Multimodal Prediction of Alzheimer's Disease Severity Level Based on Resting-State EEG and Structural MRI

Belmir Jesus,Tiago H Falk,Raymundo Cassani,Marco Cecchi,William J Mcgeown,K C Fadem

doi:10.3389/fnhum.2021.700627

Abstract

While several biomarkers have been developed for the detection of Alzheimer's disease (AD), not many are available for the prediction of disease severity, particularly for patients in the mild stages of AD. In this paper, we explore the multimodal prediction of Mini-Mental State Examination (MMSE) scores using resting-state electroencephalography (EEG) and structural magnetic resonance imaging (MRI) scans. Analyses were carried out on a dataset comprised of EEG and MRI data collected from 89 patients diagnosed with minimal-mild AD. Three feature selection algorithms were assessed alongside four machine learning algorithms. Results showed that while MRI features alone outperformed EEG features, when both modalities were combined, improved results were achieved. The top-selected EEG features conveyed information about amplitude modulation rate-of-change, whereas top-MRI features comprised information about cortical area and white matter volume. Overall, a root mean square error between predicted MMSE values and true MMSE scores of 1.682 was achieved with a multimodal system and a random forest regression model.

Highlights

Alzheimer’s disease (AD) is a neurodegenerative condition that progressively affects cognitive functioning by impairing nerve cell function in the brain (Gross et al, 2016)
We report the Pearson and Spearman correlation coefficients calculated between the predicted Mini-Mental State Examination (MMSE) and observed MMSE scores per cross-validation trial for the multimodal system
With the advances seen with deep neural network based analyses of EEG (Roy et al, 2019) and MRI signals (Jo et al, 2019), deep multimodal architectures should be explored once more data is made available to the research community. Features such as beta-amyloid or tau may be used to illustrate the progression of the hallmark neuropathology relating to Alzheimer’s disease, additional combinations of neural features captured through EEG and structural MRI may provide further insight on the neurodegeneration and neuronal injury that occurs with disease progression

Summary

Introduction

Alzheimer’s disease (AD) is a neurodegenerative condition that progressively affects cognitive functioning by impairing nerve cell function in the brain (Gross et al, 2016). This process may start 20 years or more before symptoms become evident (Alzheimer’s Association, 2019). Given the co-occurrence of neuropathological and structural brain changes that occur alongside cognitive impairment, Multimodal Prediction of Alzheimer’s Disease Severity research has focused on developing biomarkers of the disease that may help to increase confidence in the diagnosis. It is thought that diseasemodifying therapeutic interventions should be most effective when administered in the early stages of the disease, before neuronal loss occurs (Jedynak et al, 2012), early detection is crucial

Objectives

Methods

Results

Conclusion