Imaging Protocol Optimization for MR-Only HDR Brachytherapy Treatment Planning
Abstract
Purpose
To optimize a simulation imaging protocol for MR-only treatment planning (MRTP) of high-dose rate (HDR) brachytherapy (BT) in patients diagnosed with prostate cancer.
Methods
Under IRB waiver, MRI sequences were added to the treatment simulation protocol for 3 patients undergoing HDR BT to optimize contrast between patient anatomy and interstitial needles containing Orion fiducial markers (C4 Imaging). Imaging was performed on a 1.5T Siemens Sola scanner. An iterative optimization process was performed on several axial 3D MRI sequences (T2 SPACE, CISS, and TRUFI) by maximizing spatial resolution, minimizing artifacts, and maintaining an SNR of at least 25. Optimized MR sequence parameters are listed in Table 1, including FOV, matrix, voxel size, TR/TE, averages, acceleration, and scan time. Contrast-to-noise ratio, CNR = |S1 - S2| / σ0, was measured for each sequence using mean signal in the needle (S1), mean signal in the surrounding tissue (S2), and standard deviation of the noise (σ0).
Results
Figure 1 displays axial images of each 3D sequence as well as sagittal MPRs of the TRUFI sequence for each patient, used to visualize the end of the needle tip. The T2-weighted SPACE sequence demonstrates negative contrast in the needle with respect to the surrounding tissue, while the T2/T1-weighted CISS and TRUFI sequences show positive contrast of the needle due to their balanced, steady state acquisitions. Figure 2 shows the average CNR of each sequence, with error bars denoting standard deviation; SPACE yielded the lowest CNR, while TRUFI yielded the highest.
Conclusion
This work demonstrated the optimization of a simulation imaging protocol for MR-only treatment planning of HDR brachytherapy in patients with prostate cancer. A T2/T1-weighted 3D TRUFI sequence outperformed a SPACE and CISS sequence when measuring the contrast of the needle to the surrounding tissue, thus offering better visualization of the needle trajectory and tip.