3D Pulmonary Ventilation Based on 4D-CT and Deformation Image Registration

BACKGROUND: Existing methods for quantification of lung ventilation require a tracer gas and specialized imaging such as single photon emission computed tomography (SPECT) or magnetic resonance imaging (MRI). Limitations of these methods include slow imaging speed, complex interpretation, and heavy burden on patients. OBJECTIVE: To assess 3D ventilation quantification based on 4D-CT and multi-resolution 3D B-spline deformable image registration (DIR). METHODS: All 4D-CT data sets were acquired with patients during quiet breathing. A 3D displacement vector field (DVF) of two different phase 4D-CT image pairs was calculated using 3D B-spline DIR algorithms, and converting to the Jacobian determinant. The axial grayscale images generated were colorized prior to fusion with the CT images and coronal and sagittal sections were reconstructed. The contours of the ventilation regions with different Jacobian values were delineated and their volumes were calculated. RESULTS: Based on 4D-CT images of patients and multi-resolution 3D B-spline DIR, 3D ventilation images can be easily generated and quantified. The maximum lung volume changes are significantly related to functional lung volumes at a level of P = 0.05 (bilateral). CONCLUSIONS: It is feasible to quantify the volume distribution of pulmonary ventilation based on 4D-CT images and 3D B-spline DIR.