Cone-Beam CT Image Super-Resolution

Research Images

Methodology and Goals

This project involves developing a progressive curriculum learning-based CycleGAN to super-resolve cone-beam computed tomography (CBCT) images towards high-resolution peripheral quantitative computed tomography (HR-pQCT) quality. Super-resolved CBCT scans will yield more accurate detection and quantification of bone erosions, expanding the utility of musculoskeletal CBCT for rheumatoid arthritis research and analysis. Read more about this work here (*Computers in Biology and Medicine - December 2025*): [CL-GAN: A progressive curriculum learning approach for bone CT super-resolution](https://doi.org/10.1016/j.compbiomed.2025.111332)