Encrypted login | home

Program Information

Impact of Immobilization Accessories On Geometric Distortions for H&N MRI Sequences

L Claps

L Claps1,2*, E Klein1,2 , Y Cao1,2 , F Diaz-Molina2 , J Saggese2 , (1) Hofstra University (2) Northwell Health


SU-I-GPD-J-73 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall

Purpose: Geometric distortions should be considered when MR images are used for head and neck (H&N) radiation therapy planning. We investigated the impact of radiotherapy immobilization accessories on MRI geometric distortions for various MRI sequences commonly used to image the H&N for IMRT and SBRT treatments.

Methods: We used a 3T Siemens Magnetom Skyra to acquire images of the ACR MRI accreditation phantom with a flexible 18-channel TIM coil. We imaged the phantom alone; positioned on an MRI-compatible table overlay; and positioned on a headrest with a leveling aquaplast accessory made from thermoplastic pellets for stability, atop the MRI-compatible table overlay and under a thermoplastic mask. Each image was acquired via T1-weighted turbo spin echo; T2-weighted SPACE gradient echo; and T1-weighted MPRAGE gradient echo sequences. For each sequence, we resampled each of the MR images of the phantom and immobilization accessories, to the MR-image of the phantom without accessories. Using a rigid fusion tool, we fused this resampled image set to a CT of the ACR phantom positioned with the same immobilization accessories.

Results: Initial results for 54 fusions have been obtained. We compared each of these fusions’ translational and rotational matrices to quantify variations in the magnitude of the geometric distortions across different sequences and accessory-combinations as described above. The data obtained shows that these devices do not produce geometric distortions.

Conclusion: We believe that this is a viable and productive method for evaluating geometric distortions produced in MR images. Future work will repeat measurements for different combinations of immobilization accessories and MR sequences, using different phantoms, and by alternative analytical means, i.e., volume, center of mass, and surface-area measurements.

Contact Email: