Evaluation of Low-Dose CT Image Quality Metrics Using a 3D-Printed Line Group Phantom

Evaluation of Low-Dose CT Image Quality Metrics Using a 3D-Printed Line Group Phantom

Lung cancer remains a leading cause of cancer-related mortality, making early detection pivotal. While low-dose computed tomography (LDCT) significantly reduces radiation exposure compared to standard CT (STDCT), the associated increase in image noise can compromise diagnostic accuracy. This study quantifies the trade-off between radiation dose and image quality using a customized 3D-printed lung model and the Taguchi loss function. Experiments were conducted on a Philips 128-slice CT scanner with varying tube voltages (100–120 kVp) and currents (20–45 mAs). Results indicate that reducing parameters to 100 kVp/20 mAs degraded image quality by 219% relative to STDCT. However, the 120 kVp/45 mAs and 100 kVp/45 mAs protocols maintained diagnostic thresholds (MDD validation) while reducing radiation dose. We conclude that 45 mAs represents a critical threshold for maintaining structural detectability in LDCT screening.