Interrater reliability of modified visual mrı rating scale assessing atrophy and white matter changes
Melek Kandemir Yilmaz1, Zehra Betul Yalciner2, Savaş M Tepe3
1Bodrum American Hospital, Department of Neurology, Istanbul, Turkey.
Background And Purpose:
Cortical atrophy and white matter changes are common findings on magnetic resonance imaging among elderly. Several visual scales have been proposed to evaluate these changes using neuroimaging. We have recently proposed a scale (Modified Visual Magnetic Resonance Rating Scale) recently which allows us to evaluate atrophy, white matter hyperintensities, basal ganglia and infratentorial infarcts together. Our aim in this study was to evaluate the interrater reliability of magnetic resonance visual assessment using this scale between two neurologists and a radiologist.
Methods:
Randomly selected 30 patients in different ages who underwent brain magnetic resonance imaging between January 2014 and March 2015 were included. Axial T1, coronal T2, and axial FLAIR sequences were visually scored by two neurologists and one radiologist separately. Sulcal, ventricular and medial temporal lobe atrophy, periventricular and subcortical white matter hyperintensities, basal ganglia and infratentorial infarcts were graded according to our scale. The interrater reliability and internal consistency analysis were evaluated by using intraclass correlation coefficient and Cronbach's alpha tests.
Results:
The interrater agreements vary between good to excellent. The interrater correlations are moderate to excellent. Interrater correlations were excellent between two neurologists, especially on ventricular atrophy, medial temporal atrophy, basal ganglia infarcts, infratentorial infarcts. When assessing ventricular atrophy, interrater correlations between individual raters were higher than sulcal atrophy. We found good correlations between neurologists and radiologist, and excellent correlations between the two neurologists for medial temporal atrophy. We found excellent interrater correlations between neurologists and radiologist for white matter hyperintensities.
Conclusion:
Our scale is a reliable tool assessing both atrophy and white matter hyperintensities with a good interrater reliability. Ventricular atrophy seems to be a more reliable marker than sulcal atrophy when assessing the atrophy on neuroimaging of a patient with memory decline. We think that the total score of the scale will also guide us in clinical practice.
