaxial images. We then compared and evaluated the six R.I. values of seven operators. (2 radiologists and 5 radiological technologists). The results showed that the R.I. values did not differ significantly between the two methods, except in two cases. Furthermore, the coefficient of variation (CV) of the R.I. obtained from the Co-registration method was lower than the conventional method in all cases. The variations among operators were also smaller in the Co-registration method. These outcomes may have been due to the fact that in the conventional method it is difficult to identify the same slices of early and delayed images. Therefore, there were differences among operators in their ways of specifying positions of the ROI. This may have resulted in differences in the number of counts/pixel of the ROI. In conclusion, the R.I. obtained by using the Co-registration method would be more useful for improving diagnostic accuracy in ^<201>Tl-Cl brain tumor SPECT imaging than the conventional method.
