PO11: MRI for Simultaneous Skin and Applicator Imaging in Surface Brachytherapy

Abstract

<h3>Purpose</h3> In skin High-Dose-Rate (HDR) brachytherapy, CT is commonly used for surface applicator imaging and treatment planning. However, since CT does not clearly depict skin, lesion depth is estimated from additional ultrasound examinations. MRI offers better soft tissue contrast than CT, and a PETRA MR sequence has recently shown the potential to visualize surface applicators and catheters, producing treatment plans comparable to CT-based ones. Additional skin detection on MR images would provide a complete information set for MR-only treatment planning. We investigated the feasibility of simultaneous skin and surface applicator imaging using MRI. <h3>Materials and Methods</h3> A Freiburg flap applicator (14 catheters with 24 beads each) was wrapped around the right shin of a healthy female volunteer. The volunteer was scanned in a 3T Siemens Vida using an UltraFlexLarge18 coil around and a Spine32 coil under her lower right leg. A 3D Dixon VIBE sequence (TR\TE 4.02\1.32 ms, field-of-view (FOV) 350 × 350 mm², 1.09 × 1.09 × 1.1 mm³ voxels, bandwidth 1116 Hz/pixel, acquisition time 1min 38s) was acquired coronally. Obtained in-phase, opposed-phase, water-only and fat-only image series were assessed for applicator and skin detectability in all three orientations. The series providing best visibility for both features was selected, and skin thickness was measured at five points around the leg (figure top left), on five reconstructed axial slices with distance 17.5 mm from each other, using a linear distance tool in the MIM software. <h3>Results</h3> The 3D properties of the Dixon VIBE sequence allowed for coronal acquisition orientation to best cover the area of interest, with reliable reconstruction in the axial plane employed for treatment planning. Both skin and applicator spheres were visible on the opposed-phase (figure top) and water-only (figure bottom) images in the three orthogonal planes. Opposed-phase images were selected for further analysis due to the consistent sphere signal and clear demarcation of the skin from the underlying bright subcutaneous fat. All applicator beads contained inside the FOV were visualized. Empty plastic catheters produced no signal, but their position in the center of the spheres could be determined by their negative contrast. Average skin thickness from assessed points and slices of the opposed-phase images was 0.18 ± 0.02 mm (mean ± standard deviation). These results lie inside the range of dermal and total skin thickness values reported for extremities of female subjects. The small standard deviation of the mean suggests that measured skin thickness was consistent among measurement locations. <h3>Conclusions</h3> An optimized 3D Dixon VIBE MR sequence offered simultaneous visualization of human skin and a surface brachytherapy applicator for the first time. While PETRA sequences can provide positive catheter contrast, the lack of catheter signal on Dixon VIBE images can also be used to determine their position inside the brighter applicator spheres. Dixon VIBE offered the additional advantage of skin visualization on opposed-phase and water-only series with spatial resolution around 1 mm, providing reasonable skin thickness assessments. As its acquisition time is short, Dixon VIBE can be combined with T2-weighted MR sequences for contouring organs at risk, presenting the possibility to achieve surface brachytherapy treatment simulations with a single imaging modality. Its implementation may pave the path towards MR-only treatment planning in surface HDR brachytherapy.