Encrypted login | home

Program Information

Clinical Assessment of Spatial Uncertainty for a Combined 0.35T MR and EBRT Treatment Machine

J Neylon

J Neylon*, K Sandler , D Du , Y Yang , J Lamb , K Sheng , D Low , M Cao , UCLA School of Medicine, Los Angeles, CA


TU-RPM-GePD-J(B)-6 (Tuesday, August 1, 2017) 3:45 PM - 4:15 PM Room: Joint Imaging-Therapy ePoster Lounge - B

Purpose: With the advent of magnetic resonance imaging guided external beam radiotherapy, there has been increasing interest in MR-only treatment planning. Spatial integrity in planning images is an important consideration for radiotherapy. Systematic distortion can be effectively minimized, however, patient-induced susceptibility and chemical shift distortions are difficult to address. This study attempts to estimate overall spatial uncertainty on clinical patient data for 0.35T MR images.

Methods: Ten patients with head-and-neck cancer that had undergone both CT and MR simulations were selected. The two simulation images were rigidly registered, and aligned images were generated with matching size and resolution. Landmark-based analysis was performed using an in-house software, allowing an anatomical expert to select identifiable landmarks on both CT and MR images and quantify spatial distortion using the Euclidean distance between each pair of landmarks. Landmarks were also tagged by tissue interface: soft tissue, bone-tissue, or air-tissue. For baseline comparison, a distortion phantom containing simulated bone and low density regions was also imaged and analyzed.

Results: Average error was 1.15+/-1.14mm for the distortion phantom, and 1.46+/-1.78mm for the patient data. For landmarks with non-zero errors, the histogram peaked at 1.5-2mm error, which corresponded with the in-plane resolution of the CT image (1.2-1.5mm). Larger errors were observed at bone-tissue interfaces with an average of 2.01+/-2.20mm, compared to 1.41+/-1.56mm and 0.88+/-1.24mm for the soft tissue and air-tissue, respectively. Geometric error also generally correlated to the in-plane radial distance from the image center.

Conclusion: Results indicate that spatial uncertainty remains in the MR images after systematic distortion corrections are applied. Even though the observed errors were small and should have little to no clinical significance, the uncertainty emphasizes the need for continued development of quantitative methods for assessing patient-specific spatial distortions as an important consideration in moving towards MR-only treatment planning.

Funding Support, Disclosures, and Conflict of Interest: Drs. Cao and Yang reported speaking honorarium from Viewray Inc., outside the submitted work. Dr. Lamb reported consulting fees from ViewRay Inc.

Contact Email: