Assessment of total geometric distortion in MR images used in intracranial stereotactic radiosurgery

Abstract

Introduction In Stereotactic Radiosurgery (SRS), MR-images are widely used for target localization and delineation in order to take advantage of the superior soft tissue contrast they exhibit. However, spatial dose delivery accuracy may be deteriorated due to geometric distortions inherent in MRI. Purpose To develop and implement a quality control procedure for the assessment of total geometric distortion in MR-images clinically employed in SRS treatment planning. Materials and methods A specially designed acrylic-based phantom was utilized. In particular, the phantom encompasses five planes on which 947 3-mm diameter holes are drilled. The centers of mass of these holes serve as control points for geometric distortion detection. In terms of imaging the phantom is both CT and MR compatible. The phantom was MR-imaged at 1.5T (GE-Optima) and 3T (Siemens-Skyra). CT-detected control point locations served as the reference distribution. Distortion evaluation was performed for three typical pulse sequences used for SRS treatment planning using in-house developed software. The effect of specific imaging parameters on geometric distortion was also investigated. Results Overall geometric accuracy of the 1.5T scanner was found slightly superior than the 3.0T scanner for the specific sequences investigated (mean distortion values: 0.6 mm and 0.8 mm, respectively). However, the 1.5T unit exhibits increased distortion on z -axis in regions distant from the scanner’s isocenter (maximum detected distortion was 2.7 mm). Accuracy of the developed distortion detection algorithm was estimated to be 0.24 mm. Conclusion Phantom and methods developed in this multicenter study were found efficient for distortion characterization of MR pulse sequences used in SRS.