Click here to


Are you sure ?

Yes, do it No, cancel

Correlation Between Tumor and Vessel Motion in the Liver

S Jupitz*, A Shepard , P Hill , B Bednarz , University of Wisconsin, Madison, WI


(Sunday, 7/29/2018) 4:00 PM - 4:55 PM

Room: Room 202

Purpose: Intra-fraction motion management systems such as ultrasound and MR-based methods can rely on internal surrogate structures when the target motion is not easily visualized. This work aimed to assess the validity of using liver vessels as internal surrogates for the estimation of liver tumor motion.

Methods: Two-dimensional sagittal MRI cine data were collected from three treatment fractions on the ViewRay MRIdian (ViewRay, Inc., Cleveland, OH, USA). For each case, a liver tumor and at least one vessel were visible and tracked through 700 image frames. A 2D tracking approach utilizing block matching and multiple simultaneous templates was used to track the motion of each feature. The Pearson correlation coefficient was determined between each vessel and corresponding tumor in the horizontal (x) and vertical (y) motion directions. The distance between the centroids of the features and the tumors was also calculated to assess the role feature proximity may play in the relative correlation. Verification as to the validity of the tracked motion was performed by calculating the error between the tracked centroid position values and manually-defined ground truth annotations of the centroid position on 10% of the frames.

Results: The correlation coefficients between patient 1 fraction 1 (P1Fx1) tumor and vessel were 0.933 and 0.871 in the x and y directions, respectively. For P1Fx2, the correlation between tumor and vessel motion was 0.938 (x) and 0.834 (y). The correlation coefficients for P2Fx1 tumor and near vessel were 0.930 (x) and 0.918 (y), and for the tumor and far vessel were 0.928 (x) and 0.923 (y). The mean tracking errors obtained for motion validation were less than 1.24 mm relative to manual annotations for all features.

Conclusion: The motion of liver tumors and liver vessels is highly correlated, therefore vessels may make suitable surrogates for tracking tumor motion in the liver.

Funding Support, Disclosures, and Conflict of Interest: This work was supported by an NIH/NCI grant R01 CA190298.


Patient Movement, Gating, Image-guided Therapy


TH- External beam- photons: Motion management (intrafraction)

Contact Email