Expandable and rigid endorectal coils for prostate MRI: impact on prostate distortion and rigid image registration

Abstract

Endorectal coils (ERCs) are used for acquiring high spatial resolution magnetic resonance (MR) images of the human prostate. The goal of this study is to determine the impact of an expandable versus a rigid ERC on changes in the location and deformation of the prostate gland and subsequently on registering prostate images acquired with and without an ERC. Sagittal and axial T2 weighted MR images were acquired from 25 patients receiving a combined MR imaging/MR spectroscopic imaging staging exam for prostate cancer. Within the same exam, images were acquired using an external pelvic phased array coil both alone and in combination with either an expandable ERC (MedRad, Pittsburgh, PA) or a rigid ERC (USA Instruments, Aurora, OH). Rotations, translations and deformations caused by the ERC were measured and compared. The ability to register images acquired with and without the ERC using a manual rigid-body registration was assessed using a similarity index (SI). Both ERCs caused the prostate to tilt anteriorly with an average tilt of 18.5 degrees (17.4 +/- 9.9 and 19.5 +/- 11.3 degrees, mean +/- standard deviation, for expandable and rigid ERC, respectively). However, the expandable coil caused a significantly larger distortion of the prostate as compared to the rigid coil; compressing the prostate in the anterior/posterior direction by 4.1 +/- 3.0 mm vs 1.2 +/- 2.2 mm (14.5% vs 4.8%) (p < 0.0001), and widening the prostate in the right/left direction by 3.8 +/- 3.7 mm vs 1.5 +/- 3.1 mm (8.3% vs 3.4%) (p = 0.004). Additionally, the ability to manually align prostate images acquired with and without ERC was significantly (p < 0.0001) better for the rigid coil (SI = 0.941 +/- 0.008 vs 0.899 +/- 0.033, for the rigid and expandable coils, respectively). In conclusion, the manual rigid-body alignment of prostate MR images acquired with and without the ERC can be improved through the use of a rigid ERC.