Accuracy of magnetic resonance imaging for measuring maturing cartilage: A phantom study

Authors

  • Jennifer R. McKinney University of Toronto; The Hospital for Sick Children; Department of Diagnostic Imaging
  • Marshall S. Sussman University of Toronto; Toronto General Hospital; Department of Medical Imaging
  • Rahim Moineddin University of Toronto; Department of Family and Community Medicine
  • Afsaneh Amirabadi University of Toronto; The Hospital for Sick Children; Department of Diagnostic Imaging
  • Tammy Rayner University of Toronto; The Hospital for Sick Children; Department of Diagnostic Imaging
  • Andrea S. Doria University of Toronto; Toronto General Hospital; Department of Medical Imaging

DOI:

https://doi.org/10.6061/clinics/2016(07)09

Abstract

OBJECTIVES: To evaluate the accuracy of magnetic resonance imaging measurements of cartilage tissue-mimicking phantoms and to determine a combination of magnetic resonance imaging parameters to optimize accuracy while minimizing scan time. METHOD: Edge dimensions from 4 rectangular agar phantoms ranging from 10.5 to 14.5 mm in length and 1.25 to 5.5 mm in width were independently measured by two readers using a steel ruler. Coronal T1 spin echo (T1 SE), fast spoiled gradient-recalled echo (FSPGR) and multiplanar gradient-recalled echo (GRE MPGR) sequences were used to obtain phantom images on a 1.5-T scanner. RESULTS: Inter- and intra-reader reliability were high for both direct measurements and for magnetic resonance imaging measurements of phantoms. Statistically significant differences were noted between the mean direct measurements and the mean magnetic resonance imaging measurements for phantom 1 when using a GRE MPGR sequence (512x512 pixels, 1.5-mm slice thickness, 5:49 min scan time), while borderline differences were noted for T1 SE sequences with the following parameters: 320x320 pixels, 1.5-mm slice thickness, 6:11 min scan time; 320x320 pixels, 4-mm slice thickness, 6:11 min scan time; and 512x512 pixels, 1.5-mm slice thickness, 9:48 min scan time. Borderline differences were also noted when using a FSPGR sequence with 512x512 pixels, a 1.5-mm slice thickness and a 3:36 min scan time. CONCLUSIONS: FSPGR sequences, regardless of the magnetic resonance imaging parameter combination used, provided accurate measurements. The GRE MPGR sequence using 512x512 pixels, a 1.5-mm slice thickness and a 5:49 min scan time and, to a lesser degree, all tested T1 SE sequences produced suboptimal accuracy when measuring the widest phantom.

Downloads

Download data is not yet available.

Downloads

Published

2016-07-01

Issue

Section

Clinical Sciences

How to Cite

Accuracy of magnetic resonance imaging for measuring maturing cartilage: A phantom study . (2016). Clinics, 71(7), 404-411. https://doi.org/10.6061/clinics/2016(07)09