Performance study of a 360° CZT camera for dosimetry planning step in SIRT treatment of liver cancer and liver metastases
Résumé
Aim/Introduction: SIRT treatment planning requires an essential SPECT/CT-based pretreatment step to plan 90Y microspheres treatment dosimetry by means of intra-arterial 99mTc-MAA administration as a surrogate. Usually, conventional camera is used In this study, we propose to evaluate the performances of a 360° CZT camera on the SIRT dosimetric treatment planning and for the acquisition time optimization. Materials and Methods: A 6,2L uniform Jaszczak phantom flled with a 99mTc activity of 400 MBq and a NEMA IEC BODY phantom, flled with a 99mTc volume activity of 0.166MBq/mL, were used to compute mean absorbed dose recovery coefcients (mDRC). A homemade phantom, made with a water vial and a 8mL sphere inserted inside and flled respectively with 0,34MBq/mL and 1,59MBq/mL of 99mTc, was used to simulate the dosimetry of a healthy perfused liver and a liver tumor with a tumor-to-normal tissue ration (TN) at least 4. Acquisitions were performed on the energy peak centered at 140,5keV (+/- 7.5%) and a lower scatter window [109-130]keV. Reconstruction parameters (RP) were: 8 subsets, 8 iterations, CT attenuation correction, with and without scatter correction. The Partial Volume efect (PVE) correction, PSF compensation, convolution pre-flter and median post-flter were studied. The mDRC were computed with voxel-Svalues (VSV) and Local Deposition Method (LDM) from commercial software and compare to MIRD-based values in order to model the spatial resolution efect on absorbed dose as a function of phantom volumes and according to RP studied. The acquisition time optimization were consisted in studying the efect of the reduction of acquisition time on the uncertainties of the dosimetry calculation for acquisition durations of 10, 8, 6, 4 and 2 minutes. Results: For the uniform Jaszczak, RP do not infuence dosimetic results. For IEC phantom, the mDRC of spheres depend largely on RP with a variation from 1.5 to 12% depending on the sphere diameter. LDM gives better results in terms of mDRC than VSV method, whatever the sphere diameter. These observations are also made with the homemade phantom : the absorbed dose in the vial is not infuenced by RP while the sphere is strongly impacted. Conclusion: The 360° CZT camera provides reliable dosimetry treatment planning from a lesion volume greater than 8mL. The acquisition time optimization is currently being analyzed.