[OA085] CT Quantification of holmium distribution for absorbed dose calculation in a context of microbrachytherapy

Purpose Holmium 166 ( 166 Ho) is a potential candidate for internal radiotherapy due to high-energy beta radiation (1.77 and 1.85 MeV) and gamma ray (80.6 keV) adapted to SPECT imaging albeit with poor spatial resolution (cm). The spatial resolution at which activity distribution is determined impacts the absorbed dose calculation algorithm that needs to be implemented: For example, SPECT spatial resolution will often lead to the Local Energy Deposition (LED) approximation. In this study, we investigated the potential of CT for holmium quantification, and compared various absorbed dose calculation algorithms in a context of micro-brachytherapy. Methods For CT calibration, acquisitions were performed on a General Electric BrightSpeed 16 using a CIRS 62 phantom with known holmium concentration in order to get a calibration curve. We then assessed holmium concentration in brain tumours implanted in Yucatan minipigs treated by 166 Ho microbrachytherapy. Acquisition parameters were as follows: voltage = 140 kV; reconstruction filter B31s, matrix size: 512 × 512; pixel size: 0.488 × 0.488 mm 2 ; slice thickness = 0.625 mm. Holmium segmentation was based-on pre and post injection CT image subtraction. From the CT calibration curve, and with a prior knowledge of 166 Ho specific activity, it was possible to get a 3D activity distribution. Different absorbed dose computation approaches were implemented: Local Energy Deposition (LED), superposition/convolution of dose-point kernels assuming homogenous or heterogeneous media, and Monte Carlo modelling (GATE) taken as the reference. Gamma-index criteria (0.1%, 0.5 mm) were used to compare approaches. Results The LED approach was clearly not appropriated due to 166 Ho range vs. CT resolution (0.15% passing rate according to the gamma-index criteria). Homogenous and heterogeneous convolution/superposition approaches provided 78% passing rate in a situation of homogenous soft tissues, compared to GATE. Discrepancies were mainly visible in a situation of heterogeneous densities. Conclusions CT-based holmium quantification can provide relevant density information with good spatial resolution (as compared to SPECT imaging), broadening the possibilities for absorbed dose calculation algorithms. This work was partially sponsored by BPI-France under the scope of the Industrial Strategic Innovation project (projet d’Innovation Strategique Industrielle) “TheraneaM.