In beam PET acquisition on 75 MeV/u carbon beam using sampling-Based Read-out Electronics
Résumé
Positron Emission Tomography (PET) dedicated to the ion range verification in the context of hadrontherapy treatments, requires a very selective experimental device able to discriminate the two gamma from + decay from prompt particles due to the interactions of the beam with the matter. Prompt particles are a background noise in the context of induced + activity measurement. A significant part of this background noise is produced by the prompt emitted during beam spill. Therefore a precise knowledge of beam time structure is crucial to control the quality of data of in-beam PET acquisitions. In other words, beam time structure determines the number of primary ions per bunch and the nuclear induced background. An experiment was performed at the National Large Heavy Ion Accelerator (GANIL) in Caen (France) using a 75 A.MeV carbon beam in order to test the use of sampling-based read-out electronics to provide a very accurate time stamping of the detected particles. The experimental set-up is constituted by a small acceptance dual head detector positioned on each side of the beam centered on target. For this cyclotron beam, the cycle is 12 MHz with a beam spill of 10 ns. If acquisition takes into account only events occurring outside the spill, we demonstrate that the energy spectrum of events is equivalent to target radiative decay spectrum obtained just after stopping the beam. For this experiment, the technique of sampling signal was chosen, allowing to provide an accurate time measurements and adjust optimal time and energy cuts on the recorded data. With the time structure of the beam spill at GANIL, taking into account all the events, including those acquired during the spill, the measured background-to-signal event ratio is 10%