Automatic regularization parameter choice for tomography using a double model approach

2026-02-09

AI Summary

This work addresses the ill-posed inverse problem in X-ray computed tomography arising from limited projection data, which necessitates a careful balance between data fidelity and prior-based regularization. The authors propose an adaptive regularization parameter selection method based on dual-grid discretization, introducing—for the first time—a feedback control mechanism that dynamically adjusts the regularization strength across coarse and fine discretization models to drive their reconstructions toward consistency. This approach automatically determines the optimal regularization parameter without manual intervention. Experimental results on real tomographic data demonstrate that the proposed method significantly improves reconstruction quality, confirming its effectiveness and robustness.

Technology Category

• Search and Optimization: Mixed Discrete/Continuous Search
• Reasoning under Uncertainty: Stochastic Optimization
• Constraint Satisfaction and Optimization: Mixed Discrete/Continuous Optimization

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Image reconstruction in X-ray tomography is an ill-posed inverse problem, particularly with limited available data. Regularization is thus essential, but its effectiveness hinges on the choice of a regularization parameter that balances data fidelity against a priori information. We present a novel method for automatic parameter selection based on the use of two distinct computational discretizations of the same problem. A feedback control algorithm dynamically adjusts the regularization strength, driving an iterative reconstruction toward the smallest parameter that yields sufficient similarity between reconstructions on the two grids. The effectiveness of the proposed approach is demonstrated using real tomographic data.

Problem

• tomography
• regularization parameter
• ill-posed inverse problem
• image reconstruction
• automatic parameter selection

Innovation

• automatic regularization
• double model approach
• feedback control
• tomographic reconstruction
• ill-posed inverse problem