Abstract
Objective: To determine whether dynamic contrast-enhanced magnetic resonance
imaging (DCE-MRI) evaluated using semi-automatic image processing software
can accurately assess synovial inflammation in rheumatoid arthritis (RA) knee
joints. Methods: In 17 RA patients undergoing knee surgery, the average grade of
histological synovial inflammation was determined from four biopsies obtained
during surgery. A preoperative series of T(1)-weighted dynamic fast low-angle shot
(FLASH) MR images was obtained. Parameters characterizing contrast uptake
dynamics, including the initial rate of enhancement (IRE), were generated by the
software in three different areas: (I) the entire slice (Whole slice); (II) a manually
outlined region of interest (ROI) drawn quickly around the joint, omitting large
artefacts such as blood vessels (Quick ROI); and (III) a manually outlined ROI
following the synovial capsule of the knee joint (Precise ROI). Intra- and interreader
agreement was assessed using the intra-class correlation coefficient (ICC).
Results: The IRE from the Quick ROI and the Precise ROI revealed high
correlations to the grade of histological inflammation (Spearman's correlation
coefficient (rho) = 0.70, p = 0.001 and rho = 0.74, p = 0.001, respectively). Intraand
inter-reader ICCs were very high (0.93-1.00). No Whole slice parameters
were correlated to histology. Conclusion: DCE-MRI provides fast and accurate
assessment of synovial inflammation in RA patients. Manual outlining of the joint
to omit large artefacts is necessary.
imaging (DCE-MRI) evaluated using semi-automatic image processing software
can accurately assess synovial inflammation in rheumatoid arthritis (RA) knee
joints. Methods: In 17 RA patients undergoing knee surgery, the average grade of
histological synovial inflammation was determined from four biopsies obtained
during surgery. A preoperative series of T(1)-weighted dynamic fast low-angle shot
(FLASH) MR images was obtained. Parameters characterizing contrast uptake
dynamics, including the initial rate of enhancement (IRE), were generated by the
software in three different areas: (I) the entire slice (Whole slice); (II) a manually
outlined region of interest (ROI) drawn quickly around the joint, omitting large
artefacts such as blood vessels (Quick ROI); and (III) a manually outlined ROI
following the synovial capsule of the knee joint (Precise ROI). Intra- and interreader
agreement was assessed using the intra-class correlation coefficient (ICC).
Results: The IRE from the Quick ROI and the Precise ROI revealed high
correlations to the grade of histological inflammation (Spearman's correlation
coefficient (rho) = 0.70, p = 0.001 and rho = 0.74, p = 0.001, respectively). Intraand
inter-reader ICCs were very high (0.93-1.00). No Whole slice parameters
were correlated to histology. Conclusion: DCE-MRI provides fast and accurate
assessment of synovial inflammation in RA patients. Manual outlining of the joint
to omit large artefacts is necessary.
Original language | English |
---|---|
Journal | Scandinavian Journal of Rheumatology |
Volume | 41 |
Issue number | 2 |
Pages (from-to) | 89-94 |
ISSN | 0300-9742 |
DOIs | |
Publication status | Published - 2012 |