Abstract #1292

Sundaram T, Gee J, Kubo S, Takahashi M, Hatabu H

Lung MR can be used to noninvasively visualize and assess pulmonary anatomy and dynamics. We have previously demonstrated and validated non-rigid registration algorithms for computing the motion between successive images. However, the problem of spatiotemporally matching individuals with unique respiratory rates and lung capacities is challenging. Using a specialized algorithm, we can interpolate anatomies at specified positions along the geodesic path between two images. We reconstruct these intermediate images using total lung deformation as our parametric basis, and thereby interpolate at the same physiologic point in the respiratory cycles of multiple individuals to construct a normative atlas.