Method of evaluating respiratory induced organ motion by vector volume histogram.

PURPOSE: Published organ motion data have been collected from measurements of a limited number of points within the organ, the centroid, or the edge of the organ. These are derived from the spatial characteristics of respiratory induced motion; however, this approach does not consider non-rigid organ deformation. We propose a novel quantitative method for evaluating respiratory induced organ motion using Deformable Image Registration (DIR).

METHOD: Two phases from a 4-dimensional computed tomography (4D CT) dataset at maximum inspiration and expiration were each taken from five patients. The left and right lungs, esophagus, stomach, spinal cord, and liver were manually contoured in the end-expiration phase. The hybrid deformable registration algorithm of the RayStation treatment planning system (TPS) was used to deform the end-expiration phase to the end-inspiration phase. From this, the deformation vector field (DVF) was calculated. DVFs consist of DVFLR (left-right), DVFAP (anterior-posterior), and DVFSI (superior-inferior) as separate files. We calculated the vector volume histogram (VVH) and Lmax (maximum absolute vector of the organ) to evaluate every vector for each individual organ. We also measured respiratory organ motion from the position of the organ centroid in two phases.

RESULTS: VVH enabled us to find the absolute distance and volume of the organ contributing to motion points on the curve. Organ motion using the centroid method was smaller than Lmax using VVH. Using the centroid method, it is difficult to evaluate the deformable organ motion.

CONCLUSION: VVH may be a useful technique in evaluating organ volumetric change during respiratory organ motion.