Rationale and objectives: The aim of this study was to investigate a method for quantifying lung motion from the registration of successive images in serial magnetic resonance imaging acquisitions during normal respiration.
Materials and methods: Estimates of pulmonary motion were obtained by summing the normalized cross-correlation over serially acquired lung images to identify corresponding locations between the images. The estimated motions were modeled as deformations of an elastic body and thus reflect to a first order approximation the true physical behavior of lung parenchyma. The Lagrangian strain, derived from the calculated motion fields, were used to quantify the tissue deformation induced in the lung over the serial acquisition.
Results: The method was validated on a magnetic resonance imaging study, for which breath-hold images were acquired of a healthy volunteer at different phases of the respiratory cycle. Regional parenchymal strain was observed to be oriented toward the pulmonary hilum, with strain magnitude maximal at the midcycle of the expiratory phase.
Conclusion: In vivo magnetic resonance imaging quantification of lung motion holds the potential of providing a new diagnostic dimension in the assessment of pulmonary function, augmenting the information provided by studies of ventilation and perfusion.