Analytical Description of Long Time Scale Diffusion MRI of the Human Lung

Niels Buhl^1,2, Sune Nrhj Jespersen²

We present an analytical treatment of long time scale diffusion weighted MRI of the human lung. Diffusion in the airways of the lung is modeled as diffusion on a network of connected line segments, with the relative position of each line segment given by the symmetric branching geometry. Within this model we are able to average the time dependent diffusion coefficient over all airways of the acinar region. The theory is in good agreement with published 3He long time scale measurements, provided the latter may be treated as roughly satisfying the Gaussian phase approximation.