Skip to main content
ASTRO 2024 Main banner
Icon Legend
Presentation Icons
Event Add-on
Event Add-on
In Person Only
In Person Only
Off-site Event
Off-site Event
Location: Marriott Marquis
Location: Marriott Marquis
Back
Main Session

SS 14 - Phys 2: Advanced Imaging

180 - Assessing the Tumor Visibility and Respiratory Motion of Liver Cancer Patients on Multi-Task MR and 4D-CT Imaging for Radiation Treatment Planning

Monday, September 30, 2024
8:50 AM – 9:00 AM ET
Location: Room 152
O. W. Chau1, J. Everts1, J. Chen2, M. Feng1, J. Scholey3, X. Miao4, Y. Yang1, M. Ohliger1, K. Sheng1, Z. Fan2, and W. Yang1; 1University of California, San Francisco, San Francisco, CA, 2University of Southern California, Los Angeles, CA, 3The University of Pennsylvania, Philadelphia, PA, 4Siemens Medical Solutions USA Inc, Malvern, PA

Purpose/Objective(s): Liver tumors are difficult to visualize on standard 4D-CT images. A novel multi-task (MT)-MR imaging technique has been implemented on MR simulator, providing both T1 and T2 weighted 4D images in a single 8-min free-breathing scan for more accurate tumor delineation and motion evaluation. In this study, we report our early clinical experience in tumor visibility and motion assessment on both pre- and post-contrast MT-MR compared to post-contrast 4D-CT for liver cancer radiation treatment (RT) planning. Materials/

Methods: Motion p</span>hantom validation was performed on both 4D-CT and MT-MR simulating a wide range of tumor motion with breath per minute (BPM) of 6-15 BPM and amplitude of 10-30mm. Forty patients with liver tumors of HCC (n=14), ICC (n=5), CC (n=4) or metastatic (n=17) diagnoses underwent same-day RT simulation scans on the CT and 3T MRI. Image datasets of exhale breath-hold contrast-enhanced CT, post-contrast 4D-CT, and pre- and post-contrast MT-MR were acquired. Gross tumor volumes (GTVs) were delineated on the exhale breath-hold contrast enhanced CT scans and deformably propagated to each phase of the box-based registered MT-MR and 4D-CT dynamic dataset. The contrast to noise ratio (CNR) of liver tumors were compared between the 4D-CT and MT-MR image datasets at the exhale phase for both pre- and post-contrast datasets. GTV centroid motion ranges were determined from the end-exhale and end-inhale locations. Two sample t-tests were performed with a p value <0.01 considered statistically significant.
Results: Phantom sup-inf motion was validated with = 1.7mm for 4D-CT and = 2.02mm for MT-MR differences from the program motion range. For patient scans, significant better CNRs (p =0.007) were observed from MT-MR imaging for both T1 (mean ± SD of -16.0 ± 36.2 pre; -18.8 ± 23.1 post-contrast injection) and T2-weighted (25.4 ± 40.1 pre; 9.4 ± 25.0 post-contrast injection) image datasets compared to post-contrast 4D-CT imaging (-1.5 ± 3.0). Mean absolute motion ranges of MT-MR (RL: 1.28 ± 1.38mm, SI: 9.01± 4.87mm, AP: 3.36 ± 2.75mm) were observed, which was significantly greater on the sup-inf direction (p <0.001) compared to 4D-CT (RL: 1.21 ± 1.05mm, SI: 6.48 ± 3.52mm, AP: 2.97 ± 2.99mm).
Conclusion: The novel MT-MR sequence was successfully implemented on the MR simulator. The MT-MR with inherently co-registered T1 and T2 4D-MR may translate into improved tumor and motion definition in RT, compared to 4D-CT.