Research On Real-time Processing Technology For In-beam PET Of Heavy-ion Cancer Treatment

Having many unique advantages,heavy ion cancer treatment technology becomes the hotspot in the fields of radiation medicine and nuclear medicine.Heavy-Ion Medical Machine(HIMM),designed by the Institute of Modern Physics,is mainly used for non-invasive precision cancer treatment.It was registered as a medical device in 2019,which was a groundbreaking event in the field of nuclear technology for civilian use in China.Tumors can be effectively killed by HIMM through Bragg peak and relative biological effect(RBE)when the Carbon beam passes through human body,which holds great significance in improving the quality of people’s life.The monitoring of the real-time spatial distribution and measuring does of beam during the treatment can ensure the safety and accuracy of the treatment plan.Meantime,the treatment plan is accurately conducted if the real-time imaging technology is applied.Positron emission tomography(PET)installed on the HIMM treatment beam is a kind of in-beam PET that is an ideal non-invasive detection method for the area that needs treatment.In-beam PET can provide imaging reference for the heavy ion radio-therapy,which can effectively monitor and verify the implementation of treatment plan.With the detector,in-beam PET can convert the optical signals into the electrical signals which are then sent to the electronic system.In-beam PET is composed of the following four parts:detector module,Data Acquisition Unit(DAQU),Central Process-ing Module(CPM)and Clock and Synchronization Unit(CSU).The detector module is composed of Dual-Head Planar-Type detector array that contains several detector units to capture the annihilation photons in the target area,converse the optical signals into the electrical signals,and sent the signals to the data acquisition unit.DAQU digitizes the event data and extracts its information.The digitized data will soon be imported into the CPM.The Central Processing Module then dose its function to realize the real-time coincidence judgment of annihilation events and communicate with the host sever.CPM can be expanded into a distributed processing unit to improve the adaptability of the system to accommodate more detector blocks.CSU provides clock and synchro-nization signals to each DAQU to ensure a unified timestamp of the system.This paper studies the real-time processing technology of in-beam PET for heavy ion cancer treatment.A distributed high-speed data processing framework is studied and constructed.The design of real-time coincidence logic based on timestamp are carried out under this framework.The real-time coincidence selection technology is realized by digital timestamp measurement,distributed sorting and multi-pipeline technology,which can judge and select the coincidence event by extracting the time and location information form data.The high counting rate can be supported by the design of pipeline and the improvement of system operating frequency.The implementation of distributed high-speed data processing framework is constructed by the design of multi-level optical fiber transmission link and high-speed PCIe interface.In order to reduce the pressure of computer storage in processing a huge mass of data,an adaptive filtering algorithm is designed to remove the unnecessary storage caused by the invalid filler data.At present,the research results of this paper have been used in the PET system with16 detectors.Online acquisition,real-time coincidence,control and data transmission with the events rate of 16×10~6events/s are realized.A series of tests are conducted to verify its performance.The real-time coincidence judgment algorithm can adequately support the highest events rate processing of the DAQU.The real-time processing capac-ity of in-beam PET system is efficiently ensured since the adaptive filtering algorithm can relieve about 43.75%of the computer storage and the high-speed data processing can maintain up to 20 Gbps.This work lays a foundation for improving the efficiency of image reconstruction and the quality of the reconstructed images.