Background Conventional dental panoramic imaging is often limited by soft tissue overlap and shadow artefacts from the cervical vertebrae, reducing diagnostic visibility. Purpose This study presents a single-energy material decomposition (SEMD) framework that generates material-specific panoramic images directly from pre-acquired cone-beam computed tomography (CBCT) projections without additional scans. Methods Single-energy projections were separated into soft- and hard-tissue components, and panoramic projections were extracted along the desired scan trajectory to reconstruct material-decomposed images with reduced tissue overlap and vertebral shadows. The SEMD approach was evaluated using numerical simulations and physical phantom experiments. Results Numerical simulations and phantom experiments demonstrated improved anatomical delineation and higher diagnostic contrast than conventional panoramic imaging. Quantitative evaluations showed increased lesion contrast-to-noise ratios, low normalized root-mean-square error and normalized mean absolute error relative to the ground-truth material maps, and improved perception-based image quality scores in the hard-tissue image. Conclusions The proposed framework enables material-specific panoramic imaging from a single CBCT acquisition, improving lesion visibility.
Yang et al. (Fri,) studied this question.