Performance comparison of 1.5-T endorectal coil MRI with 3.0-T nonendorectal coil MRI in patients with prostate cancer

Abstract

Rationale and objectives: To compare prostate morphology, image quality, and diagnostic performance of 1.5-T endorectal coil magnetic resonance (MR) imaging (MRI) and 3.0-T nonendorectal coil MRI in patients with prostate cancer.

Materials and methods: MR images obtained of 83 patients with prostate cancer using 1.5-T MRI systems with an endorectal coil were compared to images collected from 83 patients with a 3.0-T MRI system. Prostate diameters were measured, and image quality was evaluated by one American Board of Radiology (ABR)-certified radiologist (reader 1) and one ABR-certified diagnostic medical physicist (reader 2). The likelihood of the presence of peripheral zone cancer in each sextant and local extent was rated and compared to histopathologic findings.

Results: Prostate anterior-posterior diameter measured by both readers was significantly shorter with 1.5-T endorectal MRI than with 3.0-T MRI. The overall image quality score difference was significant only for reader 1. Both readers found that the two MRI systems provided a similar diagnostic accuracy in cancer localization, extraprostatic extension, and seminal vesicle involvement.

Conclusions: Nonendorectal coil 3.0-T MRI provides prostate images that are natural in shape and that have comparable image quality to those obtained at 1.5 T with an endorectal coil, but not superior diagnostic performance. These findings suggest an opportunity exists for improving technical aspects of the 3.0-T prostate MRI.

Keywords: Prostate cancer; endorectal coil; image quality; magnetic resonance imaging; tumor localization; tumor staging.

Figure 1

Research design flow chart. In order to evaluate whether 3.0 T MRI without an endorectal coil is clinically usable, the ACRIN imaging archive with endorectal 1.5 T MRI was compared to our in-house 3.0 T MR imaging study without an endorectal coil.

Figure 2

Prostate diameter measurement results by 2 readers. LR diameter was generally the longest, and AP shortest, among the three diameters. AP diameter at 1.5 T endorectal MRI was significantly shorter than that at 3.0 T due to the rectum dilation for an endorectal coil insertion and fixation as reviewed by both readers.

Figure 3

100% stacked column plot shows the percentage of subjects with different scores, as evaluated by two readers. Highest score was given to the majority of subjects at either 1.5 T endorectal or 3.0 T MRI.

Figure 4

Image artifacts on 1.5 T endorectal and 3.0 T MRI. The typical image artifacts from 1.5 T endorectal MRI included ghosting (a) and coil-related SNR decrease (b). The typical artifacts from 3.0 T MRI are signal graininess (c) and rectum motion artifact (d).

Figure 5

Receiver operating characteristic curves of 1.5 T endorectal MRI diagnostic performances versus 3.0 T MRI performances for both readers in tumor localization (a), extraprostatic extension (b), and seminal vesicle involvement (c).