Abstract
BACKGROUND: While spectral CT using single photon counting detector has shown a number of advantages in diagnostic imaging, knowledge of the detector response function of an energy-resolved detector is needed to correct the signal bias and reconstruct the image more accurately.
OBJECTIVE: The objective of this paper is to study the photo counting detector response function using laboratory sources, and investigate the signal bias correction method.
METHODS: Our approach is to model the detector response function over the entire diagnostic energy range (20 keV <E< 140 keV) using a semi-analytical method with 12 parameters. The model includes a primary photo peak, an exponential tail, and four escape peaks. Four radioactive isotopes including Cdmium-109, Barium-133, Americium-241 and Cobalt-57 are used to generate the detector response function at six photon energies. The 12 parameters are obtained by non-linear least-square fitting with the measured detector response functions at the six energies. The correlations of the 12 parameters with energy are also investigated with the measured data.
RESULTS: The analytical model generally describes the detector response function and is in good agreement with the measured data. The trend lines of the 12 parameters indicate higher energies tend to cause grater spectrum distortion. The spectrum distortion caused by the detector response function on spectral CT reconstruction is analyzed theoretically, and a solution to correct this spectrum distortion is also proposed.
CONCLUSION: In spectral and fluorescence CT, the spectrum distortion caused by detector response function poses a problem and cannot be ignored in any quantitative analysis. The detector response function of a CdTe detector can be obtained by a semi-analytical method.