An Easy-to-use Method for CT Image Simulation with Parameter Optimization Using a Water Phantom

Abstract

We developed an easy-to-use method to generate computed tomography (CT) images that simulate the images obtained when using an actual scanner. The developed method generates images by simulating the data acquisition and image reconstruction processes of a scanner from a linear attenuation coefficient map of an object numerically generated. This approach is similar to general image simulation methods. However, we introduced adjustable parameters for the CT data acquisition process, for example, parameters related to X-ray attenuation in the anode of the X-ray tube and the bowtie filter. These parameters were optimized in advance by minimizing the difference between the simulated and measured images of a water phantom. To verify the validity of the developed method, a simulated image was generated for a torso phantom and then compared with the measured image of the phantom obtained using the scanner. The simulated and measured images of the torso phantom were in good agreement. The spatial resolution and noise characteristics of these two images were also comparable, further indicating the accuracy of the developed method. In the existing methods, various information/data related to an actual scanner, including difficult-to-acquire ones, were essential for image simulation. In the developed method, instead of determining the difficult-to-acquire information/data, we introduced adjustable parameters. Therefore, the developed method was easier to use than the existing methods.