Simultaneous water activation and glucose metabolic rate imaging with PET

Abstract

Simultaneous imaging of [ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">15</sup> O]H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> O activation measuring relative regional cerebral blood flow (rrCBF) and [ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">18</sup> F]FDG for glucose metabolism is of potential interest for presurgical positron emission tomography (PET) imaging of brain tumours. A novel imaging strategy is proposed to be able to separate the [ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">18</sup> F]FDG and the multiple [ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">15</sup> O]H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> O signals from a simultaneously acquired dynamic acquisition of the two PET tracers. The technique is based on the fact that the dynamics of the two tracers are very distinct. By adopting an appropriate bolus injection strategy and by defining tailored sets of basis functions that model either the FDG or water component it is possible to separate the FDG and water signal. The basis functions are defined as the convolution of estimated input functions (IPFs) with a set of decaying exponential functions, where the IPFs are estimated from the overall measured head curve. The estimated IPFs are not required to correspond to the arterial IPFs of the two components and the proposed method does not need an arterial blood input function, which greatly increases the practical feasibility of the method. Once the IPFs are estimated the voxel-level time activity curves (TACs) are modelled using the basis functions. The set of exponential functions used to define the FDG basis functions are predetermined and equal for all voxels while the set of exponential functions defining the water basis functions will be determined for each individual voxel. The technique can be applied post-reconstruction as a fitting routine using the MLEM algorithm, or, the model can be incorporated in a global 4D reconstruction strategy to further reduce the noise in the estimated components. In this work the technique is applied post-reconstruction. Simulation studies show that the [ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">18</sup> F]FDG and the multiple [ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">15</sup> O]H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> O components can be separated, when using simultaneous imaging, by the proposed model. The resulting errors are similar to the errors obtained when the two tracers are imaged separately demonstrating the feasibility of the method.