MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

Evaluation of Liver Motion in Real Time Using Diaphragm Tracking Versus Tracking Implanted Gold Markers for Stereotactic Ablative Radiation Therapy (SAbR)

N Hassan Rezaeian1*, (1) The University of Texas Southwestern Medical Ctr, Dallas, TX

Presentations

(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose:
Motion management is an essential component of stereotactic ablative radiation therapy (SAbR) for liver Hepatocarcinoma and metastasis treatment. Understanding the relation between noninvasive markers and implanted maker provides information about liver motion during treatment and their accuracy. In this work, we evaluated the correlation between the implanted gold markers and diaphragm motion in a custom 3D-printed phantom using both kV triggered and MV cine images.

Methods:

We used an in-house 3D-printing technique to cast a liver, stomach, and diaphragm. Lung behavior is emulated by aerated foam and a balloon. An in-house 3D-printed body, is used to encapsulate the liver, stomach, diaphragm, and lung. Respiratory motion was induced using a balloon and a cyclic air pump. Three gold markers were placed inside the liver near the hypothetical target. We used our in-house developed the projection marker matching method (PM3) to track gold markers using kV x-ray projection images acquired during treatment delivery. Concurrently, we acquired a series of ciné-MV images during treatment. We used a simple algorithm to detect the diaphragm. We cross-correlate the motion of diaphragm in superior-inferior to the 3D motion of target using a gold marker implanted in the liver. We used deep-learning approach to predict target motion in 3D from diaphragm motion accurately.

Results:
Based on our phantom study, we observed a strong correlation in superior-inferior (SI) direction and diaphragm motion. On average, the amplitude of motion estimated by gold markers in SI direction was 0.93±0.05 of the diaphragm motion. We observed the motion in anterior-posterior (AP) and medial-lateral (ML) direction strongly depends on the target size and depth of inspiration. Using our trained network, we observed an agreement of 95±6% and 94±6% in AP and ML.

Conclusion:
We have developed a quantitative approach to examine the accuracy of our trained model using a 3D-printed phantom.

Keywords

Treatment Techniques, Treatment Verification, Image-guided Therapy

Taxonomy

TH- External Beam- Photons: Motion management - intrafraction

Contact Email