Clinical Metadata Guided Limited-Angle CT Image Reconstruction

2025-09-01

AI Summary

Limited-angle CT (LACT) cardiac imaging suffers from severe truncation artifacts and an ill-posed reconstruction problem. This paper proposes the first two-stage diffusion-based reconstruction framework guided by structured clinical metadata—including acquisition parameters, demographic information, and diagnostic impressions. In Stage I, an anatomy-aware prior is generated; in Stage II, metadata is fused to enable physics-constrained refinement, with an ADMM module ensuring data consistency. The key innovation lies in the first integration of clinical metadata into LACT reconstruction, leveraging it for both prior modeling and image refinement—substantially enhancing robustness under extreme undersampling. Quantitative evaluation on synthetic and real cardiac CT datasets demonstrates significant improvements over metadata-agnostic baselines across SSIM, PSNR, normalized mutual information (nMI), and Pearson correlation coefficient (PCC). Ablation studies confirm complementary contributions from distinct metadata categories.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Medical and Biological Imaging
• Search and Optimization: Metareasoning and Metaheuristics

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Limited-angle computed tomography (LACT) offers improved temporal resolution and reduced radiation dose for cardiac imaging, but suffers from severe artifacts due to truncated projections. To address the ill-posedness of LACT reconstruction, we propose a two-stage diffusion framework guided by structured clinical metadata. In the first stage, a transformer-based diffusion model conditioned exclusively on metadata, including acquisition parameters, patient demographics, and diagnostic impressions, generates coarse anatomical priors from noise. The second stage further refines the images by integrating both the coarse prior and metadata to produce high-fidelity results. Physics-based data consistency is enforced at each sampling step in both stages using an Alternating Direction Method of Multipliers module, ensuring alignment with the measured projections. Extensive experiments on both synthetic and real cardiac CT datasets demonstrate that incorporating metadata significantly improves reconstruction fidelity, particularly under severe angular truncation. Compared to existing metadata-free baselines, our method achieves superior performance in SSIM, PSNR, nMI, and PCC. Ablation studies confirm that different types of metadata contribute complementary benefits, particularly diagnostic and demographic priors under limited-angle conditions. These findings highlight the dual role of clinical metadata in improving both reconstruction quality and efficiency, supporting their integration into future metadata-guided medical imaging frameworks.

Problem

• Reducing artifacts in limited-angle CT cardiac imaging
• Incorporating clinical metadata to guide image reconstruction
• Improving reconstruction fidelity under severe angular truncation

Innovation

• Two-stage diffusion framework with clinical metadata
• Transformer-based model generates anatomical priors from noise
• Physics-constrained refinement using ADMM for data consistency