2371 - Development of Volumetric Quantitative Evaluation Software for Assessing Respiratory-Induced Target Motion

H. Miura^1,2, M. Tanooka³, S. Ishihara¹, M. Kenjo¹, M. Nakao¹, S. Ozawa¹, and M. Kagemoto¹; ¹Hiroshima High-Precision Radiotherapy Cancer Center, Hiroshima, Japan, ²Department of Radiation Oncology, Institute of Biomedical & Health Science, Hiroshima University, Hiroshima, Japan, ³Takarazuka City Hospital, Takarazuka, Japan

Purpose/Objective(s): Most respiratory-induced target motion data have been measured at only a few points, the centroid and the edge. Due to organ deformation, volumetric measurement is more appropriate than point measurement. Previously, we proposed a quantitative method to evaluate respiratory-induced organ motion using deformable image registration (DIR), called vector volume histogram (VVH). However, the process of VVH was very complicated and no one can use the VVH because the deformation vector field (DVF) was calculated on the treatment planning system (TPS) and the file format was converted by in-house software. We developed a volumetric quantitative evaluation user-friendly software to evaluate respiratory-induced organ motion using DIR. Materials/

Methods: The B-spline-based DIR algorithm was used to compute the DVF, which includes DVF_LR (left-right), DVF_AP (anterior-posterior), and DVF_CC (cranio-caudal). The VVH function was written in Python. Two sets of images were required for DIR in the VVH software: a reference image set and a registered image set. The VVH was a calculation method similar to the dose volume histogram. In the VVH software, the user could change the result of the direction by selecting a combo box. Other features included a calculation index, changing the range of the x and y axes, a tracking bar, and exporting to a CSV file. A displaced target within a moving phantom was used to evaluate the performance of the VVH system. The 2 cm diameter target was systematically displaced 5, 10, 15, and 20 mm in the CC direction. To evaluate respiration-induced target motion, the VVH method was applied to the inhalation and exhalation phases of 4D CT scans in 5 patients with lung cancer. L_5% (length at 5% volume) and L_50% (length at 50% volume) were calculated to evaluate target motion. Target centroid was measured for comparison with VVH methods.
Results: In the phantom study, L_5% and L_50% were 5 mm versus 4.7 mm, 9.7 mm versus 10.3 mm, 14.7 mm versus 14.9 mm, and 19.7 mm versus 20.0 mm for displacements of 5, 10, 15, and 20 mm, respectively. In the patient study, the average difference between the L_5% and L_50% methods and the centroid methods was as follows: L_5% and L_50% were 0.2 mm versus 0.5 mm, 0.3 mm versus 0.6 mm, and 2.0 mm versus 0.8 mm in the LR, AP, and CC directions, respectively. Although it was difficult to decide which index to use, the centroid and L_50% results were close. The performance of VVH depends on the accuracy of the DIR algorithm and the extent of anatomic changes.
Conclusion: The performance of the VVH software was demonstrated by verifying the phantom and the patient with tumor motion. The VVH software provided a volumetric quantitative assessment of respiratory-induced target motion, which in its clinical application can provide strategy-decision at the time of treatment planning.