Optimizing the Dose-Quality Paradigm in Low-Dose CT Lung Cancer Screening: A Narrative Review

Abstract

Low dose computed tomography (LDCT) is the standard for lung cancer screening, proven to reduce mortality by enabling early detection. However, its implementation requires a careful balance between minimizing patient radiation exposure and maintaining diagnostic image quality. This narrative review examines the evolution of this dose-quality paradigm in LDCT. We deconstruct the core principles of CT dosimetry (CTDIvol and DLP) and image quality (noise, resolution), explaining the fundamental physical trade-off where noise is inversely proportional to the square root of the dose. The review then traces the progression of optimization strategies, from patient-specific Automatic Exposure Control (AEC) to the paradigm shift in image reconstruction. We detail the evolution from traditional Filtered Back Projection (FBP) to advanced algorithms, including Hybrid Iterative Reconstruction (HIR), Model-Based Iterative Reconstruction (MBIR), and the changing impact of Deep Learning Image Reconstruction (DLIR). These technologies have progressively enabled significant dose reductions while preserving nodule detectability. We also discuss the synergistic role of Artificial Intelligence (AI) in improving diagnostic accuracy and the overwhelmingly positive benefit-to-risk ratio of screening. Future frontiers, such as Photon-Counting CT (PCCT) combined with AI, promise to further refine this balance, leading to more personalized and efficient screening protocols.