DIFFERENTIATION OF MILD COGNITIVE IMPAIRMENT CONDITION IN BRAIN MR IMAGES USING STRUCTURAL BIOMARKERS OF FORNIX

Abstract

Mild Cognitive Impairment (MCI) is a transitional phase between Normal Cognitive (NC) aging and several neurodegenerative diseases. The pathological manifestations due to MCI occur in different brain sub-anatomic structures, including fornix. Studies suggest that atrophy in fornix analyzed using Magnetic Resonance (MR) images is a reliable indicator for identifying MCI conditions. In this study, an attempt has been made to differentiate MCI condition using structural biomarkers extracted from segmented images of fornix. For this, T1- weighted brain MR images of NC and MCI subjects are considered and preprocessed using FreeSurfer. Reaction Diffusion Level Set (RDLS) method is employed to segment fornix from the preprocessed sagittal view MR images. Six statistical features, namely mean, variance, standard deviation, skewness, kurtosis, and entropy, are extracted from segmented images. Wilcoxon rank sum test and student t-test are employed to identify the significant features in differentiating NC and MCI. Results show that RDLS method is able to segment fornix structure from MR images of NC and MCI subjects. All the extracted features are found to be statistically significant (p-value < 0.01). The mean values of entropy, variance, standard deviation, and kurtosis features are found to be significantly lower in MCI subjects compared to NC. As alterations in fornix have the potential to predict early stages of the disease, proposed approach appears to be clinically significant. © 2024 IAE All rights reserved.