Characterization of PET data acquisition system with compressed sensing detectors

We evaluated the timing performance of our data acquisition system (DAQ) with compressed sensing detectors we used in a PET insert for simultaneous PET/MR brain imaging. Compressed sensing is a promising multiplexing method that can achieve high multiplexing ratio while resolving simultaneous hits on multiple pixels without serious degradation of timing and energy resolution for positron emission tomography (PET) scanners. In our compressed sensing detector, each of the 128 silicon photomultiplier (SiPM) pixels produces a unique pattern on the 16 read out channels (the multiplexing ratio is 8:1), which are digitized and recorded at the DAQ system. The unique decoding pattern for each SiPM is used to recover the crystal indexes. In this study, we used a 10 μCi ²²Na radioactive point source to evaluate the coincidence timing resolution across the two lutetium-yttrium orthosilicate (LYSO) based block detectors. We also measured the timing integral non-linearity of the DAQ system by using a pulser to trigger both detectors with the triggering pulses for one detector offset by sequential ~ 2 ns steps. The integral non-linearity of the DAQ system is ~0.3 ns across the entire ADC sampling period, which is below the timing resolution for the measurements using a pulser as input (~ 1.0 ns FWHM). The coincidence timing resolution measured with a ²²Na source over the entire 128 crystal block is 5.98 ± 0.09 ns. Possible factors that are limiting the current timing performance are also presented.