Assessing large FOV MRI distortion phantoms and geometric distortion of eleven MRI scanners

Abstract

Introduction Geometric distortions are a known challenge of using magnetic resonance imaging (MRI) in radiotherapy. However, many phantoms are commercially available for assessing large field of view (FOV) distortions.

Objectives To develop a scoring system and use it to assess a number of large FOV phantoms. To assess large FOV MRI distortions in a number of scanners used for radiotherapy throughout England.

Methods A scoring system was developed and used to evaluate five phantoms: the GRADE, Wang et al, Quasar MRID 3D, Model 604 (CIRS), and Philips 2D geometric QA. To achieve the second objective, the GRADE phantom was used to assess the geometric distortion in eleven MRI scanners throughout England.

Results The scoring system evaluated the phantoms using nine criteria that included repeatability, FOV, safety, and ease of use. The Quasar, GRADE, CIRS, Philips 2D, and Wang et al. phantoms achieved first, second, third, fourth, and fifth place, respectively. The range of the mean distortions for all scanners at 100-150 mm from isocentre were 0.28-0.62 mm and at 200-250 mm from isocentre were 0.67-2.52.

Conclusions A scoring system was developed and used to successfully evaluate five phantoms. For small FOV applications such as the brain (distances from the isocentre of up to 150 mm), the mean distortions for all scanners would be < 1 mm. Except for one scanner, for large FOV applications such as the head and neck or pelvis (distances from the isocentre of up to 250 mm), the mean distortions would be < 2 mm.