CT image reconstruction methods decrease exposure to radiations while ensuring ensure high image quality in low dose CT. This model presents a new penalized weighted least squares (PWLS) reconstruction method that uses regularization based upon Union of Learned TRAnsforms (PWLS-ULTRA). ULTRA is pre-learned from various image patches taken from dataset of CT scan images. Alternation between image reconstruction step and a coding and clustering step optimizes the PWLS-based cost function. The model shows significant improvement in image quality as compared to PWLS-EP (PWLS reconstruction with a nonadaptive edge-preserving regularizer). The use of patch based weights improves both image resolution uniformity and image quality. This model is much more faster and yields better quality images.
Input variables : Low Dose CT Image
Output Variables : Reconstructed CT image
Visit Model : github.com
Additional links : arxiv.org
| Model Category | : | Public |
| Date Published | : | June, 2018 |
| Healthcare Domain | : | Medical Technology |
| Code | : | github.com |
| Medical Imaging |
| Image Synthesis |