Abstract
Computed tomography (CT) imaging with high energy resolution
detectors shows great promise in material decomposition and
multi-contrast imaging. Multi-contrast imaging was studied by
imaging a phantom with iodine (I), gadolinium (Gd), and gold (Au)
solutions, and mixtures of the three using a cadmium telluride
(CdTe) spectrometer with an energy resolution of 1% as well as with
a cadmium zinc telluride (CZT) detector with an energy resolution of
13%. The phantom was imaged at 120 kVp and 1.1 mA with 7 mm of
aluminum filtration. For the CdTe data collection, the phantom was
imaged using a 0.2 mm diameter x-ray beam with 96 ten-second data
acquisitions across the phantom at 45 rotation angles. For the CZT
detector, we had 720 projections using a cone beam, and the six
detector energy thresholds were set to 23, 33, 50, 64, 81, and
120 keV so that three thresholds corresponded to the K-edges of the
contrast agents. Contrast agent isolation methods were then
examined. K-edge subtraction and novel spectrometric algebraic image
reconstruction (SAIR) were used for the CdTe data. K-edge
subtraction alone was used for the CZT data. Linearity plots
produced similar R
2 values and slopes for all three
reconstruction methods. Comparing CdTe methods, SAIR offered less
noise than CdTe K-edge subtraction and better geometric accuracy at
low contrast concentrations. CdTe contrast agent images of I, Gd,
and Au offered less noise and greater contrast than the CZT images,
highlighting the benefits of high energy resolution CdTe detectors
for possible use in pre-clinical or clinical CT imaging.