Automatic adjustment of a medical imaging data acquisition system to unknown delays in the input communication channels

Abstract

In this article, novel FIFO and RAM-based Synchronization Modules to keep synchronism throughout the input channels of a Data Acquisition Electronics (DAE) system are proposed. DAE is a main component of a Medical Imaging System, namely, a Positron Emission Mammography (PEM) system. DAE input data comes from a scanner constituted by an array of scintillating crystals. The scanner captures radiation generated by human cells injected with a radioactive substance and converts it into electrical signals. The corresponding digital information is sent to the DAE. In order to deal with the huge amount of data, flowing at high data rates, point-to-point (p2p) communication channels are used between the scanner and the DAE. Propagation delays associated with the different communication channels may change differently. Additionally, differences among channel delays may exceed one clock period. Keeping synchronism in these circumstances requires more than the classical asynchronous FIFO solution. All these aspects motivate the work proposed in this article. The PEM DAE system is a multi-board, multi-FPGA, multi-clock domain system. Therefore, the DAE architecture follows a Globally Asynchronous, Locally Synchronous (GALS) design style. The novel Synchronization Modules proposed in this article are implemented in the DAE. The effectiveness of these new structures is validated through simulation and laboratorial test. Simulation and test results are presented.