Abstract
Introduction: Recently, a great emphasis has been placed on early diagnosis of Alzheimer's disease (AD). The new diagnostic criteria for AD involve new methods such as detection of structural and metabolic changes in the brain.
These include examination of hippocampal volume. Scheltens et al. introduced a visual rating scale for hippocampal atrophy assessment.
Our aim was to determine inter-rater variability and to test practical application of this scale. Methods: MRI scans of 70 elderly persons with cognitive impairment and persons classified as cognitively normal were assessed by eight investigators.
The investigators had a different degree of experience with Scheltens visual rating scale. Following a brief training, the investigators were asked to practice on 20 MRI brain scans in coronal plane.
Correct scores were disclosed to all investigators. Subsequently, the investigators evaluated 70 study MRI scans.
They were unaware of the participants' cognitive status. The variability was calculated for all investigators together and after dividing them into "experienced" (n = 4) and "inexperienced" (n = 4) group, where "inexperienced" refered to no previous knowledge of this visual rating scale.
Results: Inter-rater agreement (kappa) was very high among all investigators for the right (K = 0.87) and left (K = 0.88) hippocampus. After dividing the raters into experienced and inexperienced, the inter-rater variability continued to be high in both groups for the right and left hippocampus.
Conclusion: Scheltens visual rating scale is a simple and easy to use tool for hippocampal atrophy assessment even for inexperienced investigators. The final assessment variability was relatively low.
We can recommend this visual rating scale for routine clinical use.