Comparison of Non-Contrast Computed Tomography Pulmonary Perfusion Imaging and Contrast-Enhanced Iodine Density Maps
Abstract
Purpose
Computed tomography-derived pulmonary perfusion (CTQ) imaging is an emerging, accessible modality that derives regional pulmonary perfusion information from CT datasets. To facilitate clinical translation of this new technology, we compared two CTQ metrics: non-contrast CT-based perfusion (NCCT-Q) and CT iodine density (CT-ID) maps.
Methods
Breath-hold CT scans were acquired at end-expiration and end-inspiration before injection of iodinated contrast agents and repeated at end-inspiration after contrast injection for 15 thoracic cancer patients. NCCT-Q images were generated using the CT:VQTM software (4DMedical, Australia), which estimates perfusion by measuring regional blood mass variations derived from inhale/exhale CT density changes, corrected for respiratory-induced air volume changes. Regional iodine density was quantified by calculating the HU enhancement between pre- and post-contrast CT images using a previously published deformable image registration-based method. Regional perfusion matching was evaluated at both the lobar and regional levels. First, relative perfusion was compared across all lung lobes using Spearman correlation and Bland–Altman analysis. Second, a regional region of interest (ROI)-based analysis (40 ROIs per patient) used patient-specific Spearman correlation coefficients to assess spatial function matching between the CTQ metrics.
Results
A strong correlation was observed between relative lung lobe perfusion on the NCCT-Q and CT-ID images (Spearman's rho = 0.924, p < 0.001). The Bland–Altman analysis demonstrated a mean difference of 0.00% with 95% limits of agreement between -6.34% and 6.34%. In the regional ROI-based analysis, the median patient-specific Spearman correlation coefficients was 0.671 (range: 0.05–0.900), indicating substantial regional agreement between the CTQ metrics.
Conclusion
These findings demonstrate that both NCCT-Q and CT-ID may measure the same underlying physiological process, providing reciprocal validation of their physiological relevance in assessing regional pulmonary perfusion. Further validation against the current clinical gold standard (e.g., SPECT/Q) is warranted to support clinical adoption of these CTQ technologies.