M Rabe¹*, C Paganelli², M Riboldi³, D Bondesson⁴, M Schneider⁴, T Chmielewski⁵, J Dinkel⁴, M Reiner¹, G Landry^1,3, K Parodi³, C Belka^1,6, F Kamp¹, C Kurz^1,3, (1) Department of Radiation Oncology, University Hospital, LMU Munich, Munich, Germany (2) Dipartimento di Elettronica, Informazione e Bioingegneria, Politecnico di Milano, Milano, Italy (3) Department of Medical Physics, Ludwig-Maximilians-Universitaet Muenchen (LMU Munich), Garching (Munich), Germany (4) Department of Radiology, University Hospital, LMU Munich, Munich, Germany (5) ViewRay Inc., Oakwood Village, OH (6) German Cancer Consortium (DKTK), Partner Site Munich, Munich, Germany

• 
  M Rabe

(Tuesday, 7/14/2020) 3:30 PM - 5:30 PM [Eastern Time (GMT-4)]

Room: Track 2

Purpose:
Beam-on imaging of currently available MR-Linacs for monitoring moving targets is currently limited to sagittal cine-MRI, where out-of-plane motion can lead to target localization errors. The study’s purpose was to experimentally validate a propagation method that estimates time-resolved anatomy based on orthogonal cine-MRI.

Methods:
Ex-vivo porcine lungs were mounted in a dedicated phantom to reproducibly simulate periodic breathing motion in realistic geometry. Four artificial nodules (1.2-12.6cc) were injected into the lungs to mimic moving targets. The phantom was scanned with a research version of a commercial 0.35T MR-Linac. Respiratory-correlated 4D-MRI were acquired and served as ground truth images (GT). Series of interleaved orthogonal slices (sagittal and coronal) intersecting the injected targets were acquired at 8Hz (bSSFP; 3.5mm in-plane resolution; 5mm slices). Eight datasets for three porcine lungs were acquired at breathing frequencies of 6-15 cycles/minute, nodule centroid motion amplitudes of 3-23mm and different motion extents in right-left direction. The mid-exhale image of GT (REF) was deformably registered to the orthogonal slices. The resulting deformation fields were extrapolated to 3D and applied to REF to create estimated 4D-MRI (EST) at 4Hz. Differences of nodule centroid positions in GT and EST were calculated to assess the accuracy of the propagation method.

Results:
Among all datasets and series, the mean estimation error over ten breathing cycles was (2.1±1.1)mm (±1s) for nodules intersected by orthogonal slices and (3.0±1.4)mm for the remaining nodules. Estimation errors increased with increasing distance to the orthogonal planes due to limitations of the method and towards the edges of the field-of-view due to geometric distortions.

Conclusion:
The porcine lung phantom allows acquisition of realistic patient-like data with high reproducibility. The sub-voxel-size agreement between GT and EST volumes suggests that the propagation method could eventually be used to estimate time-resolved 4D anatomy for intrafractional treatment adaption and post-treatment dose accumulation.

Funding Support, Disclosures, and Conflict of Interest: This work was supported by the German Research Foundation (DFG) within the Research Training Group GRK 2274.

Keywords

MRI, Phantoms, Lung

Taxonomy

IM/TH- MRI in Radiation Therapy: MRI/Linear accelerator combined- IGRT and tracking

Contact Email