Radiographic restoration technique based on a simple scatter-and-noise degradation model to enhance image visibility in cone-beam computed tomography systems using an offset detector

Abstract This study presents a novel radiographic restoration technique based on a simple scatter-and-noise degradation model to enhance image visibility in cone-beam computed tomography (CBCT) systems using an offset detector. This method aims to address the challenges of scatter and noise in CBCT images. It involves three main steps: estimating the scattering intensity and transmission function from the acquired projections, restoring projections by removing scatter and noise using the estimated quantities, and reconstructing the image using the standard filtered-backprojection algorithm. To verify the efficacy of the proposed method, we conducted an experiment on a commercially available test phantom using a tabletop setup of the micro-CT configuration, demonstrating significant improvements in key image quality metrics such as contrast-to-noise ratio (CNR), noise power spectrum (NPS), and modulation transfer function (MTF). The CNR value of the restored CBCT image was approximately 2.1, representing a 5.3-fold improvement over the original image; the NPS value evaluated at a spatial frequency of 3 lp/mm was approximately 4.2 × 10 5 , 55.3 times smaller than that for the descattered image; and the spatial resolution evaluated at an MTF value of 0.1 was approximately 1.1 lp/mm, 8.4 times greater than that of the original. Accordingly, our experimental results demonstrate that the proposed method significantly enhances the visibility and quality of CBCT images. The method is entirely software-based, making it practical for implementation without requiring additional hardware or higher radiation doses.