Gray level co-occurrence matrix structural MRI texture analysis for Alzheimer’s disease prediction

Abstract

Alzheimer’s disease (AD) is a progressive and irreversible neurodegenerative condition that increasingly impairs cognitive functions and daily activities. Affecting millions worldwide, AD’s symptoms gradually worsen over time, starting years before they become evident with mild cognitive impairment (MCI) states. However, once symptoms manifest, the effectiveness of most therapeutic options diminishes, underscoring the urgent need for early detection methods. Given the incurable nature of AD and its profound impact on the elderly, early diagnosis and intervention are crucial, focusing on delaying disease progression and improving patients’ quality of life. Thus, this work aims to develop an automatic method to detect AD in 3 different stages, namely, control (CN), MCI, and AD itself, utilizing structural magnetic resonance imaging (sMRI) images, one of the adjunct techniques to diagnosis. For such purpose, brain sMRI images from the ADNI database were pre-processed and a set of 22 texture statistical features from the gray level co-occurrence matrix (GLCM) were extracted from various slices corresponding to the different anatomical planes. Different combinations of features have been used to feed classic machine learning (cML) algorithms to analyze their discrimination power between groups. The cML algorithms achieved the following classification accuracies: 85.2% for AD vs. CN, 98.5% for AD vs. MCI, 95.1% for CN vs. MCI, and 87.1% for All vs. All. These results are particularly significant in the field of AD classification, providing a comprehensive set of metrics beyond mere accuracy. Moreover, when compared with a similar state-of-the-art sMRI study, this approach was able to outperform the classification accuracies by 10.8%, 6.9%, and 11.8% in the AD vs. MCI, CN vs. MCI, and All vs. All study groups, respectively.

Date of Award	16 Jul 2024
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Pedro Miguel Rodrigues (Supervisor)