A method for selecting a protocol for routine body CT scan using Gemstone Spectral Imaging with or without adaptive statistical iterative reconstruction: phantom experiments

To investigate a method for selecting a protocol for body CT scan to acquire monochromatic images (MIs) by gemstone spectral imaging (GSI) with or without adaptive statistical iterative reconstruction (ASiR). We subjected a phantom to conventional scanning at 120 kVp and 50-700 mAs and GSI at 165-600 mAs; reconstructed MIs at 65 keV with ASiR 0-100 % for GSI; placed 5 regions of interest on each of 3 consecutive reconstructed slices to obtain the averaged standard deviation (SD) as image noise for conventional scan and GSI. Linear regression analysis yielded the mAs by conventional scan that could be used to achieve similar image noise by GSI. To achieve similar noise, we found excellent linear correlation of mAs between GSI with ASiR 0-100 % and conventional scan (r = 1.00, P < 0.0001), and obtained a table of equivalent mAs between MIs at 65 keV and conventional CT at 120 kVp. We can select a protocol for body CT scan for MIs at 65 keV with or without ASiR with results comparable to those of conventional CT at 120 kVp.