Volumetric measurement of the entorhinal cortex: A reliability analysis

Abstract

Background: The pathology of Alzheimer’s Disease (AD) is believed to appear first in the transentorhinal/entorhinal cortex. The usefulness, however, of entorhinal volumetric measurements to discriminate among normal and prodromal stage AD patients is debated. Some attribute this reduced discrimination ability to difficulties with magnetic resonance imaging (MRI) depiction and measurement of this structure (Xu Y, Jack et al., 2001). Objective(s): The purpose of this study was to assess intra and inter rater reliability of three different entorhinal cortex volumetric measurement approaches to identify if one of these protocols is most appropriate for visualization and quantification purposes. Methods: Volumetric MRI scans (3.0T Siemens; MP Rage sequences) were obtained on 24 nondemented older adults (mean age 71.09 6.44; Mini Mental State Exam 28.96 1.19). Volumetric analysis was conducted with MEASURE (Barta et al., 1997). Each entorhinal cortex was measured twice using three different volumetric protocols: 1) Insausti et al (1998) that uses collateral sulcus (CS) depth to define ROI boundary; 2) Goncharova et al (2001) that excludes the CS; 3) an experimental “Full” technique that includes the entire CS. Conclusions: Intraclass correlation coefficients showed high intra-rater reliability for all three entorhinal protocols (Insausti, r .98; Goncharova, r .97; Full, r .96). Inter-rater reliability was highest with the Full technique (r .95) and lowest with the Insausti et al protocol (r .86). There was adequate association among all three techniques (ICC among all three approaches, r .92). Our preliminary findings suggest that the entorhinal cortex can be measured reliably on MRI using any of three volumetric protocols. The lowest rater reliability was identified for the Insausti et al (1998) protocol. This protocol requires a rater to subjectively alter their measurement approach based on CS depth. Strengths and weakness of measurement approach along with the range of volumes will be discussed.