Author: Morgan Aire, Krystal M. Kirby, Olivia Magneson, David E. Solis 👨🔬
Affiliation: Louisiana State University, Mary Bird Perkins Cancer Center 🌍
Purpose: For soft-tissue abdominal tumors, a 4D-MR may provide improved tumor tracking over the clinical standard 4D-CT. This project aims to optimize a clinically available 4D-MR pulse sequence to achieve motion displacement accuracy which is not statistically different from 4D-CT.
Methods: Waveforms passed through a Quasar MRI4D Motion Phantom and oscillated a diluted gadolinium-filled sphere. Two different sinusoidal amplitudes with three different breathing periods were played out. The number of respiratory phase bins, slice thickness, radial views, and acquisition direction were individually varied from the clinical protocol. Scans were acquired for five repetitions on a 1.5T MR machine (Siemens Magnetom Sola) using a 3D radial “stack-of-stars" (StarVIBE) sequence, and CT images were acquired for comparison. Displacement values were measured using the full-width-half-maximum of a line profile taken through the center of the sphere at maximum and minimum amplitudes. The adjusted combination of best parameters was tested using sinusoidal waveforms and three regular, semiregular, and irregular patient breathing waveforms that were acquired from a Varian RPM system. The adjusted protocol was compared against the default MR protocol, CT protocol, and the waveform average input amplitude. Average amplitudes were determined using a MATLAB algorithm. Results were analyzed using ANOVA and linear mixed models.
Results: The adjusted MR protocol provided more accurate measurements than the clinical protocol for all sinusoidal cases, and there was no difference between adjusted MR measurements and CT measurements. For patient waveforms, measurements from the adjusted MR protocol falls 0.4mm above CT measurements on average and 1.5mm below the waveform mean using a linear mixed model. CT measurements fell 1.9mm below the waveform mean.
Conclusion: 4D-MR shows promise for motion management in radiotherapy, offering potential advantages over 4D-CT in abdominal tumor motion estimation. Future work will evaluate interplay effects between parameters and investigate failure cases for MR.