Assessment of distortion in magnetic resonance imaging for stereotactic radiosurgery

Abstract

Background and purpose: MR images are widely used in routine practice including radiotherapy planning for stereotactic radiosurgery. Image distortion is one of the major disadvantages of MRI, affecting accuracy. Cylindrical phantom was used for distortion measurement of T1-weighted pulse sequences by a 1.5 T NVi/CVi, General Electric Medical Systems. Materials and methods: The phantom was scanned with three techniques providing T1W images contrast SE, FSE and 3D FSPGR. The images were evaluated with varying parameters (TR, TE, phase encoding, bandwidth and NEX). Image distortion was measured in each zone of the slices position at -60, - 30, 0, +30, +60 mm. Results: The FSE T1-weighted image exhibited errors less than 2.0 mm in the transverse planes. No significant difference in image distortion was demonstrated among all techniques providing T1W images. Bandwidth is a single parameter affecting image distortion. The image distortion of the middle zone (120x120 mm) was within acceptable range for SRS (<1mm). The displacements of the periphery zone are greater than the middle zone about 0.5 mm. At the edge of the field (-60, +60 slice positions) showed greater displacement than the centre of the coil. Conclusions: Magnetic field inhomogeneity was a major cause of image displacement in all T1W techniques. The displacement at the centre of the image was acceptable for SRS while the spatial accuracy in periphery zone exceeded the acceptable range. Increasing the receiver bandwidth to 32 kHz is suggested to decreased distortion.