Improved Statistics Based Positioning Scheme for Continuous Thick Crystal PET Detectors

PET detectors based on continuous crystal could be an ideal alternative for current high resolution pixelated PET detectors as long as an effective positioning algorithm is provided. Statistics Based Positioning (SBP) scheme has drawn much attention for its high positioning performance. However, we found that the Gaussian distribution assumption, which the two dimensional SBP (2D SBP) relies on, is inaccurate for detectors with thick crystal. The discrepancy between the assumption and the realistic situation could result in serious performance degradation. To address the problem, we propose to improve 2D SBP as ISBP (Improved Statistics Based Positioning) scheme. Using our experimental data, the performance of 2D SBP and ISBP are evaluated and compared. By the event tracking method, the different behavior of 2D SBP and ISBP is revealed and analyzed. ISBP is more suitable for detectors with thick crystal. In addition, we compare the performance by ISBP and the artificial neural network position estimator using the same experimental data. The characteristics of both positioning algorithms are discussed.