OA165] Optimization of quantitative 99MTC-MAA SPECT/CT imaging for 90Y RADIOEMBOLIZATION: A 3D-printed phantom study

Abstract

Purpose Radioembolization (RE) with 90Y-microspheres has gained widespread acceptance as a safe and effective technique for treating liver malignancies. Accurate quantification in RE is a key step in treatment planning optimization and it is becoming a pressing issue in the light of the Directive 2013/59/EURATOM, stating that individual dose planning must be enforced in legislation by EU member states by February 2018. The aim of this study was to develop a SPECT/CT imaging protocol for quantitative imaging optimization in RE based on a novel anthropomorphic 3D-printed phantom. Methods and materials A SPECT/CT system provided with a novel software for imaging quantification was used to assess the quantitative accuracy of 99mTc-MAA SPECT imaging for treatment planning in RE. Gamma camera calibration factors and recovery coefficients were determined performing preliminary SPECT/CT acquisitions of a homogeneous and a NEMA/IEC phantom, respectively. The quantification procedure was then validated using a 3D-printed anthropomorphic phantom provided with a fillable liver section and multiple inserts. Raw data reconstruction was performed using two independent software: i) the built-in xSPECT package, using both the Ordered Subset Expectation–Maximization (OSEM) and the Ordered Subset Conjugated Gradient (OSCG) algorithm; ii) the commercially available QSPECT software based on OSEM method. All reconstructed images were manually fused and appropriate regions of interest (ROIs) were drawn on the MIM 6.1.7 system. Results The measured activity concentration is consistent within 3% with the expected activity concentration values. This is in keeping with the overall accuracy ( 10%) stated by the manufacturer. No substantial differences were reported between xSPECT with OSCG or OSEM in terms of relative activity. However, a 20% difference was found between ROIs extracted from images reconstructed with xSPECT and QSPECT using OSEM. Several artifacts were visible in the QSPECT reconstruction, likely due to the limited area used for attenuation and scatter corrections within the software. Conclusions The investigated protocol based on the innovative xSPECT tool allows an accurate absolute quantification within 3%. QSPECT software provides the poorest results with several artifacts as well as intrinsic dishomogeneities.