A Design Implementation of the Central Process Module for an In-Beam PET in HICTD
L. Y. Ke,J. W. Yan,J. D. Chen,X. L. Zhang,C. X. Wang,C. M. Du,M. C. Hu,Z. Q. Yang,Y. Qian,Q. S. She,H. B. Yang,H. Y. Zhao,T. L. Pu,C. X. Pei,H. Su,J. Kong
DOI: https://doi.org/10.1016/j.nima.2021.165572
2021-01-01
Abstract:To monitor the irradiation dose and position accurately delivered to patient during particle therapy in Heavy-Ion Cancer Therapy Device (HICTD) in Lanzhou, China, an In-Beam Positron Emission Tomography (PET) has been designed as a detection device to obtain time, energy and position information of gamma ray generated from the annihilation event, which is equipped at the beam line of HICTD. Sixteen detection blocks are adopted to construct a Dual-Head Planar-Type detector array in the In-Beam PET, and each detection block is composed of pixel detection elements. Data Acquisition Unit (DAQU) is employed as the processing module to process and digitize signals from detection blocks. Each DAQU faces a pair of detection blocks, there are eight DAQUs in total in this PET at present. A Central Processor Module (CPM) developed for the detector array is presented in this paper. CPM has two crucial functions: first it is as a hub to connect with all DAQUs through the optical links for data transfer and state control, second it is as a data process engine which collects data from different DAQUs and sorts the events into the chronological order for coincidence event discrimination, and then interfaces both control signals and event data to a host server through PCIe interface for image reconstruction. A novel parallel sorting algorithm has been explored and adopted in CPM, which can effectively sort the valid events detected by detection blocks. The key performance of the CPM has been evaluated in experiments. It can handle the events up to 16 million per second at least, which is capable of sorting and discriminating coincidence events efficiently.