Related Experiment Video
Updated: Jul 16, 2025
Magnetic Resonance Imaging of Multiple Sclerosis at 7.0 Tesla
Published on: February 19, 2021
Reliability of paramagnetic rim lesion classification on quantitative susceptibility mapping (QSM) in people with
Jack A Reeves1, Maryam Mohebbi1, Robert Zivadinov2
1Buffalo Neuroimaging Analysis Center, Department of Neurology, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, State University of New York, 100 High Street, Buffalo, NY 14203, USA.
Background:
Recent developments in iron-sensitive MRI techniques have enabled visualization of chronic active lesions as paramagnetic rim lesions (PRLs) in vivo. Although PRLs have potential as a diagnostic and prognostic tool for multiple sclerosis (MS), limited studies have reported the reliability of PRL assessment. Further evaluation of PRL reliability, through original investigations and review of PRL literature, are warranted.
Methods:
A single-center cohort study was conducted to evaluate the inter-rater reliability of PRL identification on quantitative susceptibiltiy mapping (QSM) in 10 people with MS, 5 people with clinically isolated syndrome, and 5 healthy controls. An additional systematic literature search was then conducted of published PRL reliability data, and these results were synthesized.
Results:
In the single-center study, both inter-rater and intra-rater reliability of per-subject PRL number were at an "Excellent" (intraclass correlation coefficient (ICC) of 0.901 for both) level with only 2-years lesion classification experience. Across the reported literature values, reliability of per-lesion rim presence was on average "Near perfect" (for intra-rater; Cohen's κ = 0.833) and "Substantial" (for inter-rater; Cohens κ = 0.687), whereas inter-rater reliability of per-subject PRL number was "Good" (ICC = 0.874). Only 4/22 studies reported complete information on rater experience, rater level of training, detailed PRL classification criteria, and reliability cohort size and disease subtypes.
Conclusion:
PRLs can be reliably detected both at per-lesion and per-subject level. We recommend that future PRL studies report detailed reliability results, including rater experience level, and use a standardized set of reliability metrics (Cohen's κ or ICC) for improved comparability between studies.
Insights
Paramagnetic rim lesions (PRLs) in multiple sclerosis (MS) are reliably detected using quantitative susceptibility mapping (QSM). This study confirms high inter-rater and intra-rater reliability for PRL assessment, crucial for their diagnostic potential.
Area of Science:
- Neuroimaging
- Radiology
- Neurology
Background:
- Iron-sensitive MRI techniques visualize chronic active lesions as paramagnetic rim lesions (PRLs) in vivo.
- Paramagnetic rim lesions (PRLs) show potential as diagnostic and prognostic tools for multiple sclerosis (MS).
- Limited studies have reported the reliability of PRL assessment, necessitating further evaluation.
Approach:
- A single-center cohort study evaluated inter-rater reliability of PRL identification using quantitative susceptibility mapping (QSM).
- The study included 10 people with MS, 5 with clinically isolated syndrome, and 5 healthy controls.
- A systematic literature search synthesized published PRL reliability data.
Key Points:
- Single-center study demonstrated excellent inter-rater and intra-rater reliability for per-subject PRL number (ICC=0.901).
- Literature review showed near-perfect intra-rater (Cohen's κ=0.833) and substantial inter-rater (Cohen's κ=0.687) reliability for per-lesion rim presence.
- Inter-rater reliability for per-subject PRL number was good (ICC=0.874).
Conclusions:
- Paramagnetic rim lesions (PRLs) are reliably detected at both per-lesion and per-subject levels.
- Future PRL studies should report detailed reliability results, including rater experience.
- Standardized reliability metrics (Cohen's κ or ICC) are recommended for improved study comparability.
