Program Information
Physiological Modeling of Lung Motion Based On Hyperpolarized Gas Tagging MRI
T Cui1*, L Hu2 , W Miller3 , F Yin4 , J Cai5 , (1) Duke University, Durham, NC, (2) Duke University Medical Center, Durham, NC, (3) University of Virginia, Charlottesville, VA, (4) Duke University Medical Center, Durham, NC, (5) Duke University Medical Center, Durham, NC
Presentations
TH-CD-204-9 (Thursday, July 16, 2015) 10:00 AM - 12:00 PM Room: 204
Purpose:
To model the respiratory motion of lung based on the physiological ground truth measured with hyperpolarized (HP) gas tagging magnetic resonance imaging (MRI).
Methods:
A novel 4D model was established to investigate the respiratory motion of lung based on the 3D HP gas tagging MRI (t-MRI) and 3D proton MRI (p-MRI). Images were acquired on one healthy volunteer in breath-hold at the inhalation and the exhalation phases, respectively.
We obtained the displacement vector field (DVF) of each tagged grid (t-DVF) in the t-MRI images by directly measuring the spatial difference of the centroid of the grid, which therefore represented the physiological ground truth. The DVF of each voxel (p-DVF) in the p-MRI images were generated with deformable image registration. Regional ventilation, a metric used to evaluate of lung function, was calculated using the Jacobian determinant of DVF.
The respiratory motion of lung was estimated by optimizing the p-DVF to minimize the differences between p-DVF and t-DVF, while preserving the ground truth of ventilation. The optimization was performed by considering the correction of t-DVF as a perturbation to the existing p-DVF. In order to evaluate the resultant model, the motion and the ventilation were compared to those of t-MRI and p-MRI, respectively.
Results:
Our 4D model represents a more realistic lung motion with the average deviation from the ground truth t-DVF reduced from 6.4 mm to 5.2 mm (p<0.05, student t-test) compared to the p-DVF. Furthermore, the accuracy of regional ventilation was also improved with larger cross correlation (0.98 v.s. 0.37) and mutual information values (5.84 v.s. 0.45).
Conclusion:
We have successfully created a novel 4D motion model of lung based on the physiological ground truth measured from HP gas tagging MRI. With this model, an accurate prediction of lung motion and deformation during respiration in 4D becomes possible.
Contact Email: