Experimental study of optimal energy weighting in energy-resolved CT using a CZT detector

Recent advances in energy-resolved CT can potentially improve contrast-to-noise ratio (CNR), which could subsequently reduce dose in conventional and dedicated breast CT. Two methods have been proposed for optimal energy weighting: weighting the energy-bin data prior to log normalization (projection-based weighting) and weighting the energy-bin data after log normalization (image-based weighting). Previous studies suggested that optimal projection-based and image-based energy weighting provide similar CNR improvements for energy-resolved CT compared to photon-counting or conventional energy-integrating CT. This study experimentally investigated the improvement in CNR of projection-based and image-based weighted images relative to photon-counting for six different energy-bin combinations using a bench top system with a CZT detector. The results showed CNR values ranged between 0.85 and 1.01 for the projection-based weighted images and between 0.91 and 1.43 for the image-based weighted images, relative to the CNR for the photon-counting image. The range of CNR values demonstrates the effects of energy-bin selection on CNR for a particular energy weighting scheme. The non-ideal spectral response of the CZT detector caused spectral tailing, which appears to generally reduce the CNR for the projection-based weighted images. Image-based weighting increased CNR in five of the six bin combinations despite the non-ideal spectral effects