P. Scheltens et al., VISUAL ASSESSMENT OF MEDIAL TEMPORAL-LOBE ATROPHY ON MAGNETIC-RESONANCE-IMAGING - INTEROBSERVER RELIABILITY, Journal of neurology, 242(9), 1995, pp. 557-560
We conducted an interobserver study to assess agreement on visual rati
ng of medial temporal lobe atrophy on coronal T1-weighted MRI. A total
of 100 studies of elderly individuals, using two different MRI techni
ques (spin echo and inversion recovery sequences), were analysed by fo
ur raters (three neurologists and one neuroradiologist) using a five-p
oint rating scale. Complete agreement was found in 37% of the total sa
mple. Interobserver agreement as expressed by kappa values was 0.44 (9
5% CI=0.34-0.54) and 0.51 (95% CI=0.41-0.61) for the two techniques. A
fter dichotomizing medial temporal lobe atrophy into present or absent
, a post hoc analysis revealed higher complete agreement (70%), with k
appa values of 0.59 (95% CI=0.51-0.67) and 0.62 (95% CI=0.48-0.075), f
or the two techniques (all four raters). From this study we conclude t
hat visual rating of medial temporal lobe atrophy on MRI in the corona
l plane yields fair to good agreement among observers. We recommend th
is type of visual rating for use in clinical settings when a quick jud
gement on the presence of medial temporal lobe atrophy is needed.