Assessment of different regions of interest-based methods for [99mTc]Tc DAT-SPECT quantification using an anthropomorphic striatal phantom.

BACKGROUND AND AIMS

Molecular imaging of the dopamine transporters (DAT) provides valuable information about neurodegenerative diseases, such as Parkinson's. This study assessed the accuracy and precision of DAT-SPECT quantification methods.

METHODS

Twenty-three DAT-SPECT images of a striatal phantom were acquired. The specific (caudate and putamen) and the non-specific (background activity) chambers were filled with [Tc]Tc. Different specific-to-non-specific activity ratios (10, 9, 8, 7, 6, 5, 4, 3 and 2 to 1) and the specific binding ratio (SBR) were calculated. Five methods using ROIs were assessed: (a) Manual ROIs on SPECT images; (b) TwoBox and (c) ThreeBox methods and Volume of Interest (VOI) using structural images; (d) MRI and (e) CT. Accuracy was evaluated by the concordance correlation coefficient (CCC) and precision by Pearson's coefficient and linear regression.

RESULTS

The SBR quantified in the specific and striatal chambers resulted in a CCC increase with a decrease in the nominal values. For lower SBR, MRI and CT showed higher CCCs when caudate ([Formula: see text] = 0.89 e [Formula: see text] = 0.84) and putamen ([Formula: see text] = 0.86 e [Formula: see text] = 0.82) were evaluated. For striatal assessments, the TwoBox method was the most accurate ([Formula: see text] = 0.95). High Pearson's coefficients were found in the correlations between all methods.

CONCLUSIONS

All five methods showed high precision even when applied to images with different activities. MRI and CT were the most accurate for assessing the caudate or putamen. To assess the striatal chamber and in the absence of structural information, the TwoBox method is advisable.