Physical Model-Based Contrast Enhancement of Computed Tomography Images: Contrast Enhancement of Computed Tomography

Abstract

Computed tomography (CT) can rapidly provide high-resolution cross-section images for clinical diagnosis. The image contrast of the CT strongly influences the visibility of lesions in the images. However, the low material-dependent characteristic of the Compton scattering (CS) lowers the image contrast. In this study, a novel physical model-based method was proposed to enhance the contrast of CT images. At first, relationships between CT number and tissue parameters were determined using the CT images and elemental composition of tissue equivalent rods. Then, the CT images to be enhanced were converted to tissue parameter maps using pre-determined relationships. By using a classical parametric fit model, partial attenuation images with enhanced image contrast can be calculated. A phantom CT image and an abdominal CT image were used to evaluate the performance of the proposed method. For the phantom CT image, the image contrast between rods and background solid water were enhanced. For the abdominal CT image, the visibility of a low-attenuation lesion in the right lobe of the liver was improved. In conclusion, the proposed method could be applied in clinical diagnosis to improve the visibility of CT images.