HorusEye: a self-supervised foundation model for generalizable X-ray tomography restoration.
Authors
Affiliations (27)
Affiliations (27)
- Computer Science Program, Computer, Electrical and Mathematical Sciences and Engineering Division, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia.
- Center of Excellence for Smart Health, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia.
- Center of Excellence on GenAI, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia.
- Syneron Opal, George Town, Cayman Islands.
- Department of Biomedical Engineering, Southern University of Science and Technology, Shenzhen, China.
- Computer Science Program, Computer, Electrical and Mathematical Sciences and Engineering Division, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
- Center of Excellence for Smart Health, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
- Center of Excellence on GenAI, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
- Faculty of Computing, Harbin Institute of Technology, Harbin, China. [email protected].
- Department of Critical Care Medicine, The First Affiliated Hospital of Harbin Medical University, Harbin, China.
- Computer Science Program, Computer, Electrical and Mathematical Sciences and Engineering Division, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
- Center of Excellence for Smart Health, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
- Center of Excellence on GenAI, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
- Department of Radiology, Renji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China.
- Shanghai Academy of AI for Science, Shanghai, China.
- Department of Prosthodontics, The First Affiliated Hospital of Harbin Medical University, Harbin, China.
- Department of Computer Tomography, The First Affiliated Hospital of Harbin Medical University, Harbin, China.
- Department of Cardiology, The First Affiliated Hospital of Harbin Medical University, Harbin, China.
- Department of Computer Tomography, The First Affiliated Hospital of Harbin Medical University, Harbin, China. [email protected].
- Department of Radiology, The Fourth Hospital of Harbin Medical University, Harbin, China. [email protected].
- Computer Science Program, Computer, Electrical and Mathematical Sciences and Engineering Division, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
- College of Computer and Control Engineering, Northeast Forestry University, Harbin, China. [email protected].
- Key Laboratory of Hepatosplenic Surgery, Ministry of Education, The First Affiliated Hospital of Harbin Medical University, Harbin, China. [email protected].
- Heilongjiang TuoMeng Technology Co., Harbin, China. [email protected].
- Computer Science Program, Computer, Electrical and Mathematical Sciences and Engineering Division, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
- Center of Excellence for Smart Health, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
- Center of Excellence on GenAI, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia. [email protected].
Abstract
X-ray tomography is widely used across scientific and clinical domains, yet image degradation remains a major obstacle to reliable analysis, particularly under low-dose or data-scarce conditions. Existing restoration methods are typically designed for specific modalities and predefined degradation, limiting their generalizability. Here we show that image restoration can instead be formulated as learning realistic, nonparametric acquisition degradation processes directly from data. We introduce HorusEye, a self-supervised foundation model for X-ray tomography restoration that leverages interslice contrastive pretraining to jointly learn structural priors and degradation without paired supervision or predefined assumptions. Trained on over 100 million images, HorusEye generalizes across diverse modalities, restoration tasks and previously unseen imaging modalities, consistently outperforming task-specific approaches. Extensive evaluations demonstrate improved photon efficiency and recovery of high-frequency information. Clinical studies further demonstrate enhanced detectability of low-contrast anatomy and lesions, as well as improved performance on downstream tasks, highlighting HorusEye as a general postprocessing tool for X-ray tomography.