Super-resolution and learned perceptual loss allows high-resolution imaging of trabecular bone with clinical cone beam computed tomography

Abstract

Purpose: It is well known that osteoarthritis (OA)-related bone remodeling modifies trabecular bone structure. Subtle bone changes can be detected using high-resolution micro-computed tomography (μCT). However, the current clinical cone-beam CT (CBCT) technology cannot achieve sufficient resolution and contrast to visualize the bone microarchitecture. Previously, deep learning has been proposed to enhance image quality in musculoskeletal imaging, however, most studies generate low-resolution images using downsampling.