A 4D-CT Phantom Study of Lung Tumor Volume Variability Under Respiratory Motion
Abstract
Purpose
To evaluate the impact of four-dimensional CT (4D-CT) reconstruction methods on lung tumor volume estimation and apparent tissue density under respiratory motion using dynamic phantoms with diverse geometries.
Methods
Dynamic phantoms containing spherical and non-spherical (cylinders and truncated cones) lung tumor surrogates (10–61.7 mm diameter) of varying materials were scanned using an in-house-developed respiratory motion simulation device and a clinical 4D-CT protocol. Phase-resolved (50%), average intensity projection (AIP), and maximum intensity projection (MIP) datasets were reconstructed. Tumors were delineated using a standardized absolute Hounsfield unit (HU)–based threshold in MIM. Tumor equivalent diameter, segmented volume, and mean HU were calculated. Relative volume differences for AIP and MIP were evaluated against the 50% phase baseline.
Results
The 50% phase diameters agreed with physical dimensions; discrepancies were limited to small targets due to partial-volume effects. Relative volume differences for 50% phase were within ±11%. In contrast, AIP demonstrated systematic overestimation (30–170%), while MIP exhibited larger inflation, exceeding 150% and approaching 300% for small tumors, consistent with motion-envelope effects. Trends were consistent across target shapes. Volume discrepancies were size-dependent and pronounced for lower-contrast materials. Mean tumor HUs decreased markedly from 50% phase to AIP and MIP, confirming motion-induced density dilution and explaining segmentation instability.
Conclusion
4D-CT reconstruction method significantly influences lung tumor volume estimation and apparent tissue density under respiratory motion. Although AIP and MIP images may appear visually compact, they introduce substantial volumetric and density biases due to motion-induced contrast degradation and envelope effects. For tumors near critical structures (e.g., proximal bronchial tree, spinal cord, and heart), utilization of phase-resolved gating (e.g., 30–70% respiratory phases) may reduce treatment volume and dose to organs at risk. These findings confirm the importance of phase-resolved and gated imaging for accurate motion-managed lung radiotherapy planning, particularly in lung SBRT.