Room: Exhibit Hall | Forum 5
Purpose: This work aims to investigate the correlation between diaphragm and target motions by using 4D-CT data for liver cancer patients, ultimately generating a correlation map on liver-segment.
Methods: Liver is clinically divided into 8 sections. This study was motivated from our observation that the internal marker (mostly, liver dome in absence of seed-marker) motion in liver radiation therapy differs from target motion, depending on the target location in liver. After scanning and sorting 4D-CT images in liver treatment, our protocol measures the full- (0 to top phase) and gating-motions (30-top and top-70 phases) of both tumor and liver dome for image-guide. Based on the measurements, it is possible to extract the correlating information between tumor and liver dome motions, depending on liver segments. To evaluate our assumption, a total of 119 liver patients having target volumes on different liver segments that underwent 4D-CT scanning and gating/full-motion measurement was collected. It compared target and liver dome motions in superior-inferior direction only. The degree of correlation was quantified by Pearson, Spearman and distance correlation metrics in this study.
Results: Pearson, Spearman, and distance correlation metrics had highest scores on S1 (0.75-0.80), and lowest on S3 (0.46-0.53) on Full-motion comparison, while they recorded greatest on S8, S4, and S2 (0.77-0.85), and lowest on S3 (0.24-0.47) on Gating-motion measurement. It was found that the correlation between tumor and dome motions gets bigger as the target volume is closer to the liver dome. The tendency is more explicit on gating-motion than on full-motion.
Conclusion: This study demonstrated that the correlation between target and internal marker (liver dome) motions is highly associated with the target volume location relative to the liver dome: the closer, the greater correlation. This information could be useful in determining gating window width in gated liver radiation therapy.
Not Applicable / None Entered.